A Disabled Raider's Review of Patch 7.0 Duty Design

Navigation Links

EN

This is a LONG one. And unfortunately I was unaware of the limit of 20 images per post when I made my sections, so I had to split up the subsection reviewing the Patch 7.00 MSQ duties. Sorry, Japanese readers, I would have reserved more posts if I knew that in advance!

Introduction

1) Design Principles

2.1) 7.0 Duty Review - MSQ (two posts)

2.2) 7.0 Duty Review - Optional Dungeons

2.3) 7.0 Duty Review - Normal Raids

2.4) 7.0 Duty Review - High-End

3) Honey B. Lovely, Redesigned (this is the exciting example, start here if you want to see the principles in action)

JP

これは長い記事です。そして残念なことに、セクションを作ったときに投稿ごとに画像が 20 枚までという制限に気付かなかったので、パッチ 7.00 MSQ の任務をレビューするサブセクションを分割する必要がありました。申し訳ありませんが、日本の読者の皆さん、事前にそのことを知っていたら、もっと多くの投稿を予約していたのに!

はじめに

1) 設計原則

2.1) 7.0 任務レビュー - MSQ (2 つの投稿)

2.2) 7.0 任務レビュー - オプション ダンジョン

2.3) 7.0 任務レビュー - 通常レイド

2.4) 7.0 任務レビュー - ハイエンド

3) Honey B. Lovely、再設計 (これはエキサイティングな例です。原則の実際の動作を確認したい場合はここから始めてください)

Dear Respected Forum Master / Webmaster,

Would you please kindly forward this letter to Yoshi-P and the duty design team?

Thank you very much for your help.

Sincerely,
Aurelle

--------------------------------

Dear Yoshi-P and the Final Fantasy XIV development team,

I must first apologize for not bringing you this feedback more swiftly. I experienced health and safety issues with the graphical update, and was thus unable to partake of any Dawntrail content at all until over three weeks after the beginning of Dawntrail Early Access, when I found a postprocessing method that enabled me to see the game safely. As a result, I was unable to complete the Patch 7.00 Main Scenario Quests (MSQ) until the day before the release of Patch 7.05 and AAC Light-Heavyweight (Savage), so I judged it better for me to first complete the Savage tier and review all of Patch 7.0's duty design together and thoroughly.

I see that you have been working hard to bring players enjoyable duties, in particular responding to feedback about Endwalker's duty design. I am not advocating for a return to Endwalker's duty design, nor for making the game easier in general. In fact, I have completed every single duty available in Patch 7.0, and each and every one of them has made me bored in three completions or less!

I would very much like to attempt and complete harder duties, such as Futures Rewritten (Ultimate) (FRU) coming in Patch 7.11. Indeed, I completed The Omega Protocol (Ultimate) (TOP) during Endwalker, so it would be the logical next step for me to aim to complete FRU on-patch before the release of Patch 7.2.

But I will not get to do that, and may not get to complete FRU at all. And that is also because of Patch 7.0's duty design.

You see, I am a disabled player. Without getting into medicine or biomechanics that may be difficult for readers, especially in translation, my muscles respond slower than those of a "normal" human body. I have to use both arms through the shoulders in order to give inputs fast enough to play competently at all, and even then it puts a bit of a damper on my reaction time. Obviously not that much of a damper, since I can do TOP; but enough that during progression (practicing the mechanics of a high-end duty) I study the fastest mechanics more to learn to process them efficiently. Enough that I prefer to avoid reliance on "follow the leader" or callouts, and instead solve everything myself, because the extra delay from reacting to someone else's processing the mechanic can get my character KOed while everyone else makes it to the safespot. Enough that even with all the work I put in, I move like someone who has been slowed down by simple aging or high latency.

The vast majority of Duty Finder players could play with me and never notice. (I have often been "the leader" in Duty Finder, teaching parties the mechanics!) Even the vast majority of raiders could play with me and never notice unless I told them - though I do wind up telling raid groups sometimes, when they ask me to switch to a job I avoid raiding on due to poor input patterns for my body, or ask me to go past the raiding time limits I set to avoid injuring myself.

Why the time limit? Having seven other players' enjoyment relying on my giving all the inputs fast enough means that I hold myself to a higher standard. My body can only keep that up in a healthy way for so long; remember, I am already using both arms in full to play at all. I had to build up my stamina gradually when I started playing the game, and if I go beyond my baseline of raiding hours I will start having to stretch frequently, eat more and with a very particular nutrient balance, be more strategic about when I raid to give my body recovery time... just like an athlete has to when training intensively. (I cannot keep up hours beyond my baseline for more than a few weeks, either.) I sure notice the difference between me and an able-bodied player, because I am the one putting in all the work behind the scenes to still be able to play.

(Most people do not notice when meeting me on Earth either, unless tasks requiring high physical speed or high strength come up. I can hold a variety of jobs when offices do not insist on every last employee being able to lift heavy things or work unproductive overtime. There are some sports I can do for fitness, after consulting with my doctors. I can even drive a car - I passed all my road tests and my emergency situations training without knowing I was disabled, because my unconscious compensating skills were good enough on their own. (For example, it was years later that I found out that most drivers use their shoulder check as the first sign that a car is in their blind spot, not a confirmation check on continuously tracking the surrounding traffic.) There are lots of people whose disabilities result in worse performance than mine.)

Before Dawntrail, my baseline raiding hours were 9 hours a week. I had never gone beyond 12 hours a week until the very end of Endwalker, when I pushed to 14 hours a week in order to complete TOP before Dawntrail Early Access - despite progressing in Party Finder. At the time, I thought what I had to do for that was incredibly intensive. A rigid schedule of 2 hours every single day, not counting the time I spent practicing outside the instance using completion video footage and a striking dummy, with no days off. (Make no mistake, Party Finder does not make that easy - I had to tally up my time with each party each day.) Stretching not just immediately before, at the 1-hour mark during, and immediately after each session, but several additional times throughout each day. Eating ravenously and still having to carefully plan to pack all the protein, fats, and micronutrients into my diet. I even barely typed outside of my party chat, in order to reduce strain on my hands. Even with my Earth schedule allowing all these measures, the only reason I was willing to push my body so hard was that I knew I would get a natural recovery week - or two, if I needed it - from Dawntrail's launch and the accompanying MSQ.

Spending 9 hours a week progressing AAC Light-Heavyweight (Savage) was harder on my body in every way. Not just in stretching and food and sleep; Dawntrail was the first time my hands have hurt from playing any game, despite everything I do to care for myself. (I could not just stop as soon as I felt any pain, since the pain came afterwards - and took a full month of deliberate care to stop.) Dawntrail is the first expansion where I have had to contemplate applying my time limit to regular duties, not just raiding. Dawntrail is the first expansion that has effectively pushed my stamina down - rather than it slowly increasing with exercise - and caused me to seriously contemplate lowering my time limit to match.

In light of these results of Patch 7.0's duty design, I cannot know whether I am physically capable of completing FRU at all without personally analyzing the entire duty in depth - which I of course cannot do until well after its release, when large amounts of completion footage are available to properly understand the many possibilities arising from mechanic randomness. Likewise, even if I knew that I could physically do FRU at all, I cannot know whether it would be sustainable for me to do at least 9 hours a week of FRU progression without an even deeper analysis. (I was happy to commit to TOP and even Dragonsong's Reprise (Ultimate) (DSR) before their respective launches, because I had my baseline of 9 hours a week by the beginning of Endwalker.)

And with less than 9 hours a week, clearing an Ultimate in its launch patch becomes statistically highly unlikely, even with a static raid group (often shortened to "static") that goes from launch day through the remainder of the patch and uses its time efficiently. (I have spreadsheets that examine why this is, but the raiding consistency math is beside the point.) One simply does not get enough time to properly learn the mechanics to the high standards of consistency that Ultimate demands.

Knowing this, I cannot in good conscience commit to any static intending to complete FRU - and multiple friends have asked my interest in FRU statics! Doing so would be dishonest to the group and a disservice to my own health. I knew that even before you told us all to take care of our health when challenging FRU in Live Letter 83. And progressing an Ultimate through Party Finder in its launch patch requires even more hours per week, as I experienced with DSR: not just to compensate for time wasted in groups that are ill-prepared for the party's stated objective, but to keep up with the strong players who are putting in 30+ hours a week to complete the duty quickly. Fall behind them and one's parties start getting worse rather than better, and from there it is a vicious cycle.

So it is with great sadness that I find myself forced to sit out of FRU, at least until late in Patch 7.2 (when player interest returns after AAC Cruiserweight (Savage)) and possibly permanently. I have the interest in and dedication to difficult duties. I have the complete maximum Item Level Best-in-Slot set for SCH, designed for FRU progression rather than purely for maximum damage output, and the books to adjust the set if I want to. (And I can pick up weapons for other jobs if I want, since I have been repeating the Savage tier every week.) I even have the nostalgia for Eden Savage. But as much as I find myself bored by Patch 7.0's duties, my body simply cannot take the physical demands of the current duty design direction.

It is an unfortunate result of Dawntrail's overall design direction so far, both duty design and its interaction with job design, that the current duties are mentally very easy and physically very difficult. This is an exceedingly poor state of affairs for any game that is not designed as a form of exercise - and those games accept being limited to the niche of players who want that specific form and intensity of exercise. For an MMO that wants both broad and enduring appeal, this will not do. In the remainder of this letter, I will provide more detailed feedback aimed at correcting both sides of this issue.

Not just for myself, but for the many players who complained about the Strayborough dolls and the Honey B. Lovely duties - which garnered widespread complaint precisely because their mechanics are so heavily reactive that even young and able-bodied players find them frustrating. For the world progression raiders who are disappointed that they completed AAC Light-Heavyweight (Savage) in less than 10 hours, and the experienced raiders in general who find that tasks they have little trouble with are crowding out the interesting difficulty. For players who are forced to have slow reaction times purely by aging. (This game used to be considered "the retirement home for fighting game players" - a genre where the highest echelons of play do rely on fast reactions, and so top players outright age out of the genre!) For the players who unsubscribed a month or two into Dawntrail out of sheer boredom; combat design is not the only problem leading to such a precipitous drop in subscriptions that you saw fit to run a Free Login Campaign in the launch major patch of an expansion, but it is definitely on the list. (Yes, World of Warcraft released its own expansion, The War Within, but that was almost a month after Patch 7.05. Dawntrail got to show its graphics, story, and endgame first, so many more players would have stayed if it gave a good overall showing.) For the players who stopped saying "this is too hard" not because they adapted - they cannot adapt if they hit a physical or working memory limit - but because they were shouted down by other players or quit in frustration. (Notice that the interview in which you mentioned that was over two months after the start of Dawntrail Early Access. Even players who subscribe through time cards would have had the opportunity to unsubscribe by then, and would thus no longer be able to post here.) For everyone, really.

(I am not expecting you to change FRU's mechanical design at this stage. Surely it is already locked in by now. And Ultimate design is so exquisitely precise and interlinked that even if I had delivered my feedback upon Patch 7.05 launch, you probably could not have fully responded to it in time for FRU. (I have a great respect for the Ultimate duty designers, who take on the task of pushing us to our limits but not beyond them!) But I hope that you will be able to make use of it for Patch 7.3 onwards, if not also Patch 7.2 duties such as AAC Cruiserweight (Savage).)

In the interest of thoroughness, I have divided the remainder of this letter into three major sections. First, I will discuss design principles and overall style, with particular emphasis on those aspects where current duty design is falling short. Second, I will break down the Patch 7.0 repeatable instanced duties from my perspective, both the good and the bad, in order to provide many examples of how I experience mechanics. For those readers who find that tedious, I encourage skipping over the second section to the third: my complete redesign of Honey B. Lovely's duties, both Normal and Savage, to show the design style I want applied to a Dawntrail concept.

As I will be extensively discussing duty mechanics throughout the game's history, and sometimes their relationships to the lore behind the enemies, fellow players take heed: the remainder of this letter is full and open spoilers through Patch 7.05, including duties outside the MSQ.

As a reminder, I am not asking for the game to be made easier in general, nor for a return to Endwalker's duty design. In fact, if I had to describe the duty design style I want by reference to expansions rather than principles, the closest existing expansion would be Stormblood.

(Note to newer players: Do not judge Stormblood's duty design style by how its duties play with the Dawntrail jobs, or even most of the Endwalker jobs. Duty design style must shape the duties to create the demands on the player from playing their job in the duty, as that combined demand is what the player will experience. The Stormblood jobs were significantly more demanding across the board to reach acceptable competence for every type of duty, from MSQ dungeons to Ultimates. On top of that, there are several forms of expansion power creep involved: better food and potions becoming available each raid tier, better effective gear from syncing down, lenience in the stat squish applied with Patch 6.0, newer expansion jobs having outsized potencies below their original cap level, and easier raid buff synchronization due to the "2 minute meta" implemented with Patch 6.0.)

Stormblood's duty design style aims to challenge players at every level of play. It is not just the expansion that introduced Ultimates, it is also the expansion that put trios - otherwise an Ultimate mechanic type - in an Alliance Raid (Ultima, the High Seraph in The Orbonne Monastery). The expansion that gave us the original Thunder God Cid, which has since been nerfed three times. (And two of those are because it dared to ask Duty Finder for Extreme levels of party coordination.) The expansion with the infamous MSQ solo instance fighting Sadu. The expansion that made tanks check boss buffs for Local Resonance / Remote Resonance and move Omega-M and Omega-F accordingly... in Normal Mode. Of course, it has memorable and iconic Savage mechanics too - Playing Field, Grand Cross Omega, Forsaken, Hello World - but it does not limit the challenge to high-end duties.

Remember Omega's motto: "Fight, win, evolve." As a duty designer (and as a job designer), you should be inspiring players to apply it to themselves.

Challenge is necessary to avoid boredom and spur growth. Accordingly, you can and must challenge players. That means having penalties for failure, as it is those penalties that create the difference between success and failure. Hand out the avoidable hits, the Vulnerability Up stacks and Damage Downs, the KOs, and even some party KOs in regular duties.

But the immediate goal of duty design is not for the duty to "win" over the player by pushing the player into giving up. The goal is not even to push the player into trying to complete the same duty indefinitely, since eventually the player will lose patience and give up - and "eventually" is not that many attempts for most combinations of player and duty. The goal is for the player to complete at least the regular duties and improve in the process, so the failures need to be motivating rather than demotivating.

--------

Which raises the question: what is the difference between a motivating failure and a demotivating failure?

We have some intuitive understanding of this, but we also have existing scientific research that can give us more details. From my previous writing on balancing the magical ranged DPS role, in January of this year:

Quote:

---

I must make a brief detour to psychology. In particular, I must explain Brehm's motivational intensity theory, which has not only been tested in video games and held up, but overall has been replicated so extensively that Richter, Gendolla, and Wright's summary paper had to direct readers to narrower summary papers. For brevity and clarity despite machine translation, I will explain only the points of the theory that relate to the question of balance at hand, and I will explain via example.

Suppose a player is considering a goal with a fixed difficulty, such as completing a dungeon. If they had to do something ridiculous to accomplish that goal, like find three other people and teach them the game in order to have party members, the player would almost certainly reject that goal and find something else to do with their time. Likewise if they had to do something blatantly impossible. This illustrates that they have a maximum amount of effort that they find worth it or possible, which we call their "potential motivation" in the context of that goal.

However, even if the player can complete the dungeon within their potential motivation, that does not mean that they will put forth that maximum amount of effort - they may be willing to spend an hour in the instance, but if the party completes the dungeon in twenty minutes, they will not then spend the remaining forty minutes in the same instance pressing their buttons with no targets. (They may well spend it in the same instance exploring the environment, but that is a new goal.) Rather, they will put forth the minimum amount of effort needed to accomplish the goal, and no more than that.

This theory makes intuitive sense from the perspective of conserving energy. A rice farmer who insists upon moving to a desert and constructing new rice paddies there will probably produce no rice and starve. A rice farmer who insists on hauling their harvest back and forth for no reason, not even physical training, will not put themselves in a good position either. The many experiments confirming the theory show that it is very much correct - conserving energy in this way is an evolved imperative of the human brain, and games for humans must bend to it.

---

At the risk of stating the blatantly obvious: Motivation is all about the player's perspective. They are working with their own information, interests, and capabilities, not those of a duty designer or even a different player.

(In general within this letter, I would rather state something you already know than omit something and be misunderstood. It is all too easy to wrongly believe that some knowledge is already shared, and working through a language barrier does not improve the situation.)

--------

From the player's perspective, the effort they must put in to complete a duty is fixed but initially unknown. (The same is true of longer-term goals, such as completing the MSQ up to a fixed point or completing a Savage tier.) Before entering a duty, the player has an estimation of the effort they must put in to complete it, based on factors such as their prior experience with the game and their understanding of the available duty difficulties. If that initial estimation gives too high a chance that the effort they would need exceeds their potential motivation, they will preemptively decide "not worth it" and not even attempt the duty. (Which is often a good thing - this is what keeps players with no desire to raid from joining progression parties for high-end duties anyway. But it can also keep a player who perceives themselves to struggle with MSQ from pursuing optional regular duties such as Normal Raids and Alliance Raids, even if they would perform acceptably there.) If they enter the duty, they will continuously refine their estimation of the effort they would need based on the feedback from the duty itself, the effort they put in and the mechanical successes and failures they experience... and if at any point that refinement gives too high a chance of exceeding their potential motivation, they will declare "not worth it" and quit the duty in frustration.

Hence the first major design principle of this letter:

Every time you tell a player that they have failed something, they must correctly perceive a way to do better next time that will lead them to success within an effort they find reasonable.

(There are two exceptions. One is when dealing with the small minority of players who thoroughly refuse to learn, as they will do that no matter how carefully you teach. The other is in high-end duties that players can enter while woefully unprepared in prerequisite foundational skills, where it becomes necessary to tell unprepared players "this is not currently a productive use of your time" to discourage them from making themselves and their party members miserable. TOP is admirably designed in this regard: its first two phases carefully test many mechanical skills required for the later phases, so players who are not ready to tackle the later phases will generally fail out in frustration early on, rather than after their whole party has each invested several dozen hours into the duty to reach the special surprises.)

Allow me to go over this wording in more detail to present sub-principles.

"Perceive": The existence of a route to success does no good if the player does not understand at least the first step.

For example, it is well within casual players' capabilities to avoid KO from Judgement Nisi in Alexander - The Burden of the Father (A4N); they need only arrange their characters such that the two healers avoid contact with each other and all other player characters for the 30 second duration of the Nisi debuffs, and have the healers heal themselves periodically. (The Nisi debuffs start with one type on each healer, spread by contact, and having more than one type is an instant KO.) But casual players cannot be relied upon to read their debuff bar, even in the healer role but especially outside it - and that is the only indication of the Nisi debuffs in A4N. Thus, players whose characters are KOed by Judgement Nisi often do not perceive even the immediate cause of the KO, only that their character "mysteriously exploded". And they are correct about their own state of confusion!

Accordingly, they become frustrated at a penalty that seems out of their control, even though technically the duty has given them both a way to succeed and the information to discover it. If preceding duties had taught casual players to reliably notice when their characters are debuffed, then they would perceive the information given to them and take it as a clue. Likewise, if the Nisi debuffs had an additional indicator that casual players would reliably notice, such as the overhead markers in The Epic of Alexander (Ultimate) (TEA), once again they would perceive the information given to them and take it as a clue.

In the Patch 7.0 duties, violations of this sub-principle (and the next) regularly manifest in the form of mechanics that discourage the player from looking in the correct place to understand the mechanic, whether by overwhelming visuals, forcing the player to fight their interface, or sometimes even actively unhelpful non-diegetic cues. When that happens, the player enters a mindset of "try to constantly look everywhere", which is not a fun kind of stress. (Notably, it is not physically possible at high fidelity, due to the layout of the human eye where details require using central vision. Plus, the player has no way of knowing if they can relax out of that state without learning the duties.) In high-end duties, it can sometimes be appropriate to ask the player to figure out what information is mechanically relevant, but that requires too many attempts to be viable in regular duties.

"A way to do better next time": The player must learn something from every failure, but not necessarily a complete solution to the mechanic. The player's evolution is the cumulative effect of many small steps they take to improve.

If the player correctly thinks "okay, I got hit by this attack, I should watch out for where it comes from", that is acceptable. If the player correctly thinks "I misjudged the timing of that attack, I needed to move earlier rather than staying to deal damage", that is acceptable at low frequency; it is necessary feedback for a player developing their timing, but they will require long practice over many duties to significantly improve those skills rather than always moving early and cautiously. (Which is why so many players come to love snapshots at the end of castbars, despite the visual mismatch with attack animations - the castbar provides a fair and consistent cue to use in timing judgements.) If the player correctly thinks "I need to learn the duty timeline to plan for that before it appears", that is only acceptable in high-end duties; not only is the practice and memorization too much to expect for regular duties, but requiring advance knowledge of the duty timeline often diminishes the story experience of regular duties. If the player correctly thinks "I need a reference sheet for this mechanic", that is only acceptable if it is one small reference sheet for the entire duty, that they can write or draw out themselves rather than relying on a second device or monitor. (Which is why The Ridorana Lighthouse is acceptable - a dyscalculic player can write themselves one small reference sheet based on the notes in the instance before the "math boss" or given to them by other players.)

If the player correctly thinks "I moved as soon as I knew where to go, but I was not fast enough", that is only acceptable insofar as they perceive an actionable step to know where to go faster on their very next attempt and are willing to take it. (Which not only limits the speed one can demand of players, it largely confines such demands to high-end duties.) "Oh, this cast is always followed up with another attack in a corresponding pattern, I do not have to wait to see the follow-up" is actionable so long as you can get the player to notice the connection between attacks. "I should be looking at my debuff bar when this cast finishes" is actionable, but only reliably perceived in high-end duties, since casual players as a whole have not currently learned to reliably look at their debuff bar at all. "I should drill my planned movement sequence for this mechanic" is only actionable for players who already have such a sequence and are willing to dedicate external study time to a single mechanic, qualities you cannot rely on at all below Savage - and even in Savage and Ultimate, there are strict time bounds to that before the player gives up instead.

"I just have to react faster" is not actionable under any circumstance, and therefore not acceptable. Even the vast majority of Ultimate raiders will not perceive "make time in one's schedule for generic rapid processing drills, and keep that up over weeks to years for at most 10%-20% improvement, before returning to progressing the duty" as an option - and if you directly tell them to do it, they will balk!

Proposed changes to the player's Earth circumstances are almost always not actionable either. The speed of light and the Internet infrastructures of different regions introduce unavoidable latency, but that does not make "move to live beside the servers" actionable unto the player. (Indeed, players may have moved away from their original region due to a variety of Earth circumstances, but not be able to move their character without losing various server-bound features (like housing!) and/or the ability to play with their friends on that region's Data Centers.) Even "get a VPN" may not be actionable for a player on a tight budget, let alone "change Internet Service Providers" or "find housing that allows for a hardwired Internet connection". Likewise, "get a better computer / console beyond the declared minimum specifications", "play on PC to use accessibility third-party tools", and "get another monitor to flip through many diagrams" are not actionable on budget grounds.

"That will lead them to success": The player must succeed eventually, rather than getting stuck in a dead end or an infinite regress of failing by ever-slimmer margins, or you will exhaust their potential motivation and they will give up. Simple enough, and another reason players find the prospect of "just react faster" demotivating - they often know intuitively that training reaction time reaches a dead end eventually, and are loath to invest heavily into it without knowing if they can improve sufficiently.

(Puzzle mechanics are an interesting case here. From the perspective of a player aiming to solve the puzzle, some reasonable directions for potential solutions may be dead ends, and that may even be necessary for the difficulty of the puzzle. But a well-designed puzzle mechanic will give the player clear feedback on all wrong potential solutions - even if only with some random variations of the mechanic - that tells them their plan is wrong rather than imperfectly executed. That feedback is what prevents the player from getting stuck in a dead end even if they enter it, since it shows them to rule out the potential solutions in that direction eventually. As for that "eventually", see the next sub-principle.)

"Within an effort they find reasonable": The player's cumulative effort put into the duty, both every mechanical failure you hand them and every mechanical success they work for, must not exceed the effort they are willing to put into it. The player does not have Omega's relentless drive to evolve, at least when it comes to your game rather than their life as a whole. They can walk away at any time. You have to work with the time and effort they are willing to give you.

Obviously the player's definition of "reasonable effort" will vary massively depending on the context - just about anyone will be quite frustrated if they cannot complete a regular duty before its instance timer expires without advance knowledge of the mechanics, while raiders tackling an Ultimate would be understandably angry if they could complete one that quickly.

"Correctly perceive": If the player believes incorrectly that a way forward exists, they will try what they believe to be a way forward, continue failing, then give up. If the player believes correctly that a way forward exists but is incorrect about what it is, they will try the route they perceive rather than the route that works, adding to their cumulative effort spent. Accordingly, the further your intended solution is from the player's existing mechanical knowledge, the more extensively you must guide them to the answer. The easier you want the duty to be, the more extensively you must guide the player to mechanic solutions. This is why regular duties cannot have puzzle mechanics - Duty Finder's matchmaking cannot even guarantee the presence of a single player who can solve the puzzle in the party, let alone the patience from the whole party to wait - and all the most elaborate puzzle mechanics are in Ultimates.

But the implications are not limited to puzzle mechanics. The Duty Finder population can actually learn quite a wide range of mechanics, but teaching through regular duties must be done in small incremental steps. There, the player is trying to finish their quest or roulette, and does not have the interest in studying anything outside of attempts at the duty to learn a complex concept all at once. (A key point in which Shadowbringers and Endwalker did better than Stormblood is the guidance and teaching: using markers, NPC commentary, and sometimes even NPC demonstrations to teach players in those small pieces.)

"You tell a player they have failed something": If a player in some sense fails but perceives no penalty, it does not matter whether they perceive any steps to improve; they have no reason to take them regardless.

This is often a matter for job design, as duty mechanics are themselves pass/fail, but it is relevant to duty design where jobs create multiple degrees of mechanic success. If a casual player only undertakes regular duties and never actually sees an enrage cast, it is fine for them not to understand any steps they can take to increase their damage output without upgrading their equipment - including mechanic resolution techniques such as slidecasting for jobs with frequent cast times or gap-closing back to a boss immediately after the snapshot of a mechanic that requires leaving melee range. Not accomplishing such things will not frustrate them, even though it is "failing" relative to optimal performance. Should they choose to pursue greater skill at their job, whether to undertake harder duties or simply for intrinsic enjoyment, then they may come to perceive the downtime / lower potency output as a penalty, and at that point the rest of the principle once more applies.

This is also relevant to regular duty design when it comes to giving feedback to the player(s) who can act upon it.

For example, Alexander - The Fist of the Son (A5N) contains a "tank swap" mechanic: a series of attacks targeted at the player character with highest enmity that are especially dangerous if the same player character takes all of them, encouraging or requiring the two tanks in a full party to swap who has highest enmity in order to split up the series of attacks. In A5N, the relevant series of attacks is very slow, with layers of feedback. The boss, Ratfinx Twinkledinks, grows dramatically before using four individual Gobjab attacks, each of which inflicts a stack of Headache. The Headache debuff description reads: "Suffering mild head trauma. Damage dealt is reduced. Increased trauma results in a Concussion." If this does not clue the player in, the Concussion debuff makes the player character unable to act with a special animation, which becomes quite obvious to the player. Only after that does the boss use the Gobstraight tankbuster, which deals massive true damage if taken by a player character with Concussion, thus producing a KO. Tank swapping at any point between the first Gobjab and the start of Gobstraight will avert the KO, and tank swapping at any point between the first and fourth Gobjabs will avert the Concussion as well.

The biggest problem is that all of this feedback is given to the player whose character has the current highest enmity, the "main tank". The player who needs to act in order to take highest enmity, and therefore needs to receive feedback, is the other tank, or "off tank". (Nowadays, the main tank can Shirk the off tank, yes, but that will not actually swap the off tank to highest enmity unless the off tank already had at least half the main tank's enmity. And Shirk was added in Patch 4.0 Stormblood, so tanks did not have that option when A5N was current. The off tank absolutely had to be the one to act, in an environment of job designs where the easy ways to have high enmity involved sacrificing damage output.) And the off tank perceives none of this unless they not only look at their co-tank's debuff bar, but read what the debuffs do - which the vast majority of casual players will not do! They will perceive only that their co-tank "mysteriously exploded", and will likely attribute it to a failure of mitigation or healing.

Given this duty design, what must happen for the casual player off tank to learn to do better next time, without the unrealistic expectation of reading their co-tank's debuffs? First, the main tank (or another party member watching the whole party's debuffs) must work out that it is possible for the off tank to act to split the attacks. Second, whoever figures this out must tell the off tank that they failed the mechanic, which requires both social willingness and some typing speed. Third, the off tank must accept this rather than blaming another party member. Fourth, some party member must break down the relatively complex concept of tank swapping so that the off tank understands when to take the highest-enmity spot, how to do so, and that the former main tank will have to take the highest-enmity spot back in a future tank swap at some point.

Casual players see all of these steps as significant hurdles, and so A5N could not teach the Duty Finder population to tank swap. (Players pursuing high-end duties do manage it, because the expectations surrounding high-end duties create a player population that is more dedicated, expecting to put in more effort to learn, and willing to kick players who cannot or will not take feedback from others.) More broadly, casual players cannot be relied upon to give others feedback or receive it, so you as the duty designer have to give the feedback to players who can act upon it.

(Indeed, Heavensward is further from the design style I want than Stormblood not because of its attitude to challenge, but because it did worse at teaching players. Alexander - The Burden of the Son (Savage) (A8S) is the duty that convinced me to pursue Ultimate, because I thoroughly enjoyed the difficulty and the intense focus required.)

Want to teach casual players to tank swap? You can, but not like that.

First, create a marker for attack sequences that should be tank swapped, so that players have a cue to apply the concept in future duties before seeing the enemy finish the dangerous series of attacks. (That could be on the castbar, like the "interruptible cast" marker introduced with Patch 5.0 Shadowbringers, or it could be on the boss's current target like the various tankbuster type markers introduced over the course of Shadowbringers and Endwalker. Either offers the opportunity to give extra feedback on a successful swap, over the "target of target" HUD element or by moving the marker to the boss's new target.) Then create a solo instanced duty with an NPC co-tank who can instruct and demonstrate. For example, if you wanted all players entering Dawntrail's endgame to learn it, whether they currently play tanks or not, the solo instance could have been in the MSQ as an Echo vision of Wuk Lamat's combat training with Gulool Ja Ja. (Even players who dislike Wuk Lamat as a character can still come away with new mechanical knowledge, and it would fit the themes of the existing MSQ.) In that duty, have the player character be the main tank first, while the boss begins a very slow and flashy tank swap mechanic that will show the player the new tank swap marker for the first time. (A5N did the "slow and flashy" part right.) During the series of attacks, have the NPC co-tank comment on the danger and demonstrate a correct tank swap to the player, taking the end of the series of attacks and moving into position to keep the boss still. Then, after a short "spacer" period of familiar mechanics, have the boss begin the same tank swap mechanic again targeting the NPC, showing the player the tank swap marker for the second time. Now that the player has seen a tank swap demonstrated, the NPC can comment that the player character needs to take the boss's attention back, and the duty can even give a non-diegetic prompt as some solo instances do. If the player correctly takes the highest-enmity spot from the NPC during the mechanic and keeps it through the end of the series of attacks (using Provoke and their tank stance), give them some appreciative feedback from the NPC and carry on. If not, KO the NPC and start the duty (or section of the duty) over.

(Concepts building on that foundation, such as when to use Shirk and double tank swaps, could be covered immediately afterwards but are best covered once players have solidified the basic form.)

Do that, and they will try to tank swap when they see the marker again on their co-tank in future duties. (And since they have seen both sides, they will not be confused when another tank swaps off them.) They will likely not be quick about it the first few times, but that would have been a good opportunity for the first two Dawntrail trials to allow players to practice with Duty Support without any fear of embarassment. For some concepts, you really have to guide the player's hand to get casual players to pick them up.

"Must": Why not "should"? Because "should" creates temptation to put the responsibility back on the players instead of the game design, without first seriously considering how the game can be designed to improve the player experience.

If a casual player mistakenly thinks that Normal Raids are Serious Raiding and avoids them accordingly, that is not the community's responsibility to fix - that is your responsibility to tell them otherwise. (Players who complete the MSQ through Duty Support may barely talk to other players at all!) If a colorblind player struggles to tell the difference between the two types of fists in Alphascape V3.0 (Savage) (O11S) or TOP, that is not their fault - that is because you did not make the shapes distinct enough without relying on color cues. (Everything else in TOP is thoroughly distinct for colorblind players... in contrast to the tethers in Asphodelos: The Fourth Circle (Savage) (P4S) Act 2, which have no distinguishing features at all other than the colors shown over player characters' heads. Likewise, the orbs in Emerald Weapon (Extreme) are distinguished only by their colors.) If a player cannot complete the Active Time Maneuver (also known as a Quick Time Event or QTE) in Seat of Sacrifice Normal, or injures themselves trying, that is not their failure to learn - that is your design error in considering what pace of inputs the game should really require or make players think it requires.

And yes, I am aware that that particular QTE has a "fakeout" where the first 75% of the bar depletes rapidly to make the player think they are always on the brink of failure. I have seen the metronome demonstrations that it can be completed at 200 actions per minute (APM), even though it looks like the player is about to fail. I am aware that QTEs in this game do not require clicking specifically; any button that does not open a menu, send input to the chat window, or cause the operating system to take focus off the game will do. It is still an unreasonable demand to make of players, because your actual gameplay is not nearly that fast even on fast jobs, so it needlessly excludes players who are capable of the interesting challenges you want to set them.

Let us take a very fast job and work through the math. Suppose a player is able to do a perfect MCH Hypercharge window, hitting Heat Blast / Blazing Shot every 1.5 seconds and hitting a Gauss Round / Ricochet / Double Check / Checkmate evenly in between every Heat Blast. That is two inputs every 1.5 seconds, so one input every 0.75 seconds. Suppose the player is able to keep up this pace of inputs continuously, not just during the brief period of Hypercharge. One input every 0.75 seconds is 80 inputs every 60 seconds, so 80 APM. Even if the player uses only one hand for actions with animation locks (primarily job and role actions), leaving their other hand completely free for movement and targeting, and they are capable of the same APM with their other hand while keeping both hands perfectly staggered... that is still only 160 APM. A player can be comfortably able to play one of the fastest jobs in the game and still be physically unable to succeed at that QTE. (Or be at high risk of injury from attempting it repeatedly!)

And since the penalty for a single player failing that is an instant party KO, in an MSQ duty, that locks them out of the rest of the game.

That math pretty much matches my APM, by the way. I deal with QTEs in this game by giving up on controlling what keyboard keys I hit for the duration, and instead devising an elaborate method of hitting multiple keys with each finger motion by "rolling" each finger along one axis with staggered timings while "rolling" each hand along a different axis. (Note that this method will not occur to most players, and is very demanding on the player's motor coordination, which means that a player with limited APM is quite likely to be unable to do it even if they think of it.) And even then I nearly did not make it through the sustained input requirement of the Seat of Sacrifice Normal QTE. It is absurd that anyone who can clear an Ultimate was nearly hard blocked from completing the MSQ. This is a clear design error.

In contrast, the QTE in Alexander - The Heart of the Creator (A11N) is entirely acceptable, as it requires one input in several seconds to avoid an individual KO, and that capability is likewise required for even beginner gameplay. Likewise, the required use of the Omega Jammer in Alphascape V3.0 (O11N) is entirely acceptable even though failure generally results in a party KO, since the Omega Jammer has a 1.5 second cooldown and the party has significant leeway on the number of collective uses they need to form the shield. That does not require any more input speed than playing a very slow job casually, nor does it require better input cadence, and the players who struggle most can still be helped through the duty by their party members contributing more inputs. They match up to tasks that are both much less demanding and inherently necessary for core gameplay, so you already have a good reason to require them.

Since QTEs are not used in this game for anything mechanically interesting, you could have an accessibility setting to remove them or turn them into single button presses like the A11N one. Or you could replace them for everyone with forgiving cadence checks like the Omega Jammer, which would still function well to represent the player character fighting independent of their job actions. You have the technology. But you need to think about that sort of thing every single time, or design errors will slip through.

Every expansion has various good mechanics, even Dawntrail. But having mostly good mechanics is not enough. It only takes one bad mechanic to ruin a duty. It only takes one bad duty to ruin a roulette. It only takes one mechanic a player simply cannot do to prevent them from continuing.

You must meet the bare minimum standard of the duty being playable without pain or undue injury risk every single time. It is simply the rigorous nature of the task.

--------

The first major principle comes of considering how the player's required effort estimation changes. What about how the player's potential motivation changes?

Well, what moves the player to enter a duty in the first place? Not the effort that they expect to put into the duty, but everything they expect to get out of it: story progression, power fantasy of playing their job, access to and power in later duties, access to other game systems and areas, cosmetics, light entertainment, personal accomplishment, socializing, completionism, ... Gear is either cosmetic, for access to and power in later duties, and/or a form of accomplishment. Gil is similar, though it can also be used for non-duty gameplay. The meaning of an in-game item comes from the journeys we take for them and/or with them, without which they are just pixels.

And from the player's perspective, the total rewards of completing a duty are fixed but initially unknown, just like the required effort to complete the duty. (Even if they know how many times they must do Expert Roulette to get all the tomestone gear they want, they do not initially know how much and what forms of enjoyment they will get out of the duties or the gear.) So they likewise continually refine their estimation of what they will get from completing a duty, changing their potential motivation. If their potential motivation drops below their estimation of the required effort to complete the duty, they will also declare "not worth it" and quit the duty - depending on the cause, this can be out of frustration, boredom, or apathy.

Hence the second major design principle of this letter:

When trying to make the game challenging, do not lose sight of making it fun throughout. You must challenge the player to avoid boredom, but you can only challenge the player in proportion to what they get out of undertaking that challenge.

Let us examine some corollaries.

The player is constantly refining their estimation of how much they will enjoy the game later based on how much they are enjoying the game now.

You cannot expect the player to persist in improving at combat for very long at all without giving them regular positive feedback of actually succeeding at something related to that improvement, even if the success does not take the form of a duty completion. Thus, especially for casual players, "fight, win, evolve" is actually the correct order of operations. The improvement comes by learning from failure, but it needs success too. Even Ultimate raiders get frustrated by spending hours on progression without doing the mechanic they are working on correctly.

Likewise, the player's estimation will naturally include factors outside of duty design, and outside of combat entirely. If the player is interested in the story, their finding Wuk Lamat poorly written will mean that you have less opportunity to challenge them in the MSQ, where they reasonably expect to see more Wuk Lamat for their trouble. If the player wants something specific out of their job, not having and funneling them into a job that matches that desire will make them get less out of duties, and so have less tolerance for putting effort into duties. If the player wants to play their Warrior of Light as a morally upstanding hero - which most players do - and considers the Endless to be people, forcing them to genocide the Endless has likely permanently put them off of all future storylines and the associated duties, even if they continue the game at all. Dawntrail has done rather poorly at giving players who are not already "good at combat" a reason to persevere.

(Personally, I started playing the game for the old crafting system, and pushed through my initial struggles with the MSQ duties to reach to new zones to craft and gather in. Indeed, I first started attempting high-end duties in order to prepare myself for future MSQ requirements, and then discovered that I prefer the planning-oriented design style of raiding and its encouragement of job mastery to any design style so far for regular duties. And that got me far enough to start seeing improvements in how I work with my physical limitations outside the game, which is an extremely strong motivation to raid. When it comes to player persistence, I am very much an outlier... and yet someone just like me who started playing today would not make it to the endgame at all. The new crafting and gathering actions are good for restoring rotation depth, but they are not nearly enough to sustain someone like me through the discouraging aspects of the Patch 7.0 MSQ duties.)

Of course, the player's expectations of enjoyment of future duties will be informed by the duty they are currently in, whether they complete the duty or not. (And the player almost always has some interest in completing future duties of similar difficulty if they have a healthy interest in the game - pretty much the only non-concerning reason for a player to lack that interest is pursuing a single Ultimate for glamour purposes.) Push the player too hard, and they will not enjoy the duty even if they complete it. And if they keep not enjoying a category of duties, they have good cause to give up on the category. This can arise from their party "carrying" them to completion, or in solo instances where the panic of repeatedly scraping by and worry about future duties push out enjoyment. Most players want to feel useful to their party, even in casual story-oriented play.

(When The Copied Factory first released with Patch 5.10, I struggled with certain mechanics and found them overly reactive. Having that experience once is one thing. Completing the duty several times in the first week, in order to get a specific gear piece for raiding purposes, and still having reaction problems with mechanics is another thing. When subsequent weeks still did not improve matters, I made a detailed plan for each such mechanic just like I did for final Savage floors, complete with prepositioning not only my character, but my hands, eyes, and camera... which resulted only in my character being KOed in the safespot instead of out of it. I then asked several raider friends to help refine my plans, including one who had already completed TEA, and none of them could suggest improvements.)

(At that point I had gone above and beyond to rise to the challenge, only to find that my initial impression of a reaction time problem was correct. So not only did I stop entering The Copied Factory, when The Puppets' Bunker released (with Patch 5.30) I refused to unlock it - to avoid further worsening my chances of getting a personally enjoyable duty in Alliance Roulette - and pointed anyone who protested on gearing grounds to my Eden's Verse Savage gear. To anyone who protested because they thought I would enjoy it, I replied "ask again when I clear an Ultimate." Not as a deliberately absurd statement, but as a logical conclusion: having already exhausted all the rapid ways to improve my practical response time by better planning and HUD Layout, the only remaining option was to undertake hundreds to thousands of hours of practice efficiently processing mechanics and hope for improvement... and already being comfortable in Savage, the only remaining higher difficulty that could show that improvement was Ultimate.)

(I knew full well that the other 23 players could carry me through an Alliance Raid even if my character spent the entire time KOed. I was even already playing a DPS. I was miserable regardless, and naturally unwilling to attempt the extended reaction time drills in the same duties that were making me miserable. Needless to say, the vast majority of players who see frustrating lack of personal improvement in a regular duty are not going to even attempt to fix that in Ultimate - they are just going to quit the game.)

It is also worth noting the game's repetition-based gameplay loop. If the player knowing all the details of a duty makes it boring - or frustrating - they will not want to repeat that duty. In particular, based on the Patch 7.0 duties:

• Delaying the first indication for what random variation of a mechanic the enemy is doing does not make the duty interesting or fun. It only teaches players to learn when that first cue appears and wait until then in boredom, despite the brief moment of stress from resolving the mechanic once the delayed cue is finally given. The result is a boring experience punctuated by brief moments of stress, and at no point is it fun. Even stressful and difficult duties can be boring; this is why I have been more likely to fall asleep in Dawntrail duties than those of any previous expansion!
• Movement sequences that push reaction within the sequence are not fun, nor are they interesting. The player is denied the opportunity to plan, and instead learns that there is no recourse but reacting repeatedly - with the result of being bored even during the reaction stress. (If you are expecting the player to get some sort of enjoyable adrenaline rush out of reacting, well, I do not get that, and the mixed player feedback makes it clear that I am not alone in that. It is annoying at best to people like me, and your player population has been shaped by the older turn-based Final Fantasy games, along with this game itself historically being quite lenient on reaction time for a real-time game. Anyone who does get a reaction thrill has lots of other options.) Instead, give the player the information needed to plan the movement sequence, and then ask them to execute it at a speed appropriate to the duty and intended effect of the mechanic.
• Making needed visual indicators difficult to see is anti-fun. (Mind that there are significant differences in style of visual processing even between able-bodied and fast-reacting players using the same console / computer setup! Plus a screenshot that gives a clear view of a mechanic is not an accurate representation of what the player sees with limited time, since the human eye does not have high detail outside central vision and various action effects briefly blur the screen.) For players with any sort of visual impairment, photosensitivity, reliance on rapid processing, or simply visual processing style you did not expect, it is an active detriment. And for everyone else, it does not matter either way. Have visuals that suit the lore of the duty, yes, but do it in a way that prioritizes visual clarity.
  (The word "needed" is important. The A8S Mega Beam obscured by explosions is quite fun, because it is actually a planning test - its earlier appearances teach the party that it is baited on a random player character, so they need only arrange their solution to the preceding mechanic to create a safe zone no matter who the boss chooses for the bait. Plus every duty has visual elements that are not mechanically relevant, included for other reasons, and those may be as subtle as the artist and/or lore writer creating them desires.)
  In fact, ensuring visual clarity makes for more fun mechanics that are easier to design. Why? Obscured visuals delay some players much more than others, while clear visuals eliminate that delay, therefore yielding a relatively uniform discernment time among the player population attempting the duty. (Many forms of obscured visuals also vary how much they delay the player based on random variations of the mechanic.) You can then plan around that discernment time when leaving time for other tasks in the mechanic, such as making a decision or moving their character. This gives you finer control of how long the player has for those tasks, allowing you to tune the challenge level more precisely while ensuring that it remains achievable.
• Adding randomness to a duty does not automatically make it more fun, or even less boring. To be helpful, the randomness must contribute to something interesting, while ensuring that all possible duty timelines created by random variation have a similar skill floor (required competence to complete) for every job. (Branching timelines regarding the order of mechanics, such as those in Alphascape V1.0 (Savage) (O9S) and Abyssos: The Eighth Circle (Savage) (P8S), are generally fine. (There is a limit, in that the player who puts in the effort to plan out all possible timelines reasonably expects to get to use all those plans before they stop returning to the duty.) Major variation in what mechanics are presented at all is prone to unexpected differences in skill floor, and therefore is best avoided.) Randomness in the wrong places can even make the duty less fun, by denying the player opportunities to plan uses of powerful job actions.

To reconcile giving the player an encouraging first-time experience with avoiding boring repeat experiences, encourage and reward skill expression so that the player will add more challenge to the duty at their own pace. Skill expression should be welcomed, not feared. (Which does not have to make fellow players feel bad, since the skill expression need not be clear to any player other than the one doing it.) This is tightly linked to job design, as it relies on the jobs to define what skill expression for each job is. Again from my previous writing on balancing the magical ranged DPS role:

Quote:

---

The fifth problem [with relying only on the duties to provide difficulty] concerns the longevity of any given duty, even one of appropriate difficulty for the player. As I am sure you have experienced in raiding, one finds that any given duty gradually becomes much easier as one learns that duty's mechanics. Even if the duty incorporates randomness, one must still learn the variations to reliably complete the duty. In effect, familiarity with a duty lowers the skill floor of completing that duty with any given combination of job and equipment. (Though less so for jobs in different roles that have different parts to play in mechanics.) As a result, any duty that one can complete occupies less than one's full attention on subsequent completions, at any difficulty level from story to Ultimate, even without better equipment. What, then, can occupy the rest of one's attention in order to prevent boredom? Not the duty, by necessity - only the jobs can do so.

---

However, duty design can and should give the player both the opportunities and the tools to express their skill with their job. Having frequent cast times is what makes moving no more than necessary valuable, but it is duty design that can allow the player to know exactly where an AOE will hit so that they can remove unnecessary movement. (The Patch 7.0 duties have a recurring lack of floor geometry for the player to use as reference points.) Being limited to melee range to effectively attack is what makes staying close to enemies valuable, but it is duty design that can allow a tank to reposition enemies towards a safespot for a later mechanic. Having limited usage of mitigation and healing tools is what makes planning their use valuable, but it is duty design that creates the timeline and its patterns to plan around. (In particular for dungeons, fighting two packs at once - known as "double pulling" - is a form of skill expression. If you are going to deny the player that option for a given trash pack, you should present some other form of skill expression in that trash pack instead.)

Satisfying mechanics actively reward the player's skill expression, adding enjoyment and longevity to the duty. What makes a mechanic or player strategy satisfying to execute is not just the opportunity for the player to show their skill and/or knowledge of the duty, but also clear confirmation of doing well. An earlier mechanic leaving room to handle a later mechanic, or even being used to counter the later mechanic, because the party set it up that way. A spinning AOE dodge where the final hit comes almost but not quite to the player character's position, not by accident but because the player knows how far it turns. A healing and mitigation plan that effectively uses a cooldown exactly on cooldown multiple times in a row, because the party mapped out the fight timeline to make that happen. (This is not even limited to high-end duties - Eden's Promise: Anamorphosis (E11N) allows a magical ranged DPS to Addle three Burnished Glory raidwides in a row by using Addle early on the first Burnished Glory and then exactly on cooldown twice thereafter. Drift the cooldown and the last Addle will be too late.) And yes, sometimes a sequence that the party gets to watch resolve after they have set it up - a flashy sequence is a good way to let the party see the results of their work and create a breather before the next mechanic. The existing Dawntrail duties are sorely lacking in such mechanics, which does their longevity no favors.

--------

Since players have so frequently remarked on the demands on working memory and reaction time in Dawntrail so far, these two aspects of duty design deserve special attention.

Working memory mechanics are like soy sauce on sushi: a small amount enhances the flavor, too much overwhelms it. (Which is why the soy sauce is supposed to go on the fish, not the rice!) Totally aside from the appropriate difficulty of memory mechanics, Patch 7.0 has too many of them, crowding out other mechanics that could provide other forms of difficulty. The result is boring and repetitive duties.

As for the appropriate difficulty of memory mechanics, the player does have a working memory limit inherent to their brain, which they cannot change... and which they need to use for playing their job in the duty, not just doing the mechanics. (Yes, even for downtime mechanics - jobs still have healing, buff-based mitigation, and resource-building tools to think about.) The player finds a job and/or a duty less memory-intensive with practice not from expanded working memory, but from learning bigger "chunks" that they can work with fluidly. For example, an intermediate RDM can think "basic melee combo" for what a beginner has to break up as "Enchanted Riposte, Enchanted Zwerchhau, Enchanted Redoublement, Verflare / Verholy, Scorch, Resolution".

Plus the player cannot be expected to offload the task of remembering to their chat log - even if they have the APM to type in combat, they may be using a controller without a keyboard in reach or extra hotbar space for text macros. They can only be relied upon to have one tool for storing a duty-provided piece of information outside their brain, and that only if surrounding mechanics permit: moving their character to a location related to the mechanic, such as a direct or conceptual "first safespot". (Prior to Patch 5.2, the party as a whole could also agree on moving waymarks to remind themselves of the random variation shown, but since players can no longer move waymarks in combat, that is no longer an option for memory mechanics.)

What is the working memory limit of the human brain, then? Well, the classic "seven plus or minus two [items held in memory]" is at best an upper bound, and newer experiments best fit the idea that there is no fixed number of items at all, but instead pushing to remember more items results in less complexity and detail for each, until eventually there is too little detail to usefully be a memory. If all the items to be remembered are very simple and familiar, such as individual digits of a number, then "seven plus or minus two" is roughly accurate. (Indeed, telephone numbers without area codes provide a ready demonstration of that specific working memory task, called "digit span".)

While memory mechanics can often be broken down until the player only needs to remember appropriately simple and familiar concepts like "spread" or "left", the rotational chunks of the jobs are not that simple - and making them that simple would be terribly dull. So we must take the low end of "seven plus or minus two" - that being five items - and set aside not just one item, but at least two and possibly more, for the player to continue playing their job. (As well as potentially more room to plan ahead for other mechanics!) Therefore, the total of "stored items not yet released" from memory mechanics may never exceed three simple items at any point in a duty's timeline, and should almost always be less. If the player has a good working memory, they can and should occupy themselves with their job, looking for hidden patterns, planning for later mechanics, ...

Examples:

• Some recurring sets of mechanics where the boss randomly choosing one counts as "one item" for a memory mechanic: {spread / pair / light party stack / full party stack}, {left / right / front / back}, {in / out}, {Deep Freeze / Pyretic}.
• In Eden's Promise: Eternity (E12N), the stored primal combination ground AOEs count as two items, since first-time players will generally remember the primals individually and work out the overlapping safespots. In the Savage version (E12S), the same primal combinations start as two items and end as one item, since the player will see them enough over the course of progression to memorize the overlapping safespots and work with those rather than the individual primals. (And for a final Savage floor, if the player is having problems with how fast they remember the safespots, they can actively study the five possibilities outside the duty, which does not take that long.) FRU would be able to treat them as one item from the beginning, due to the high expectations for external study and number of attempts at the duty in Ultimate.
• In Seat of Sacrifice Extreme, Quintuplecast looks like it violates the three item rule. However, the party can arrange themselves to do the whole mechanic in their clock spots for the Absolute Stone III proteans and close to each other (and the boss) for the Absolute Holy III stack, which allows the player to not bother remembering where those attacks are in the sequence. The remaining three attacks perfectly fit the three item rule, even if the player is not comfortable reconstructing the last attack from the previous two.
• Mascot Murder, the Living Memory World FATE against Mica the Magical Mu, is in flagrant violation of the three item rule. First the boss prompts the player to remember cards (all numbered from one through six) and displays six cards covering the arena, putting the player up to six items to remember as the boss hides these cards before the next section of the mechanic. Then the boss draws a card out of its hat, which the player must match to the cards covering the arena, though they stay at six items to remember since they can move their character to the matching card. (Though later repeats of the mechanic drop puddles on player characters after the hat cards are shown, which can drive them off the card in the arena if they are not careful about how they dodge.) Then the boss draws a second card out of its hat, which the player must again match to the cards covering the arena, and this time they have to remember the location as a seventh item. Then the boss draws a third and final card out of its hat, allowing the player to match the drawn card to those covering the arena once more before dropping the memory of the latter. Depending on how fast the player does the matching, they may briefly jump to eight stored items before dropping down to two. (Finally, the boss asks the player to answer which card it drew first, second, and third, with later repeats doing so in a random order and with other AOEs in between.) But even seven is wildly unacceptable, seeing as they are playing their job the whole time and not just doing a digit span test!
  Just because World FATEs are casual duties and players who struggle with working memory can be carried, that does not make it appropriate to make them miserable! (In my experience there is a hidden pattern (that the safe cards on the arena floor do not share any edges), but that does not matter - not only does the player need to have a good experience before finding any hidden patterns, but noticing them is harder under high attention demands.)

--------

As for reaction time, I wrote about this back in January, but I must expand upon what I wrote:

Quote:

---

The second problem [with relying only on the duties to provide difficulty] you have also already experienced with Anabaseios Savage, but more subtly: there is only so much you can demand in duty mechanics. A player can only react so quickly, move so precisely, focus for so long, and so forth. They cannot improve these attributes much by "natural" practice once they have learned the basics of gameplay. And even a player who is dedicated enough to undertake deliberate and thoughtful practice on these attributes will only have so far to go before they run into the hard limits of their body, including their brain. Running into such limits as a player feels awful, especially if it is due to disability or being outside the human body's brief physical prime - a concern for both the aging MMO playerbase and the family members they may want to share the game with! (As a disabled raider, I could talk about this all day, but that would be outside the scope of this series.)

Both regular duties and Ultimates have been grappling with these limits for years, and Extreme and Savage have gradually started facing them too. (For one example, consider Anabaseios: The Tenth Circle (Savage) and its Jade Passage orbs. The fact that they appear within the Pandaemoniac Ray animation and fire shortly thereafter makes them quite reactive, especially when they are later combined with Daemoniac Bonds adding more flashy animations and movement requirements. This sort of design even encourages players to use third-party tools to remove the flashy animations or otherwise get information more quickly, and Anabaseios Savage is full of it.) Players have already started remarking on the fast pace of duties.

These physical attributes limit not only the range of battle mechanics you can make, but how hard you can make them, and how hard you can make them if you want high-end duties to be possible outside a brief player age range. If you refuse to place difficulty in the jobs, these same limits will extend to how difficult you can make the top end of the curve at all, and not in a satisfying way. A player who is currently stuck on an aspect of skill they can improve may aspire higher, reach further, remain motivated to play. A player who is stuck on a physical limit and knows it - or sees the highest difficulty held back by other players' physical limits - is necessarily demotivated, for they know that no amount of effort they can put into the game will fix it.

---

When I wrote that, I was imagining a continuation of the game's slowly increasing pressure on reaction time over the course of Shadowbringers and Endwalker. I was not expecting Patch 7.0 to take a flying leap into more reactive mechanics! (Especially since your engine and the inherent latency of long client-server trips do not suit reactive mechanics well.) In this regard, Patch 7.0's duty design direction has been the exact opposite of helpful.

Obviously this is a real-time game, and therefore requires the player to meet some standard of reaction time. (If a prospective player sets up their HUD Layout and keybinds to their liking, and still cannot react to a basic ground AOE in less than ten seconds, they probably do have some limitation you cannot reasonably accommodate.) But the real-time nature of the game only requires the player to meet a fixed standard and stay there - it is not necessary to demand that the player react ever faster. Indeed, demanding that the player react ever faster is only useful in niche games that are designed as reaction time trainers, since even action games will eventually have players hit their physical limits. Since Final Fantasy XIV is carrying on the legacy of the originally turn-based Final Fantasy series and attracts many "non-gamers" to play with their family and friends, a generous reaction time standard is best for the game even aside from the aging MMO playerbase.

Why, then, is it intuitively correct that the updated version of Copperbell Mines paces its mechanics slower than The Royal Menagerie (Shinryu Normal) and The Ghimlyt Dark? After all, all those duties are required by the MSQ.

• The player cannot be assumed to start the MSQ with familiarity with video games in general, let alone the genre specifics of tab-target and holy trinity MMOs. They can be expected to pick those up over the course of gameplay. In Copperbell Mines, they may still be getting used to the sheer volume of information on their screen, along with foundational concepts like "class / job", "party", and "target". You can show them something slowly, but at that point you cannot rely on them seeing it at all! By The Royal Menagerie, not only are they used to the game's basic systems, they have built on their fluency with foundational concepts to be comfortable with recurring mechanics like "stack", "tankbuster", "raidwide", "add", ... Without any change at all in their reaction time, they understand more quickly what the duty is asking of them because they are better able to organize the information given to them and fit it into their existing concepts.
• The player starts the game with the default HUD Layout and keybinds, which are almost certainly not ideal for them - and may even be downright difficult and counterintuitive to them. (For example, nothing in the game tells the player to try both Standard and Legacy movement modes.) In Copperbell Mines, they may well not know what to consider in setting up the interface to suit them, and may even still be slow with their movement keybinds. By The Royal Menagerie, they have gotten used to their interface and looking over it, even if they stick to the default rather than adjusting it for themselves.
• One way to improve and maintain the human brain's generic rapid processing is real-time video games. Like this one. If the player does have any easy improvement in reaction time available to them, they get it over the course of the early MSQ and other regular duties!

Out of all these reasons, the only one that continues essentially indefinitely is expanding the player's library of game-related concepts, the mechanical layers of which are also known as "mechanical vocabulary". The average player in Stormblood duties will handily beat the average player in early A Realm Reborn duties in mechanical vocabulary, overall game familiarity, interface comfort, and reaction time. The average player in Dawntrail duties will handily beat the average player in Stormblood duties in mechanical vocabulary, but nothing relevant to their speed at dealing with mechanics.

Accordingly, Dawntrail duties can expect the player to make better guesses about what the duty is asking of them during their first time, use their existing concepts more effectively, pick up new mechanical concepts faster, ... than they did back in Stormblood. But the player in Dawntrail duties cannot practically respond to mechanics any faster than they did back in Stormblood. Indeed, one reason that Stormblood is the closest existing expansion to the duty design style I want is that it is the last expansion before mechanics that were still too fast even with practice started creeping in.

Some players have complained of mechanics being too reactive and feeling split-second, and other players have replied that the game does not actually give the player less than a second between presenting the first cue for a mechanic and snapshotting whether the player character has completed the necessary action(s) to resolve the mechanic successfully. While that latter statement is true, it is not a useful response to player frustration for two reasons:

• There are many steps other than the player's conscious perception of time to decide / react between the server sending the first cue for a mechanic and the server checking whether the player character is in the correct state to resolve the mechanic successfully. Those include the server sending the packets indicating the cue to the player's game client, the client system's latency in processing those packets and giving the cue to the player, the player's sensory nerve conduction time, for visual cues the player moving their eyes to focus on the cue (unless they are already looking at it or trained to manage it by peripheral vision, whether by planning or accident), the player's sensory perception processing time, the player's motor planning time converting their decision / reaction into nerve signals, the player's motor nerve conduction time and muscular response, the client system's latency in processing the player's input, however long the player character takes to perform the required action(s) (such as moving over a distance) client-side, and finally the player's game client sending the packets indicating the completed action(s) back to the server.
  Subtract all those other steps out, and the player may well be left with less than a second to decide / react even for a completely individual mechanic, and be accurately reporting that length of time! (For mechanics that involve acting relative to the party, including such basics as stacks and spreads, there are more steps unless the party has prearranged positions.)
• More importantly, [B]it does not matter to the player's motivation whether they have actually hit the hard physical wall of minimum reaction time imposed by their body. What matters is whether they have hit the soft wall of the practical reaction time they have now[B], such that their wanting to do better next attempt is not leading them to an actionable way to improve - and if they are complaining, that is already the case! (If they perceived a way to improve that they consider reasonable and actionable, they would take it instead of complaining.) Remember, motivation is all about the player's perspective.

When I find something too reactive now, I probably really am hitting my hard physical wall. But if you ask a player to jump through all the hoops I did to get there, they are almost certainly not going to oblige you. They are instead going to say "that is ridiculous to expect me to think of, let alone do" and quit.

--------

Let us examine what it means to avoid pressuring the player's reaction time in different types of duties through an example.

The basic spread is an A Realm Reborn mechanic, that has recurred over and over again in both regular and high-end duties. It recurs so often that it is the very first marker you teach in the new section of Hall of the Novice added with Patch 7.10. So why has the Duty Finder population never mastered it after 11 years? Why will experienced players still sometimes get their characters hit by two or more spread AOEs in regular duties, and even occasionally have a slow party KO by stacked spread markers KOing all player characters who can raise?

Because Duty Finder does spreads by reacting to each other, and asking the player to react to another player's choice will always inherently pressure their reaction time. No matter what fixed standard of reaction time you or I desire to set, one player meeting it in such a situation will give their party members less warning - and thus less time to react - due to client-server delays. The player improves up to a point by learning heuristic guidelines like "run away from other party members" and "let those with frequent cast times stand still", but that route will never save them from getting stuck in a loop of adjusting to another player. Even continuously tracking the positions of every party member will not lead to perfection.

(Client-server delays are not the only reason for adjustment loops. People do experience them face-to-face, classically when trying to pass each other in a hallway and both moving the same way. But the resolution is usually much smoother in a hallway - there is no "I was ahead of you on my screen!")

In contrast, the Party Finder and Raid Finder populations spread perfectly, even players who are new to high-end duties and doing their first Extreme. And this is not because the players in Party Finder / Raid Finder are better at the game - Duty Finder matchmaking can put a party full of world-class raiders together, and they still might fail a spread in that environment! Why is this?

Because Party Finder and Raid Finder do spreads by planning, not reaction. The party agrees upon a spot for each player character to go to, and they expect each other to go to their own assigned spots. They may react to seeing the mechanic appear again in progression, but they are not reacting to each other. (And they can make such plans for regular duties if they please, as world progression groups used to do back when Normal Raids were released alongside their Savage versions.) In fact, trying to react to each other instead of following the plan is often frowned upon, with phrases like "never adjust".

--------

More broadly, regular duties must be designed for the Duty Finder population and player expectations.

Imagine the "worst reasonable" party that Duty Finder could matchmake together: all first-timers with no foreknowledge of the duty, all at minimum Item Level, only casual knowledge of their jobs, no knowledge of how to plan for specific mechanics even if they want to, never met before, some returning to the game, others tired and under-hydrated after a long day of work and/or caregiving. (Yes, the latter will make their reaction time worse.) They are all trying to finish their quest and continue the story attached to the duty, and they are a little impatient. They are trying their best in the moment, but not looking beyond the moment, and they certainly have no grand plan of personal evolution.

That party should generally complete the duty before the instance timer expires, and if they do not complete the duty they must correctly perceive that they have come away with enough knowledge to individually queue back into matchmaking and get it the next time. Either way, they should enjoy their time spent in the duty. Challenge them - perhaps even hand them a few party KOs - but without discouraging them from continuing the story. As a result:

• Duty Finder does not prepare outside the game, and rarely prepares outside attempting the duty. Part of why the Hall of the Novice update is so welcome is that "try this solo instance" is about the limit of what casual players can get each other to do, even when the whole party is cooperating. (After all, the player may be on console without a separate device to watch videos or browse guides.)
• Duty Finder is perpetually operating on reaction, since first-timers cannot plan for what they have never seen and even players who complete the duty frequently are unlikely to have its timeline memorized.
• Duty Finder's practical response times are not fast, so the player needs a generous lead time and long periods of information visibility. They are not looking in the right place when the duty first gives them the information they need, they have never actively thought about how they can efficiently move their eyes between information sources in the interface, and they will take time putting it all together - which have so much impact that I often beat my matchmade parties to safespots. (They may also be looking away from cues entirely in order to raise or read party chat.) I have even been regularly beating able-bodied and experienced healers to raising KOed party members in current duties ever since I settled into endgame, despite that being a reaction test that is not even from the duty timeline - because I have taught myself to notice KOs while keeping my party list in my peripheral vision, and they have not. (Sometimes my party position tracking is enough to alert me of the impending KO before it happens, and I can target the player character and use Swiftcast before they even go down on my screen.)
• Duty Finder does not prepare for specific mechanics, as a natural result of not knowing what mechanics are coming. This is why Thunder God Cid did not work out and had to be nerfed: getting three player characters into each tower and getting all of the six Shadowblade AOEs to avoid overlap are more advanced versions of spreading, and thus reliability requires planning who will go where in the Party Finder and Raid Finder style.
• You can get Duty Finder to do a tightly limited amount of general preparation - roughly one thing that is useful without foreknowledge of specific mechanics - before players run from the starting area of the duty in impatience. In Stormblood, that one thing was "wait for the healer(s) to cast Protect on every party member." In Heavensward and A Realm Reborn, that also included waiting for Stoneskin if the party had a CNJ / WHM. In an alternate timeline, that one thing easily could have been agreeing on "clock spots" via a macro or around a waymark, that would then be reused for a variety of mechanics just as Party Finder and Raid Finder do now. (If you really wanted to, you could still do that, but it would require a substantial amount of teaching players new expectations from Hall of the Novice up.)
• Duty Finder requires most mechanical mistakes to allow for recovery and duty completion within that attempt. With no plan and no foreknowledge, they are not going to reach the standard of making very few mistakes per attempt in any reasonable period. As such, mechanics should be carefully examined for what sorts of mistakes can result in an immediate or delayed party KO. Most mistakes should not even guarantee an individual KO for players with good equipment, healing, and mitigation.
• Duty Finder accepts only infrequent penalties for its lack of preparation specifically. The player, so long as they are reasonable, will not be angry at their party members or the game for occasional stacked spreads and similar mishaps. But they will get frustrated by too high an incidence of such things, so mechanics that invite such mishaps are ill-advised.
• Skilled players in Duty Finder will have a completely different experience than the baseline, and it is up to you to make sure they also have a good experience by giving them opportunities and rewards for skill expression. Do not neglect those opportunities in even the easiest of duties. As I have explained regarding healer design and magical ranged DPS design, anything that is appropriate to put in front of casual story-oriented players will not challenge a party that is comfortable in Extremes, let alone Ultimates. (And even the "worst reasonable" players will have a very different experience the 90th time they see the same dungeon in Expert Roulette.) Since your gameplay loop puts all those players in regular duties, that different experience is either going to be occupying themselves with skill expression or it is going to be boredom. Simply completing the duty faster is not enough - strong players can do that while falling asleep by already-acquired skills alone. This is why I made a point of bringing up the connection to job design and satisfying mechanics.

This same approach is why the casual player does not learn complex concepts unless the game (or a fellow player) takes the time to break down the concept into small pieces and teach them that way. Their tracking of effort and reward is long-term, but they are only willing to put in effort in the moment. Guide the casual player to one step at a time, and they will climb a mountain. Set a hill in front of them with no explanation, and they will perceive it as a wall.

--------

In contrast, high-end duties must be designed for the different expectations of players tackling them. The player still likely does not have any grand plan of personal evolution, but they do have a plan to improve at the duty in front of them. While Duty Finder lives in the moment, Party Finder, Raid Finder, and statics live in the future: job performance standards, party coordination, advance planning based on the duty timeline, practicing their execution of those plans in progression... and for putting more effort in, they expect more rewards out in how the duty treats them. Not just rewards in the gear, but in the gameplay experience of the duty itself.

(Of course, there are degrees to this. A player in Extreme probably remembers their plan for a mechanic upon seeing the castbar and executes it with some hesitation, which is the right general idea of planning, but not nearly enough for Ultimate plans. For example, consider shielding TOP's final iteration of Wave Cannon correctly on SCH: in both cases the party must move around extensively during the Wave Cannon castbar due to Alpha Omega's preceding mechanic, so the correct time to cast the first shield without expending Swiftcast is during the preceding castbar while looking at the preceding mechanic's cues, planned carefully not to be so early that the shield expires before absorbing damage. (Raising is almost useless there due to the structure of the Dynamis buffs, so one can expend Swiftcast to shield without much of a tradeoff, but it is more fun not to.) And then the second shield must be in place 6 seconds after the first shield is fully used, in which time the SCH must also move and help heal the party back to full HP, so the slightest hesitation will lead to a party KO. Oh, and they must also remember the other mitigation that the SCH and the party as a whole must apply to every hit, regardless of what cooldowns they have planned to use where. The whole plan is based on knowing when each of Alpha Omega's attacks will land before they even begin, and it is a delight, but a player new to high-end duties will not be doing it.)

• In high-end duties, the player expects and wants to prepare outside attempting the duty, including for specific mechanics. Remember the community reaction declaring Deltascape V1.0 (Savage) (O1S) too easy because the world first completion took only a single attempt? It is too easy for Savage, but let us be precise: the world first team was able to complete it in a single attempt because they could use their brief exposure to the Normal version and their general preparation to do the Savage mechanics by reaction. They had good reaction times on their side - that is a physical attribute world progression teams select for - but that would not have averted a party KO if the duty had forced them to make a mechanic-specific plan. (The very first time I entered O1S, without a timeline, I could not do its mechanics on reaction, but I could see exactly why the world first team could.) Nothing in O1S demands a mechanic-specific plan, and that makes it too easy for even Extreme.
• In high-end duties, the player's practical response time for a mechanic improves significantly over the course of learning that mechanic, and more so the more extensive their planning. Here the player can plan to always be looking in the right place when the duty first gives them the information they need, to move their eyes efficiently between information sources, to preposition their character for the possible random variations of the upcoming mechanic. They can work out how different sources of randomness combine in advance. They can practice executing their plan efficiently and without hesitation, even if they never actively think about prepositioning their hands or distilling animations down to the first visual cues.
  TOP provides an excellent example of this in the opening to Run: ****mi* (Omega Version), popularly known as "the Predation Dodge" due to its resemblance to Ultimate Predation in The Weapon's Refrain (Ultimate) (UWU). Those readers who are unfamilar with TOP can see many examples of the Predation Dodge in real time, courtesy of Sausage Roll's study playlist. Without knowing that the footage is from TOP, the casual player observer can tell that the mechanic is not suited for them; they may not identify it as an Ultimate mechanic, but they understand that putting it in Alphascape V4.0 (O12N) would be a miserable failure. And they are correct - even though every single one of those AOE shapes is known to and manageable for the casual player in Alphascape Normal, as is the notion of AOE combinations! (Omega-F casts her cross AOE, Optimized Blizzard III, and her "Hot Wing" front-to-back-line-safe AOE, Superliminal Steel, in O12N. Omega-M likewise casts his point-blank AOE, Efficient Bladework, and his donut AOE, Beyond Strength, in O12N; in fact, the safespot for Beyond Strength is bigger in TOP to accommodate the combination with Optimized Blizzard III. Both have the same weapon tells in O12N that they do in TOP. (They do cast their AOEs simultaneously in O12N too, though there their positions and facing are dependent on the tanks' choices and the combinations are more limited.) And while the casual player does not get to fight Final Omega, they have seen the hourglass AOE sequence with the same safespots in Alphascape V1.0 (O9N) as Latitudinal Implosion / Longitudinal Implosion.) The casual player even has the notion of remembering an AOE pattern without orange telegraphs on the floor. So why is the Predation Dodge out of reach of the casual player?
  The mechanic is purely individual dodging, so there is no coordination requirement. There is no unavoidable damage in it, so there is no healing or mitigation requirement either. And while there is a hard enrage to the phase as a whole, the damage output requirement is not what the casual player is looking at. What they correctly perceive is that they do not know where to go fast enough, even though all the information has been given to them.
  The casual player does not know that the Omega-M and Omega-F pairs always spawn on opposing intercardinals, and so does not know to begin the mechanic looking at an intercardinal with their character prepositioned on a cardinal such that all intercardinals are visible on their screen, nor to look immediately at the opposite intercardinal after locating one clone and processing its weapon. The Ultimate raider at this point in TOP progression does know all that, and even in blind progression has been primed to expect opposing spawn locations by the earlier appearance of a single pair in TOP Party Synergy. (In fact, I check the pairs for the Predation Dodge entirely by looking at the arms of each clone, since three of the weapons are visible there and "no held weapon" is always the skates for Superliminal Steel. Only Party Synergy has a decoy clone.)
  The casual player has to work out what the overlap of safespots is for each combination while the cues are being shown to them. The Ultimate raider at this point in TOP progression memorized each clone pair combination back when learning Party Synergy, and now can memorize how each clone pair overlaps with the hourglass cleaves from Final Omega. (And from there, how each combination flows into each possible other combination, all outside the duty. That is a large part of why the study playlist exists.)
  The casual player has no practice in picking the light gray weapons on the dark gray clones out of the mid-gray floor, gold overlay or not. The Ultimate raider at this point in TOP progression has practiced that abundantly for Party Synergy and still fully expects to take several individual KOs - potentially cascading into party KOs - here. Indeed, by this point the Ultimate raider has gone through hundreds of party KOs in TOP, and the casual player is not willing to persist anywhere near that long for a duty completion.
  If you immensely slowed down the presentation of cues in the Predation Dodge to allow the casual player time to work out the overlaps while in the duty, and guided the player's eyes around the arena by presenting one clone at a time with the matching weapon transition / creation animation followed by a "freeze" to have them go off in pairs, that version of the mechanic could have gone in O12N and posed little difficulty to the casual player, even with the resolution of the AOEs at full speed. (The only remaining pain point would be working out the correct distance from Omega-M without waymarks for the Optimized Blizzard III + Beyond Strength combination.) The difference between this hypothetical Normal Raid version and the extant Ultimate mechanic is purely the processing speed demanded of the player, which the Ultimate raider attains through planning for and practicing upon the specific mechanic. Their underlying reaction time need not change at all - for me the Predation Dodge went from scary to comfortable within a few hours of study and practice, long after I had already plateaued in reaction time. Even before knowing the snapshots, I have extra time for the movement in all but the "outer" combinations (Optimized Blizzard III + Efficient Bladework) that have the longest distances to move between the two safespots.
• In exchange for their effort in planning and holding themselves to a standard of job skill, the player expects that a reasonable amount of effort put into a high-end duty will yield a plan that (when executed correctly with practice) leaves them ample leeway on their reaction time and results in a "clean" duty completion every single time. ("Clean" as in "no KOs, Damage Downs, or other penalties for failing mechanics". Recovering from failed mechanics necessarily involves reaction to the failure.) High-end duties can, should, and generally do move faster than regular duties, but only ever to force the player to have a plan and execute it with appropriate fluency. Give them all the information they need well in advance and then demand that they act at a specific time, or even perform a sequence of actions without pausing, yes. Have some of that information include debuffs and/or markers given to other players' characters, yes. Demand that they act while receiving and processing information about the next mechanic or portion of a mechanic, yes. Display some necessary information for a limited time, yes, though the tightest limits are only viable in Ultimate to force the player to work out an order of processing and remembering cues. And all that will result in mechanics that can feel reactive while the player is still learning them, because the player does not have the practice with their plan to process the specific mechanic efficiently at that point. But that feeling must disappear with a reasonable amount of practice, as the player's practical response time for that mechanic improves; once the player learns the mechanic, they should have enough time left over that reaction time is not practically a factor. Mastery should feel turn-based, recovering the feeling of the older Final Fantasy games in the real-time format of this one - half the battle was already won in the planning stage.
• Accordingly, any mechanic that requires the player to react to another player's choice, or where all plans that eliminate such reaction are unsuitably elaborate for the duty, is unacceptable in high-end duties. (Responding to information that another player must pass on, such as in Eden's Promise: Anamorphosis (Savage) (E11S) Prismatic Deception, requires a long lead time - for two players' reaction times and an extra client-server round trip - and a way for players to pass on that information without relying on typing, macro space, or out-of-game communication. E11S did it right, but such mechanics can never be fast.) Yes, this means that some mechanics are permitted in regular duties but not in high-end duties. Players put more effort and resources into individual attempts at high-end duties, and taking a party KO or even a personal penalty there because of forced reaction is annoying.
  (TOP is perfectly fine in this regard, as players can devise a priority system for the "open choices" presented by Run: ****mi* (Sigma Version) and Run: ****mi* (Omega Version) without undue effort in planning or execution - that players in Party Finder turned to third-party tools instead is not the duty designer's fault. (Technically, there is also an "open choice" in Run: ****mi* (Delta Version), but that is assigned even more easily using the Remote Regression and Local Regression tethers from Patch.) I even know how to take the priority system management load off my party with manual sign placement, I just do not have the APM to personally place the signs!)
• In high-end duties, the player expects the duty to present a deadly dance with harsh penalties for failure. That is not just part of the difficulty, but also part of the thrill of attempting these duties. If you feel inclined to build in lenience for failing a specific mechanic in a high-end duty, rather than overall tuning allowing for a small number of mistakes, the mechanic is almost certainly not suitable for high-end duties in the first place. That desire for lenience is a sign that the mechanic does not yield to appropriate planning.
• While high-end duties require a degree of skill from the player to have the baseline experience, the player actively expects satisfying mechanics for their trouble, along with ample opportunities to express their skill beyond the baseline. Not every mechanic must be satisfying in itself to execute correctly, but it is a disappointment that there is not a single such mechanic in all of AAC Light-heavyweight Savage.

In high-end duties, the player will solve puzzles and learn complex concepts from fellow players entirely to the extent that they grasp the bigger picture. (Arena floors may be almost entirely flat after players learned to tilt Twintania's divebombs in The Binding Coil of Bahamut - Turn 5, but their real third dimension has always been time.) The best players work with the whole duty timeline and the whole party as one, creating a single branching plan that is executed by eight players - and in return, they expect both the planning and execution layers to provide them the joy of smoothly countering the duty's every move. The party of players as a whole takes on the role of the single "planner" player controlling a party of characters in the older Final Fantasy games, as well as their individual surface roles in executing their plan.

So in high-end duties, give us planning problems that are worth the effort. Reward the spreadsheets, the hours at a striking dummy, the hours in a single duty. Reactive, otherwise easy mechanics can never do that.

--------

For all their faults, Heavensward, Stormblood, and Shadowbringers were close enough to these principles to inspire me to fight. To work my way up from being a terrified new sprout to being Ultimate-ready. (It definitely helped that Shadowbringers had very nearly the perfect job for me.) But also to do roulettes long after I had hit my weekly cap of Allagan Tomestones, with no jobs to level, just for fun. I did not mind even Sastasha, because I could still enjoy myself there by working with my job.

Endwalker broke my job, and that was a massive blow to me; this letter is about duty design, but I have always been a systems-oriented player and the jobs are the core systems of combat. I role swapped to keep raiding, because the personal improvement in working with my physical limitations was and is important to me. I did roulettes comfortably to reach my weekly cap of my Allagan Tomestones and level jobs, but not beyond that.

Dawntrail essentially did not change my job. Yet it is the first time I have had to actively remind myself to reach that weekly cap, despite having jobs that could use the Allagan Tomestone gear. A habit sustained for years was broken in less than two weeks, because the Dawntrail duties are so physically demanding that my brain notes the excessive injury risk and pushes me away from them. (Of course, the fact that I am forced to sit out of FRU by Patch 7.0's duty design direction does not help. Ultimate-level skills do not do much for me if I cannot safely commit to the Ultimate.) And injury risk aside, the Patch 7.0 duties are boring!

What Dawntrail has moved me to do is play more RuneScape. In some ways that game is less accessible than this one - its combat system has a far higher APM ceiling than even the old manual Embrace SCH - so I cannot play it perfectly, but at least I can play its combat system solo in my own way and the game will reward me for that. They have something for me, and that is better than having no duties that suit me.

Their latest boss rush dungeon is not only great fun, it copied your attack telegraphs and your system of limited KOs in Duty Support trials to teach the player mechanics... to wild popularity of the Hard Mode that turns the KO forgiveness off! Their "tutorial boss" for recurring mechanical concepts has not only instructional NPCs but individual mechanic selection with a randomized order, allowing the player to choose what they want to focus on practicing... and its own Hard Mode that activates and amplifies all the mechanics. With the mechanics still randomized and repeating once the boss has finished the set, so the player must have at least two plans to avoid KO from each mechanic, in case the boss picks it twice in a row and their cooldowns from their first plan are unavailable. And the Hard Mode smoothly scales from low Extreme to mid-Savage difficulty for an individual completion, while offering a "killstreak" system that gives better loot for chaining up to 200 kills in a row with increased scaling each time - successfully rewarding Ultimate-level consistency! Even their latest gathering boss deliberately allows for strategies ranging from slow but low-effort to fast but high-effort, to encourage and reward skill expression.

(RuneScape is not just copying the best aspects of your duty design history, by the way. They have also been seriously improving their tooltips, combat feedback, and overall new player experience, including by launching their own version of a pet class last year with a full teaching questline. Which inadvertently resulted in their player community learning your term "job gauge", because a player made one for them and it took off enough to start supporting their other forms of combat. Their players can literally bring them the best parts of your interface.)

There is nothing quite like raiding in your game when you design it right. RuneScape would not be in serious consideration for combat challenge if you did. But Patch 7.0's duty design direction has lost sight of both long-term fun and appropriate challenge.

Give us challenges we can all rise to meet and aspire to master, not barriers we cannot do anything about.

Now to go over the existing Dawntrail duties and their mechanics from a duty design perspective, both the good and the bad. As a reminder, this section is necessarily tedious, and any player readers who are not interested should skip over to my redesign of the Honey B. Lovely duties. (Navigation links are at the top of this thread.)

Certain duties are very time-consuming to repeat with exact mechanics and tuning for review, namely MSQ solo instanced duties, hunts, treasure dungeons, regular FATEs, and World FATEs. Due to the inordinate time that would be involved in reviewing them thoroughly without access to test servers, I have omitted these to focus on the 21 readily repeatable group duties available. Most of the omitted duties follow much the same design style as the ones included, so this along with the discussion of principles should apply well to their issues.

----------------

Due to length, I have split this section into four subsections, in increasing order of acceptable and expected difficulty.

I will begin with the duties appearing in the Patch 7.00 MSQ, in the order of their appearance there. As completion of these duties is functionally mandatory for a player to continue with the game, these duties must be held to the highest standards for providing an encouraging experience to even the returning and very weak player. They know that their access to all further content - combat or otherwise - relies on their completion of these duties, and that knowledge becomes a heavy weight if they worry about their ability to do so. Enjoyment for the stronger player and in subsequent completions of these duties must be created without compromising those standards.

----------------

Ihuykatumu (Level 91 Dungeon)

https://i.imgur.com/SiqXaNm.png

Mobs Before First Boss

Three Packs (On River Boat)

These are forced single pulls (one pack at a time), and they do nothing more than a small amount of basic ground AOEs. If you are going to make the player do single pulls, make them individually interesting so that stronger players do not get bored. Throw in some more fish that come to feed on the remains of the previous pack as each mob dies, or some pattern AOEs from Bakool Ja Ja's "gift", or something. It does not have to be as resource-draining as the Holminster Switch forced single pull section, which has no downtime.

Being the opening of the first dungeon of an expansion, it is entirely possible that this is the first group duty the player has done in a while, so a gentle opening is reasonable - but that can still be interesting! This is not, and that makes the story sequences all the more grating when running the dungeon to level rather than for MSQ.

First Boss - Prime Punutiy

https://i.imgur.com/LIrdYH7.png

Hydrowave

Narrow cone AOE, easily dodged, nothing special. It is fine.

Resurface

The AOE spam after the castbar is visually overwhelming with its speed, since the Flytrap add casting the donut AOE appears during that AOE spam and forces the player to look at more than just their character's feet. If the player could just look at their character's feet, it would be visually on the high side but acceptable, because the player could and would remove distractions and speed up their time to see the relevant cues by focusing under their character's feet.

I can do this, but it is not fun - all it encourages me to do is try to constantly look everywhere as best I can, especially the first time when not knowing about the Flytrap. Even when I do know about it, the Flytrap is usually pretty far from my character's feet, forcing me to visually bounce back and forth between looking up at the arena as a whole and down at my character's feet until the AOE spam ends.

Suppose instead that the boss pulled up the Flytrap during the castbar, putting it in the arena and on the player's enmity list. The AOE sequence could then play out the exact same way, but the first-time player would have a reasonable chance to see the Flytrap and think "this probably does something" before the donut telegraph appears and probably covers the whole area around their character. On future attempts, stronger players would have the option to path to the known location of the Flytrap during the AOE spam, making it more interesting in how that interacts with their job.

Song of the Punutiy (Adds)

This is a good mechanic, with light damage from the spreads going into increased tank healing from the adds. Allowing the party to bait most of the adds' positions for when they first become targetable could provide an opportunity for skill expression, although here it does little because two of the spreads are so large.

Shore Shaker

Easy but fun - just step in when the telegraph in front of one's character disappears, and since the player knows to look there based on seeing the sequence of telegraphs appear, the timing is fine.

Overall Impression (Boss)

This boss is pretty easy other than Resurface being an issue. As the first group boss of the expansion, starting off with a visual design issue raises concerns to the player immediately.

Mobs Between First and Second Boss

First Pack - Bats and Widowmaker

Nothing interesting here. This does allow for a double pull, though, so it is fine.

Second Pack - Doblyns and Treant

Ditto.

I appreciate not having to target and interact with the ladder once this pack dies, and the same goes for the later ladders in this dungeon.

Third Pack - Mimiclots

The idea of "they've copied our forms" is cool, but nothing seems to be done with it other than apparently making these enemies stun immune - and the player is not going to see that without a WHM in their party. Even something simple like giving them each a basic AOE that draws on a "classic" job in the role (though some of that recognition has been eroded by job changes) would sell the concept mechanically: front-facing cone for the tank (MRD/WAR), point-blank AOE for the healer (any healer but AST), line AOEs for the DPS (LNC/DRG).

(The healer does have such an AOE, but none of the others do and the cast does not look like Holy. Plus I had to find that out from videos, not from my own completions, despite using Duty Support first for immersion and then to avoid hampering other players' experiences when reviewing the duty.)

This pack does allow for a double pull, so it is not boring, just a missed opportunity.

Fourth Pack - Doblyn and Widowmakers

Still nothing interesting here.

Second Boss - Drowsie

https://i.imgur.com/KbPsEVV.png

Sow

If you are expecting the player to see where the seeds dropped from this cast, that expectation is unreasonable - they are too small and show for too short a time for the player to track four objects. Since players have sufficient time to see them from the later Drowsy Dance, telegraphs, and Wallop, this winds up mostly being a "warning, something happened" cast.

Drowsy Dance

Grows the seeds with telegraphs, fair enough.

Wallop (cast by adds)

I appreciate the facing markers here, which make this reasonable to dodge since the player can just look at the base of each root. Having to follow each root upwards to find its facing direction would be too much.

Mimiclots

Since the Mimiclots return from the earlier trash, this is a great opportunity for them to do a mechanic (with the player having a good guess based on the trash) every time, rather than simply being AOE fodder when I completed this with Duty Support. Sometimes having AOE fodder in a boss fight is fine - it provides a change to the healing requirements and an opportunity for everyone to switch between single-target and AOE rotations - but this is a missed opportunity to do something more than that...

From seeing videos of other players' completions, the Mimiclots do have a mechanic here! Why can Duty Support completely eliminate a mechanic?

Sneeze

Big cone, long cast to dodge it, no problem.

Overall Impression (Boss)

The name of this boss and the fact that it starts sleeping do suggest sneaking past, which inevitably creates tension between players who want to leave it be based on their character and players who want the XP for defeating it. To remove that tension, perhaps this should have been a point where one of the other claimants' teams forced the fight?

Mechanically, this is okay.

Mobs Between Second and Third Boss

Four Packs

Nothing mechanically interesting here, but can be double pulled.

Thematically, the dead animals are a cool hint at the boss, but that is not mechanics.

Third Boss - Apollyon

https://i.imgur.com/kDPMpD9.png

Razor Zephyr

Basic line AOE, reasonable cast time.

Appearing Beasts

A thematically cool "warning, something happened", and I like the red text that hints at how the boss has powered up.

Swarming Locust

I do not love the boss jumping around without prior telegraphs for the sequence, just from a visual tracking perspective, but it is acceptable in moderation in contexts like this. The jumps are not that fast, and the next attack does not require knowing the precise position of the boss.

Blades of Famine

The empowered line AOE, and since there are four of them, being away from the boss after the jumps is helpful to the player rather than annoying.

Repeats twice, including the jumping around.

Levinsickle

I quite like this mechanic. Each of the sets of radial cones covers the arena across its sequence, and there are two of them so the player wants to step from the empty space into the overlapping space. It encourages the player to think, with all the information clearly given and a reasonable amount of time.

(no castbar) Appearing Beasts

A different mini-phase, this time adding a donut after the tankbuster.

Windwhistle

I do not love having to keep an eye on the resulting Whirlwind for so long as it pulses the snowflake of line AOEs. I find it doable, but it is definitely a visual stress since its movements are not cleanly predictable.

Overall Impression (Boss)

This is actually a pretty solid boss, despite Windwhistle. The concept of a boss that eats other predators and takes on their attacks is a good one, and could allow for a degree of "mechanics the player has already seen, but empowered / in succession" without the overwhelming nature of Tower of Zot's final boss.

Overall Impression (Dungeon)

The boring opening is not great... and Resurface looms over the entire dungeon. It put me in "try to constantly look everywhere" mode from the beginning, and that is not fun.

--------

Worqor Zormor (Level 93 Dungeon)

https://i.imgur.com/Ec8sDtk.png

Mobs Before First Boss

Four Packs

Nothing mechanically interesting here, but can be double pulled.

First Boss - Ryoqor Terteh

https://i.imgur.com/t3L2W4N.png

Fluffle Up

An AOE sequence from adds, that the boss modifies by freezing some of the adds. The first two show the patterns from the different types of adds; the small Rorrloh Teh on the cardinals hit a quadrant of the arena in front of them (safespots for each hit opposite each other), and the large Quorrloh Teh on the intercardinals hit a point-blank around them (safespots for each hit beside each other). These first two are fine.

Good visual clarity with the ice tethers, guiding the player to look from the boss to the adds.

Snowscoop

Nice sidestep between the two sets of line AOEs through the center of the arena.

Sparkling Sprinkling

These spreads appear during the preceding mechanic, allowing the player to notice others near them and try to dodge away. You still are going to get players overlapping spreads, because Duty Finder does not plan spread locations, but this is a mechanical overlap that is easy enough for MSQ.

Fluffle Up (repeated)

Now the two patterns are combined, and the boss puts out four ice tethers. While the boss is reversing both patterns, with each reversal previously shown individually, the first-time player does not know that and may not even know that the individual patterns are always reversed. For all they know, the boss could be manipulating each add individually to create a sequence of three hits instead of two, or only reversing one of the patterns.

So when the tethers come out, the previously working method of "stand in / note the first safespot, then if tethers appear, move over to the area that would be hit by the tethered adds" no longer works. The first-time player must try to remember which of the adds were going to go off first, which is itself not easy as a spatial memory mechanic with four things presented at once, and match those up with the tethers. That is not really reasonable to pick apart in the time given.

Once the player has time to think about this mechanic after the battle, it becomes fine or even too easy. "The boss is going to reverse both patterns, so stand where all the orange is" is the obvious guess and the correct one. But this is an MSQ dungeon, where the player's first attempt is a critical story experience, and many casual players do not have the notion of hidden patterns in mechanics at all.

This is a difficult mechanic to fix, because you also do not want to slow down the battle so much as to bore players on repeat runs. It may well be more suited to a Trial or a Normal Raid than a dungeon, as those allow for longer bosses that can do things like refresh the player on Spell-in-Waiting markers and still have time for another concept.

Overall Impression (Boss)

This is a good boss concept, focused on manipulating the timing of AOE sequences by freezing adds. It only gets to be too much for the first-time player during the later Fluffle Up casts.

Mobs Between First and Second Boss

Four Packs

Nothing mechanically interesting here, but can be double pulled.

Second Boss - Kahderyor

https://i.imgur.com/a8N44lx.png

Crystalline Crush

The stack tower here sets the party up for success with the immediately following Wind Shot, which is especially nice since this boss has a substantial amount of unavoidable damage.

Wind Shot (Player Donuts + Crystals Safe)

If the party stays where the tower was and the crystals appeared, they will be fine! Good introduction to what this wind mechanic does. The color for the wind AOE could use a sharper green, though.

Earthen Shot (Spreads + Away From Crystals)

Again a good introduction, since the player should reasonably guess that a different element requires a different response, and the last one was "stay where the tower was". Though the color for the earth AOE could use a sharper yellow even more than the previous mechanic, or potentially even replacement with another element given the later mechanics.

Crystalline Storm (Line Crystals)

Three small line AOEs that leave behind crystals.

This is why the elemental AOEs could use sharper colors. When the player has to check the AOEs from all three lines at once and react to their party's movements for both donuts and spreads, high contrast really helps them process the visuals quickly. (Having only seen each element once before, the first-time player may not have picked up on what the tie is between the crystal locations and the elemental safespots. Casual players have a lot to think about going in!)

Wind Shot (Player Donuts + Crystals Safe)

Stepping into any line will avoid the ground AOEs, so the player need not remember or remind themselves of all the line positions. There is a decent amount of room for stacking or spreading the donuts.

Earthen Shot (Spreads + Away From Crystals)

Now there is little room, the party is reacting to each other for spreads, and the earth AOEs are yellow on light brown. I imagine this is miserable for anyone with reduced color vision in the red-green range, even if not full-blown colorblindness.

Seriously, perhaps the earth mechanic should have been replaced with fire or lightning instead? Fire is already associated with ground AOEs, lightning is already associated with spreads, and the embedded crystals are purple.

Seed Crystals

Break out of the crystal adds. Easy enough.

Sharpened Sights

Enters a gaze + AOE mini-phase, casting first Stalagmite Circle (point-blank + look away from boss) and then Cyclonic Ring (donut + look away from boss).

The gaze combined with AOE is very well done, showing the player the combination where they will naturally succeed before asking them to respect the gaze while dodging the ground AOE. Using the contracting gaze ring rather than the icon on the enemy is an excellent way to get the player to see the gaze in this context, especially since the head of the boss may be off their screen.

Overall Impression (Boss)

The elemental effects really could use some more pop on the ground, to maintain visual clarity when they are coming from the crystal lines. Otherwise, solid boss.

Mobs Between Second and Third Boss

Four Packs

Not much mechanically interesting here, but can be double pulled. A good tank can and should bring the Cloudtrap into the group or vice versa.

Third Boss - Gurfurlur

https://i.imgur.com/WI7qdQj.png

Stonework (Fire, Expanding Square AOEs + Spreads)

A combination mechanic right from the start, but giving the first-time player plenty of time to expect something big and understand that the mechanic comes from the stone plate.

Stonework (Water, Knockback)

This one is not a combination, but it shows the player that the water knockback comes from the side with the plate, not the plate itself, which is important later.

Sledgehammer (Multi-hit Line Stack)

From a raid healing perspective this does not hit that hard, but anything that did would be too much for casual players. As it stands with the big marker and threatening name, this slow cast is enough to invite a casual healer to prepare to heal the party rather than reacting, and to invite the rest of the party to mitigate, both of which are good skills.

Arcane Stomp (Damage Up Orbs)

The large number of orbs to intercept here creates a faster section, encouraging players to use their movement tools efficiently and split up the arena between themselves. Good for making the boss "less of a striking dummy" without compromising on the slow, powerful attacks of a giant. (On repeat runs, the player may well notice that the boss only intercepts the orbs when they reach the middle of the arena, not his hitbox, and use that fact to make smaller movements.)

Notably, the buff from the orbs only lasts a minute on player characters, but lasts indefinitely on the boss, which looks scary from the casual player's perspective. And this being a leveling dungeon, it has a tight implicit Item Level sync.

Enduring Glory (Raidwide)

Again, this does not hit that hard from a raid healing perspective, but a raidwide after a damage up is scary in concept.

It does cleanse all the Damage Up stacks from the boss, which is really probably not necessary - Sledgehammer as it stands can be completely countered by Liturgy of the Bell / Seraph / Macrocosmos / Panhaima if the healer brought in Anabaseios tome gear. With this much lead time and seeing a whole mechanic dedicated to the Damage Up stacks, a casual healer can be asked to find more than one AOE healing tool; this could cleanse half or two-thirds of the stacks instead. Even if they do not find more than one AOE healing tool, the party will probably just be left at low HP instead of KOed.

(And if the party does take a party KO to some of the Damage Up stacks remaining, they can just try again with knowledge of the Damage Up. The blatant obviousness of Arcane Stomp makes sure that the (reasonable) player notices the orbs and infers that the orbs power up the boss even if they do not look at enemy buffs, so they are unlikely to blame the healer.)

Stonework (Wind, Tornadoes + Double Knockback)

A combination that invites the player to think about using their knockback immunity! The first knockback from the center plate will likely be taken to a corner since the player does not know the distance, which leaves the player well set up to position for the tornadoes before they start. The second is much less immediately pleasant, having the potential to knock the player character right into the tornadoes just before they leave the arena - but from a duty design perspective this is great, as it presents a problem to the player while making sure the tools to solve it are at their fingertips.

Stonework (Fire + Water)

This looks scary, and what that does is encourage the player to think about the combination and how the mechanics interact. The water knockback is timed to match the first hit of the square fire AOEs, so the player can have it wash their character right into the safespot and then just move for the spreads. Having seen the pieces of the combination earlier, the player is empowered to make a guess at the timing and prepared to move their character shortly after the knockback.

Sledgehammer (Multi-hit Line Stack)

As before, but coming right after a raidwide this time. The long cast makes it still very reasonable to heal.

Overall Impression (Boss)

Great! More like this for dungeon bosses. Visually clear, the voice sells the "weight of history" idea, and the mechanics are gentle enough for a casual first-time player to understand their mistakes while having frequent combinations and optimization opportunities for strong players. The spectacle both tells the story and helps cue the player, which is not possible for every concept but awesome when you make it work.

Not every dungeon boss should be this slow, but for a giant it works to have slower and more powerful attacks.

Overall Impression (Dungeon)

So close to being fun all around without issues, and at least the issues do improve on repeat completions. The mechanics carry out the themes very well.

--------

Worqor Lar Dor (Valigarmanda, Level 93 Normal Trial)

https://i.imgur.com/JZXYC62.png

(no castbar) Animation Tells

This boss has three recurring animation tells for basic attacks using three elements: the wind swirling donut (center of the platform safe), the fire stomp (leave melee range), and the lightning beak cone (front corners of the arena safe). These animation tells are supposed to be taught by the red text the first time, and then the player is supposed to recognize them with no other cue thereafter.

So let us consider these three animation tells and the text that is meant to teach them.

• Wind: "A tempestuous wind gathers around Valigarmanda." This directs the player to look at the boss model for a swirling wind, suggesting a donut around the front of the arena... which is not where the wind is, the wind is around the platform instead and Valigarmanda later enters the swirl. The most accurate lesson the player can learn from this is that your pop-up text can be wrong.
• Fire: "Valigarmanda gathers its strength." This says absolutely nothing about where to look. (Where to look is at the boss's legs, since its body starts glowing red and its legs are close to the platform.)
• Lightning: "Blinding energies build within Valigarmanda's wide-open beak." This gives a correct description of where to look - but Valigarmanda raises its beak off the screen during this animation at the default camera tilt or even slightly higher, and most players are never going to adjust their camera tilt. Indeed, there is no way to get Valigarmanda's beak comfortably on the screen without sacrificing ground visibility either camera-ward of one's character or by looking very side-on, which results in the player fighting their interface to both see this tell and dodge other ground AOEs smoothly. To make matters worse, the boss also glows red during this animation!

https://i.imgur.com/bS0ncqM.png

Every one of these teaching texts is either useless or wrong. One of the tells is not even reliably on the screen. And this is for a mechanic set the player is meant to recognize by animation alone while other things are going on in the duty?!

What I did in response to this is launch straight into the "try to constantly look everywhere" mindset and stay there even for subsequent MSQ duties, combined with following the Duty Support NPCs for the animation mechanics in this duty once I realized I was better at tracking them than finding the then-mysterious-unto-me animation tells. This is not fun, but it was the best option open to me short of finding a guide and finding and memorizing a timeline for a Normal Mode trial.

You can have animation tells without other cues; the old version of Aurum Vale did it and was fine. You can have animation tells on a large boss model if you are careful; Alphascape V2.0 (O10N, Midgardsormr) did it and was fine. But you have to make sure the player actually gets to learn the tells, and can see them consistently without having to try to constantly look everywhere. "Welp I got hit by this animation tell mechanic shortly after this castbar attack, somehow I am supposed to look at both at once" is not actionable!

(There is something of a rhythm to when this boss uses its animation tell attacks; I was much more comfortable returning to this trial after doing the Extreme version. But that does not help any player who does not complete the Extreme version. Plus between this being a high-level trial and having Duty Support, even players who queue Trial Roulette will see it very infrequently beyond their own first time - the first-time experience is not just critical but the default, and then the player does not have any knowledge of that rhythm.)

One way to significantly improve this without making major changes to the boss concept:

• Wind: Change the teaching text to "Valigarmanda creates a tempestuous wind around you.", "Valigarmanda creates a tempestuous wind around the platform.", or something along those lines. The player must be directed to look at the tell that appears with a reasonable lead time, which is the wind gathering at the corners of the arena. The wind effect could also use a little more green for contrast.
• Fire: Change the teaching text to "Valigarmanda channels fire into its legs."
• Lightning: The teaching text is accurate, if incomplete. Change the animation tell to start with the boss lowering its beak right onto the platform, so the player can reliably see it at the default camera tilt. Change the glow on the animation to be a purple lightning glow from Valigarmanda's head only. These changes would make the color and motion also dramatically distinct from the fire attack.
• All: These changes allow the player to "rest their eyes" on a spot roughly around the resting position of either one of Valigarmanda's legs, because all the animation tells are in close proximity to those spots and the teaching texts show them that. This notion of a "resting point" is essential to having the player see animation tells without giving them another cue for when to look at the animation. (Aurum Vale's "resting point" was the bosses themselves, those being significantly smaller and having few castbars to look at. O10N cues the player every time, on top of having very dramatic animation tells that are visible across a large portion of the screen.) "In this duty, look at this spot when not looking at the enemy's castbar / my character's feet / the party list / ..." is actionable, and allows you to have an animation tell mechanic whenever the player has some time to return their eyes to the "resting point" after the preceding mechanic.

A second issue with these animation tells is that this boss is not actually programmed to show the player all the different elements before the repeats begin. It can roll wind, fire, wind, (other attacks)... and then finally wind up teaching the lightning attack in the middle of the duty. This is not appropriate when the duty is already ramped up, and contributes to the "oh gods look everywhere" mindset.

You do not have to show the player each animation tell multiple times the way O10N did, or even show them in a fixed order, but you do have to make sure the introduction period for a mechanic consistently introduces the mechanic. If you want high randomness, have the first one be completely random, the second be completely random out of the two elements not already used, and the third one be the remaining element.

Opening Phase

Aside from the animation tells, this is simple enough. Really, it is almost entirely the animation tells, which does little good when the player has been directed poorly on how to learn them.

Skyruin

Nice clear phase transition warning raidwide, with the random element clearly visible at the front of the arena.

Ice Phase

(no castbar) Avalanche

Here the red text usefully tells the player to turn their camera around and check the avalanche, and it does in fact appear every time so the player does not have to worry about constant camera turning. This is fine as is, though I prefer the Extreme version where the player can learn to dodge by the text alone.

Freezing Dust

An excellently marked Deep Freeze, with both the warning marker counting down to the start of the movement period and the Freezing Up debuff counting down to the end of that period.

(no castbar, snowflake AOEs) Arcane Spheres

The shrinking yellow underline to the red text lightly indicates when the AOEs are about to go off. I can work with that, though red-green colorblind players may well have a bad time.

However, the safespots for these are relatively small, and there is no geometry on this floor for the player to work out where the edges of the safespots are. Trying to reactively correct one's positioning based on the brief telegraphs is not fun. To support mechanics like this, give the player some floor geometry to visually trace out the snowflakes. This is perfectly possible on a snow / ice floor, as demonstrated by Eden's Verse: Refulgence (E8N) and its Savage version (E8S).

This becomes even worse after the fire phase, when these can be combined with avalanches or Freezing Dust.

Thunder Phase

Thunderous Breath

I like this. It is a logical empowerment of the lightning beak, and the telegraph clues the player in right away that there must be some tool other than getting out of the orange available to them. This gets them to try the levitating sections of the floor, making sure they know the usage for later mechanics.

Hail of Feathers

Here - unlike the Extreme - the player does not know the pattern that the feathers will come in, so the cast itself winds up as a multi-hit raidwide that places the feathers to avoid later. The feathers could stand to have more contrast against the floor, though their large size does help.

Blighted Bolt

I like the concept of this. It shows the player that levitating is not always desirable, complete with telegraphs that disappear when leaving the levitating sections. The red text also nudges players to think about how they might counter lightning bolts, and "be grounded" makes logical sense.

However, there are two problems with the execution. First, the player can get hit even if they do it correctly, by another player not quite making it and having their bolt hit the nearby otherwise-safe area. Second, the feathers from earlier have point-blank AOEs at the same time as this cast finishes, and the lightning strikes that are meant to clue the player in to that happen while the player is likely looking under their character's feet for "do not levitate". This again discourages the player from seeing what they are intended to see, particularly since none of the feathers levitate so the logic for feathers is different than for player characters.

(no castbar, line AOEs) Arcane Circles

These are the same circles as the snowflakes in the ice phase, and these ones work well. The red text tells the player to look out, the tethers from the boss tell the player where, and the levitating / not levitating floor sections provide sufficient geometry to at least locate one edge of the safespot. It might not be the edge the player wants for optimization, but the player will never have to react to the brief telegraphs.

These continue to work well after the fire phase, when they are combined with Thunderous Breath.

Add Phase

Ruinfall

The stack tankbuster tower is very clear despite combining three markers, and is a cool concept that encourages the tanks to use Arm's Length on the knockback immediately afterwards

(no castbar) Repeating Line Stack + Cone AOEs

Duty Finder does not make mechanic-specific plans. Even Duty Support will not give the player a clear plan of where the NPCs are going to dodge the ground AOEs. Dodging the cone AOEs is not interesting, and trying to get the full party to stay together to take the stack is a fiasco.

If the player character has the stack with Duty Support, they have a coin flip of blowing their one allowed KO right here by not dodging the same way as the NPCs. And if they already used it on learning mechanics, they are sent back to the beginning for something that is not really a mistake.

Fire Phase

Tulidisaster

An appropriately big, long, scary cast, with multiple hits that go through the different elements of previous phases before starting the fire phase with the permanent DOT.

Permanent / long-lasting DOTs as a different style of heal check are interesting, and this is an appropriate introduction that is gentle enough for casual players.

Eruption

These puddles should be significantly more opaque and red. A stack full of them is visible enough, but the party may not be stacked - especially after the first dodge - and just one is not cleanly visible against the reddened floor.

Overall Impression

Please for the love of Hydaelyn, if you are going to do animation tells, then make sure the player can actually learn the tells and know when to look for them! The fun ideas like branching phase order and permanent DOTs are ruined by the constant stress of trying to look everywhere.

--------

The Skydeep Cenote (Level 95 Dungeon)

https://i.imgur.com/MXRKnOY.png

Mobs Before First Boss

Four Packs

Nothing mechanically interesting here, but can be double pulled.

First Boss - Feather Ray

https://i.imgur.com/aXBVeOk.png

Troublesome Tail

The cast, Nuisance debuff effect around each player character, and "You are cursed to be a nuisance to your party..." text provide an appropriate warning that the player should look out for what mechanic the boss is putting on their character.

Worrisome Wave

A narrow front cone AOE, followed after several seconds by each player character repeating it. The overhead marker provides an excellent cue that something is about to happen, having seen Troublesome Tail the player has a good guess as to what, and from there they have enough time to do okay even though they are inherently reacting to their party members' movements.

Hydro Ring

A donut that leaves its area covered in water, but really is the setup for the next mechanic.

Blowing Bubbles

Starts a bullet hell sequence signified by "Feather Ray produces copious bubbles...". There are four sets of bubbles, one pointed at each player character, so the player is expected to move their character at least once.

This is not my type of mechanic, but in this form it is tolerable, since the player has lead time to see where their party members were and therefore where the bubbles are aimed.

Rolling Current

Creates an array of large bubbles that stay still on the arena before being pushed by a knockback, with enough time to see the array. The first time, the player does not know if the knockback will hit their character or the bubbles, but they can understand it once they see the knockback resolve. When this repeats, the player has the knowledge to succeed.

Hydro Ring

Effectively shrinks the arena again, setting up for more bullet hell.

Troublesome Tail

Here is where the real problem begins.

Trouble Bubbles

While the boss does not directly emit bubbles at the party here, each party member emits bubbles at every other party member, starting as soon as the "Feather Ray produces copious bubbles..." text appears. This means that the player must process bubbles coming at their character from three different directions, filter out the bubbles their own character is emitting, try to find a nearby open space, and repeat because the next bubbles will be aimed at their new location - all reactively! The result is a reaction sequence from visually overwhelming input of fast-moving bubbles.

Oh, and individual KOs from failing this come very quickly - even if the player did the first Rolling Current successfully, two bubbles are enough to KO a non-tank from full HP. A player can easily have their character KOed before the "Feather Ray produces copious bubbles..." text disappears, regardless of their physical reaction time! (Remember that this is the first time the player has seen the Trouble Bubbles castbar. They have reason to expect bullet hell and a mechanic repeated by the player characters, but they also have reason to expect a different pattern for a different castbar and an overhead warning before the repeated mechanic begins.)

To make matters worse, the fact that these bubbles move so fast also means that if the player has even moderate client-server ping, each bubble travels a significant distance in the round trip between the server sending a packet with that bubble's position to the player's client and the player's client sending back the player character's position to the server. (Remember, some regions span very large geographical areas and older Internet infrastructures, even before getting into high ping players connecting to a Data Center whose region does not match their current residence.) As a result, the positions the player sees on their screen are substantially different from the true state of the battle on the server. Not only does this give them less time to dodge than is visually apparent, the form of the discrepancy often results in the player seeing that the server says their character was hit (based on the damage and Vulnerability Up) while the positions they see show their character sidestepping that same bubble. A perfect recipe for the player shouting "I was out of that!"

Of course, the same client-server delays apply in principle to every mechanic. This is something that raiders learn to work around when studying moving object mechanics in high-end duties, and even casual players learn with practice to intuitively account for their ping when judging the end of a castbar. But the speed and targeting of these bubbles creates a visual state that looks unfair in an already reactive mechanic, readily apparent to the casual player, leading to annoyance at both the mechanic and the broken immersion.

If the bubbles were in a very simple pattern fixed relative to the player character emitting them, like "always behind the character", that would be acceptable in this context. More complex fixed patterns like "every intercardinal of the character" would be acceptable in Criterion, where the party can plan around the pattern. As it stands, this is a pile of unfun stress, since failures will likely come from visual overload and/or reaction time.

Overall Impression (Boss)

If the player wanted a bullet hell game, they would be playing a bullet hell game instead of this one. Reacting to where fellow players emit their bubbles from is annoying, not fun!

Mobs Between First and Second Boss

Two Packs

Nothing mechanically interesting here, but can be double pulled.

Skydeep Quarriers into Skydeep Vessels

Now this is a reason to force a single pull on occasion - the Speaking Stone makes it into a forced chain pull (next pack added before the mobs already pulled die) and therefore a cooldown and rotation management test.

Second Boss - Firearms

https://i.imgur.com/zAIw0zJ.png

Mirror Maneuver

A short cast, but a good opportunity for the player to see - or at least be prompted to look for - the mirrors that complicate the next mechanic.

Thunderlight Burst

This is a fun mechanic. The dangerous AOE is massive, with the same radius as a side of the arena, and the player has to figure out where the AOE will explode from by following the boss's aim through the mirrors to an orb. But the long, clear cast makes this very reasonable for the player to do, or understand what happened if they mess up the first time. Later repeats add more mirror bounces, but remain reasonable.

Ancient Artillery

Sets up where the AOEs expand from for the next mechanic.

Emergent Artillery

Another fun mechanic. Combines cross AOEs from the beginning, introducing it with two AOEs (four safespots) the first time and using three AOEs (one safespot) thereafter. The iterations with three AOEs also have spreads afterwards, and the timing is very nice for a gentle combination - the spreads appear near the end of the cross AOE expansion and resolve after the cross AOEs, so the player can naturally move their character to an edge of the preceding safespot and then out.

While on the simple side at first, it fits with the large arena making the player leave more time for their character's movement, and it is still fun.

Overall Impression (Boss)

Good! More like this for dungeon bosses. The player might fail a mechanic, or even multiple, but they have all the information to understand why they got hit. And while it is simple, the large arena means that players who want to tackle movement optimization will.

Mobs Between Second and Third Boss

Skydeep Packers

I understand the reasoning for the barrier, and even the sharp turn to reveal the Alexandrian section of the dungeon, but why is this a single pull? Toss in a barrier guardian or something.

Two Packs

Not much mechanically interesting here, but can be double pulled. The Stonemasons having higher HP is a nice opportunity for skill expression without affecting the experience of weaker players, as it allows stronger players to deliberately target the Stonemasons with single-target cooldowns and falloff AOEs.

Third Boss - Maulskull

https://i.imgur.com/gHEuY7P.png

Stonecarver

A "punch" AOE sequence with very good teaching, between the castbar, "Maulskull draws power into its fists!" red text, and audio cue with glow pulse for each AOE locking in. Since the telegraphs are not shown on the floor, this is technically an old form of memory mechanic, but it is completely fine since even the first-time player only has to remember one {step through / stay} for the second hit.

The teaching is important to set the foundation for when this comes back in combinations.

Skullcrush

The player should be well familiar with knockback circles, but this may be their first time seeing the front-center placement in a relatively small arena. This gives the player the chance to think about aiming to ride the knockback towards a back corner of the arena.

The follow-up spreads are fine, giving the player limited directions to spread out in and a chance to practice that before later mechanics.

Maulwork

A faster, dramatic sequence of AOEs, which is fine even though it feels fast.

Deep Thunder

Big, scary multi-hit stack tower. More of a spectacle than a heal check, but regular duties cannot have difficult heal checks anyway.

Ringing Blows

Combines Stonecarver and Skullcrush, requiring the player to move fast coming out of a knockback. This is a great way to create excitement in a casual-friendly way, since it is a fast and flashy movement sequence but the player has already seen all the pieces.

Colossal Impact

A knockback from a back corner of the arena, followed by either a stack or a spread. Occurs several times.

My one problem with this is that in the spread variant, the player's view of their character can wind up blocked by the boss. (Pointing the camera at a knockback circle can help the player aim the knockback at their desired angle, which after the knockback leaves the camera off to the front and side of the arena. A top-down view has the same problem, since the boss is so large and looms over the arena.) The player should be fighting the boss, not their interface.

If the knockback was from a front corner of the arena instead, the player's view would be much less likely to be blocked by the boss. As a bonus, this would provide a more natural opportunity for the player to think about job optimization, moving into melee range (or staying out when playing a ranged job during the spreads) after the knockback. (As it stands, the player can try to optimize on a melee DPS or tank by moving out later when not using Arm's Length, but that is unintuitive since it leaves them less time to aim.)

Overall Impression (Boss)

Almost good throughout, except for the moments of fighting the interface. Veteran raiders get the bonus of recognizing many of the mechanics from Eden's Gate: Sepulture (Savage) (E4S), though of course the iconic E4S mechanics are not suitable for Duty Finder and the short format of a dungeon boss.

Overall Impression (Dungeon)

The only major issue here is the first boss, but Trouble Bubbles really sours the whole dungeon. The rest could use some polish, but is still decently fun.

----------------

Due to the 20 image limit for a single post, I must split this subsection into two posts. It continues below.

Due to the 20 image limit for a single post, this is a continuation of the same subsection as above, on duties within the Patch 7.0 MSQ.

----------------

Vanguard (Level 97 Dungeon)

https://i.imgur.com/ZV58QV6.png

Mobs Before First Boss

Two Packs

Nothing mechanically interesting here, but can be double pulled.

Vanguard Sentries

Not only can these be double pulled, but their opening AOEs add some mechanical interest and change up how the tank approaches the first pack. (The mobs notice the tank from a long distance, but will generally wind up behind the party from their telegraphed dash.) Nice to see things like this.

First Boss - Vanguard Commander R8

https://i.imgur.com/VKkmBq2.png

Enhanced Mobility

A spatial reasoning mechanic right away, with the "Beware the wingblade!" text to warn the player that there is an aspect beyond the telegraphed AOE. Great work.

Each execution of the bike dash creates a brief forced downtime. Having some forced downtime is totally fine, and the timings are very clear so strong players can plan around the downtimes.

Dispatch

Summons bike adds that do their own basic mechanics, the first set doing the dashes the player saw in the trash and the second doing large circle AOEs that expand from the initial telegraph. While both types have the same castbar, the former adds stay on the ground and the latter fly up into the air, making them visually distinct early on.

Later iterations have two sets of adds, with either simultaneous dashes (perpendicular to each other) or sequenced circle AOEs. Big danger zones, fair timings, especially since the circle sequence is corner to center rather than corner to corner.

Electrosurge

Spreads that go off at the same time as the simultaneous dashes, telling each player to find a different safespot for the dashes. There is not a lot of time with the spread markers visible, but that is fine since the player should be well into working out the dash safespots before the spread markers appear.

Overall Impression (Boss)

A very fair, spatially oriented boss. (It is on the slow side for me, but that is probably a boon to those who struggle with spatial reasoning, and this is a MSQ dungeon.) Not every boss should be this focused on spatial reasoning, but this is a solid one with that focus. It definitely has a place in the design style I want.

Mobs Between First and Second Boss

Two Packs

Gives the option to double pull, with the second pack having a ranged sentry that good tanks will bring into the group. Its moderately large Spread Shot AOE draws attention to its presence.

Two More Packs

Again can be double pulled. The pair of Vanguard Aerostats is very reminiscent of The Twinning, sharing the same trash model and donut AOEs as a pair there, which helps sell the high-tech theme for players who remember that. Like in The Twinning, this is a fun AOE type to use sometimes in trash, since the melee and casting uptime strategies are completely different.

The turret Electrobeams are less good, because unlike the earlier dashes there is no path to walk through. The player must either be at the front of the party, wait, or take the AOE, and the latter two options are not fun or interesting. If they fired in sequence or had spaces between them, they would be fun.

Second Boss - Protector

https://i.imgur.com/2N7Jlyi.png

Search and Destroy

Circle and line AOEs under every player character in every round, creating a visually dense bullet hell effect reminiscent of Another Mount Rokkon. This is the annoying kind of stressful, since it forces the player to rapidly react to the AOE placements and constantly move rather than doing anything interesting. (The later combination with Fulminous Fence does not help, despite Fulminous Fence being a good mechanic, since the underlying requirement is still repeated reaction.)

This boss could have player-baited AOEs on theme in a much more interesting way, fitting both the name of the cast and the idea of the boss using turrets. Since this boss already uses Acceleration Bomb, it could start by placing Acceleration Bomb on every player character (with the warning it gives now) and then drop a single slower circle AOE under every player character as the Acceleration Bomb is about to expire - creating a test of sticking to the plan of not moving and only then acting on the AOE, rather than running away in reaction. Then it could progress to planning various simple baits around repeated Acceleration Bombs.

Fulminous Fence

Creates lightning fences that inflict Paralysis when moved through. This is a neat way to change up the movement for later AOEs from the player's normal patterns, though it cannot be used too often by necessity.

Battery Circuit

The spinning cones and periodic circle AOEs are made interesting by the presence of the lightning fences. If this was the only movement-heavy mechanic on this boss, it would feel good to plan as a place to use movement tools efficiently. But as it stands, movement tools for jobs with frequent cast times have probably just been used on Search and Destroy, which furthers the annoyance with Search and Destroy.

Motion Sensor

This is a great warning for Acceleration Bomb, between the cast name and the red text "A motion-sensitive bomb has been affixed to you." The player has a fair shot at understanding the debuff without reading its text, even if they have never done non-MSQ duties to see the concept regularly before.

Blast Cannon (cast by turret adds)

A simple sequence of line AOEs covering the arena, that is complicated by the Acceleration Bombs. Very fair, great mechanic.

Heavy Blast Cannon

This standard line stack functions as a spacer mechanic, leading directly into a raidwide so the healer has to work a little. Nothing too hard, but good.

Overall Impression (Boss)

Why the spam of AOEs under the entire party? This could have been a great, really fun boss if it focused on using different movements than normal around the Fulminous Fences and Acceleration Bombs instead.

Mobs Between Second and Third Boss

Two Packs

Nothing mechanically interesting here, but can be double pulled.

Two More Packs

Again can be double pulled, and the first pack of this pair adds some mechanical interest by opening with a mix of line and donut AOEs.

Third Boss - Zander the Snakeskinner

https://i.imgur.com/m0KEM0H.png

Soulbane Saber

A line AOE directly in front of the boss that then explodes into a half-room cleave hitting the whole area that was in front of the boss during the cast.

After showing this once and having a tankbuster where the player can think about what they saw, the boss launches into repeating it while rotating in the middle of the arena - a smooth combination with itself, as the first AOE of the current cast goes off at about the same time as the second AOE of the preceding cast. This is also a fair combination, even once it ramps up to add spreads, since the player has had the chance to see all the pieces first.

This mechanic is a great choice to work with the story constraint of a dungeon boss needing to die and then revive stronger without lengthening the dungeon enough that players try to avoid it. Since it combines well with itself, it allows the boss to show a combination within a very short phase, rather than needing to do only simple attacks or else go too fast to give the player a fair shot at learning. With some concepts it is better to have such a short phase be all introduction in order to use more mechanics later, and with other concepts this way is better.

(no castbar) Revive Sequence

This is short enough that I do not mind it, though I am sure some players got annoyed with it while grinding the dungeon for glamour.

Mechanically, this serves as an opportunity for the healer to heal up from the repeated stacks, a forced downtime that the player can work with rotationally regardless of their job, and an opportunity for the first-time player to think about what mechanics might come from the snake transformation. (Even though they do not know what a feral soul is yet if they are going through MSQ unspoiled, they can still see the changed boss model and the dialogue "I call on the soul of the serpent, the great devourer.")

Syntheslean

A simple 90 degree cone in front of the boss after it turns, but it helps the player pick up on Syntheslither at full speed.

Syntheslither

The indicator showing the boss's attack path and four swords for four attacks is a good one, but alone it would not be enough for the speed of this sequence. (If the player is not looking directly at one of the AOEs, they will probably miss the shape since it appears so briefly.) Showing the player the Syntheslean cone first makes it much more likely that the player will pick up on what attack is being repeated in a reasonable time, allowing this attack to be fast and snake-like while remaining fair.

Slitherbane Foreguard / Rearguard

An upgrade to Soulbane Saber, having the half-room cleave go off immediately and be either at the front or the back of the boss depending on the cast name. The fact that this upgrade also combines smoothly with itself is great, mechanically selling the boss's power-up.

Overall Impression (Boss)

This is a fun boss, and an excellent way to work with a very difficult story concept. Dungeon bosses are time-crunched even before splitting the boss into before and after its first death!

Overall Impression (Dungeon)

This would be an amazing dungeon if not for Search and Destroy. Everything else in here is very well done.

--------

Origenics (Level 99 Dungeon)

https://i.imgur.com/O1cj7Jo.png

Mobs Before First Boss

Two Packs

Nothing mechanically interesting here, but can be double pulled.

One Pack

A dull forced single pull, followed by an open section for story reasons. The story reasons are all well and good, but you could have found some way for mobs to be there and be cleared out.

First Boss - Herpekaris

https://i.imgur.com/VgqcmAs.png

Vasoconstrictor

Sets up venom puddles that the boss then punches.

Venomspill (repeated many times)

The fact that the boss lifts its arm and punches a side is fair enough. The player has seen lots of enemies with the same arm animation before, and it is intuitively "hits the side with the raised arm" regardless, so the only real question they should have is how far the arm punch extends.

But the way this mechanic actually works is not only reactive, it is actively working against the player's desire to learn and improve.

The first problem here is that where the attack actually hits is a random selection of puddles that can cover almost any area except the edge of the arena furthest from the lifted arm, but generally does not cover that whole area. So at the most basic level, the player will look at the first set of telegraphs to learn where this cast hits, and will reliably learn the wrong thing from that first set. At higher skill levels, they are denied the ability to dodge just outside the AOEs (and thereby enjoy themselves) because the AOEs are not in a consistent place every time for them to learn. The player is left with no recourse but to run from one end of the arena to the other or risk getting their character hit.

The second problem is that the player cannot even start moving their character preemptively, because the repeated Venomspills are random rather than alternating. Even for a player who understands prepositioning and begins moving their character back to the center every time, it is still quite reactive, and the casual player cannot be relied on to have that knowledge!

(Plus sustaining a movement input to preposition partly pins down the player's hand (or other body part used for input), which makes pivoting to run the other way take slightly more time unless the player splits their movement inputs between both hands. Whether such a hand split is even possible using a controller is irrelevant, since a controller cannot split the character movement joystick. With a mouse and keyboard it is possible, but counterintuitive: on Legacy movement the player must face their camera sideways to the boss to make the mouse forward movement useful in the split, and on Standard movement there is the added complication of turning the player character around to face the camera for the "keyboard direction". I am not convinced that the Standard movement split method is actually an improvement, since turning the player character is added inputs, and even the Legacy movement split method is absurd to expect here. Not only does it make assumptions about the player's input scheme, but this is a leveling MSQ dungeon boss and the struggling player will generally not think of that.)

(I do in fact use the Legacy movement split method in some duties, because it is slightly faster and I naturally prefer Legacy movement. However, I need to know the duty to know to do it in the first place - which is completely not the context here even before the anti-learning nature of this mechanic.)

This is the core mechanic of the boss, and it is awful. (Even before being combined with spreads.) It could have instead been something interesting like "hit one side with the edge of the AOE right down the middle, which flings the venom to consistently cover the other side and little more in a fixed pattern". That would have been a reasonable two-step dodge.

Writhing Riot

An acceptable memory mechanic, but nothing special. The player can stand in the first safespot and then either remember both subsequent hits or remember the second hit and use it to work out the shared safespot with the third hit. This puts it under the three item rule, and allows strong players to plan to dodge in a single movement.

Overall Impression (Boss)

The core mechanic of this boss is so intensely anti-learning, and little else is in here.

Mobs Between First and Second Boss

Two Packs

Nothing mechanically interesting here, but can be double pulled.

Two More Packs

These are mechanically interesting as a double pull, combining the Aerostats (with donut AOEs) and a ranged enemy for the tank to bring into the group.

Second Boss - Deceiver

https://i.imgur.com/s7lMwpL.png

Bionic Thrash

Hits two quadrants of the arena opposite from each other, shown by the arms. Nothing special in itself, but returns later in combination with the boss's core concept.

Initialize Androids

A functional warning cast, telling the player to look for the adds at the front of the arena.

(no castbar) Line AOEs From Androids

The concept, that some of the adds are not fully spawned in and therefore do not attack, is a good one. The telegraphs covering the whole floor prompt the player that something must be going on to create at least one safespot.

However, the boss dialogue "Active camouflage operational." makes it seem like the adds that do not fully appear are the ones that will attack, since the other adds appear normally and cannot really be considered "camouflaged" in any way. The player will learn after the first hit, but it is annoying to be misled by the text when instead the text could have said something that is better than guessing without it.

The visual effect for "this is not fully spawned in" is also not really adequate for the player to rapidly scan rows of adds. If they happen to be looking when the "flicker" effect is not active, all that distinguishes the adds is the incompletely spawned ones having a vague green aura, which does not stand out in the bloom-heavy environment.

The adds then come in to be killed, which is good to have sometimes for the same reasons as previous bosses.

Initialize Turrets

Another functional warning cast, this time for large blocks on both sides of the arena that will fire larger line AOEs. I do not like calling these "turrets", since that has been long associated with smaller enemies and line AOEs, but there is nothing strongly obvious to call them. If I had to pick something for the castbar without help, I would call them "cannons" or something similar that at least does not point the player away from that AOE size.

(no castbar) Line AOEs From Blocks

The blocks fire one side after the other, creating a sequence of AOEs to dodge where the whole arena is covered over the course of the sequence. The player can see which side will fire first, as the telegraphs roll out from that side to cover the other side.

However, the visual clarity of which blocks will fire is worse than for the androids, as all of them have particle effects and substantial glow in their charging animation, which mimics the "not fully spawned in" visual effect.

Surge

These spreads with line knockback are good for the player to utilize the AOEs they just saw (or look back at the blocks), since this time the fully spawned blocks prevent their character from being knocked into the wall. Fun follow-up to the sequence.

Bionic Thrash + (no castbar) Line AOEs From Androids

A solid overlap. This time there are fewer "not fully spawned" androids, which is fine in concept since the player should know which type to look for.

Overall Impression (Boss)

Good concept, and well executed except for the introductions and visual clarity not being the best. This would be a good boss given a better presentation.

Mobs Between Second and Third Boss

One Pack

A dull forced single pull.

Two Sentries

Another dull forced single pull, except that the Sentries burst out of the door as soon as it opens rather than letting the tank pull. This does not change much, as they are not strong enough to create the resource-draining effect of the Holminster Switch forced pull series and the party has downtime while they run to the next pack.

Two Eyeborgs

These have multiple telegraphed mechanics, reminiscent of Coincounter in Aurum Vale, and can be double pulled with the next mob / miniboss. Pulling them together is fun, having fewer enemies but more complex AOEs than a normal pull.

Automatoise

I do not mind the large AOE Hard Stomp from this single mob at all. The player has ample time to dodge, they can strategize regarding range, and once they see it repeat they can strategize about it. In particular, the party benefits from dodging together, for both healing range on the tank and melee range on the Eyeborgs if double pulling.

Third Boss - Ambrose the Undeparted

https://i.imgur.com/ha3VyhV.png

Overwhelming Charge

The way the boss holds its spear is not initially clear about hitting the front or the side, which is not amazing, but the player will learn that it hits the front from the first hit so it is acceptable.

Psychokinesis (Cages)

Comes with red text "The nearby cages are under strain from psychokinesis!", which is good at prompting the player to look at the sides of the arena with the cages, even if they have forgotten the dungeon after a long break and did not get the cutscene replayed. The castbar is also suitably long.

However, the breaking cage doors could use some more visual pop. They shake, their bars are warped, and they glow slightly, but this is a high-bloom environment and the player has to split their time between checking two opposite sides of the arena. (Having the doors shake is a nice touch to not rely exclusively on contrast for clarity, since peripheral vision is relatively motion-sensitive.) The floor lines for where the cage doors can hit are clear, but the doors themselves are not until the very end when they shake heavily.

(no castbar) Uncaged Adds

A very logical place to have some adds to kill.

Extrasensory Field

The player need only move to one quadrant that does not point off the arena (or use their knockback immunity), for which the tell is clear enough even though the arrows fade out quite strongly during their pulses. Since the player has long enough to look to see an arrow pulse bright several times, they can make multiple eye movements to locate a suitable spot and still have time to move their character.

If they did not notice already, seeing this go off gives them another chance to pick up on the pattern that two opposite quadrants point off the arena to the sides, and the other two point in to the arena to the front and back. (Which is good, as it further distinguishes the quadrants rather than making it a natural guess that the player character would be swept through the midline of the arena and then off it.)

Psychokinetic Charge

Combines Overwhelming Charge and Extrasensory Field, asking the player to use the knockback to dodge the front AOE. A nice fair combination.

Electrolance

A simple circle AOE from the thrown spear, nudging the player towards the boss.

Psychokinesis (Spear Dash)

This is not very clear about how wide the spear will hit, nor does it have easy floor references, but the dash sequence starts away from the boss so the player character is naturally positioned to avoid the first couple of dashes and see the rough size. This is not good for enabling precise dodging, so it gets dull after the first time.

(From here on, the two forms of Psychokinesis and the Psychokinetic Charge repeat with added spreads. None of these mechanics play poorly with spreads, and this is a decent mechanical expression of "the boss is desperate to win" for an MSQ dungeon that a very casual healer must be able to clear unaided.)

Overall Impression (Boss)

An okay boss, nothing special even if the visual clarity was up to par.

Overall Impression (Dungeon)

Many boring sections including a horrible first boss, and nothing particularly special or exciting.

--------

Everkeep (Zoraal Ja, Level 99 Normal Trial)

https://i.imgur.com/EV8NrhF.png

Double-edged Swords

This is a great cast name, cueing the player to expect the second hit even before its telegraph appears. And since the two hits cover all of the arena exactly once, the player must move their character to succeed but is likely to get it the first time based on the castbar.

Calamity's Edge (Activates point-blank AOEs from the Visions)

Nothing special, but fair enough. After all, anything on the battlefield is not likely to be completely harmless.

Vorpal Trail (Dashing Swords)

My first reaction to this was an excited "newbie's first Exasquares!" ("Exasquares" being the colloquial name for the upgraded Cosmo Arrow in the final phase of TOP.) I saw the swords dash at the edges of the arena and come in, and it was cool. Having them start at the edges is a good choice combined with this small boss hitbox, since the player character is likely not near an edge of the arena when it starts, and so the player has a chance to see the pattern before interacting with it.

My second reaction to this was "oh no the floor is not helping". TOP gets away with that, putting a mechanic that operates on a square grid into a circular arena, by both having a very crisp visual effect for each hit and being an Ultimate. This duty has neither of those to make the mechanic work.

Getting hit by this once would be fine. What makes the lack of floor references actively annoying is that this repeatedly combines with Double-edged Swords, and the player has neither a timing reference nor a position reference for this mechanic to plan their movement for the combination. (Especially since the first section of this trial has very few mechanics in total.)

Since this is meant to be a precursor to the swords from the distorted worlds in the rest of the trial and the Extreme - where the floor does have appropriate lines - I would suggest having the floor shimmer and bring in that reference grid beforehand, both providing the mechanical reference and representing Zoraal Ja's twisted thinking. That would make the mechanic less frustrating and so allow the excitement of the combination to shine through.

(There are no further notable mechanics until the cutscene and phase transition.)

Smiting Circuit

The red text "Zoraal Ja readies his blades!" is a good cue to look into the arena for the environment tell. My only frustration here is that the chosen glow effect for the large swords is very similar to that of the "death wall" at the edges of the arena, making it needlessly difficult to pick them out of the environment. (The player does not have the boss's stance to help them, as Zoraal Ja is not holding the swords for the point-blank / donut, and the glow effect distorts the shape of the swords.) In the Extreme, the player learns to overcome that through duty knowledge and practice, but that is not useful here in the Normal.

Since the glow effect is thematically showing that Zoraal Ja is burning through souls, I would instead change the effect on the death wall. It could have been in glowing electrope purple like that in the background of the instance, with a brighter section that rises up to create the sharp wall contrast, and that would maintain the sharpness of the death wall while making the swords visually clear.

Dawn of an Age

This is a really nice way to shrink and transform the arena, allowing for the "distorted worlds" mini-phases. The AOE is itself simple enough and shown for a long enough time that the player will almost always dodge it successfully, allowing for the covered parts of the arena to disappear. Great work.

Vollok

Putting the swords down directly on the available platform the first time is good, priming the player to look further afield for the swords when this cast repeats. The first cast also shows the player that the swords hit on the grid.

Sync

A nice long cast paired with the red text "The distorted world comes into synchrony." This very fairly asks the player to look for which additional platform(s) have been activated and will perform their AOEs shortly after the end of the cast.

If this did not need a delayed activation to allow for an overlapping cast in the Extreme, for consistency and clarity I would keep the timing and lengthen the castbar so that the AOEs hit at the end of the cast.

Bitter Reaping

Double tankbusters cannot hit that hard in Normal, but it is nice to have one so that casual players remain familiar with the concept.

Gateway

Sets up the tethers that move dashing swords between the platforms. Giving each piece of this setup a separate cast is good for encouraging the player to see all the pieces in the order they appear.

Blade Warp

Another setup cast, placing the dashing swords. Again, this is good, and allows a relatively complex mechanic to work fairly in Normal.

Forged Track

The combination of the cast, the tethers to the swords being activated, and the red text "Blades go soaring through dimensions." really makes this work. The dashing swords need the tethers since they are so far away from the player character, while the earlier planted swords do not. (Indeed, the planted swords are too numerous to tether without clutter.)

Actualize

Transforms the arena back, allowing the fight to switch back and forth between mini-phases.

Half Full

This cleave is significantly easier for the first-time player to find than Smiting Circuit, because they can follow the boss's raised arm to the sword. This allows the cleave to work with the blue background tint of the "distorted worlds" arena later on.

Half Circuit

Logically combines Half Full and Smiting Circuit, which would be very reasonable with the changed death wall and background.

Duty's Edge

The big multi-hit line stack is scary, but it hits slowly enough to be reasonable in Normal.

Overall Impression

The first section (before the cutscene) weighs this trial down, between the repetitiveness overall and the positioning frustration of Vorpal Trail. Everything after the cutscene would be a sound, fair Normal given a touch of visual polish. Working with a diagonally oriented boss is a nice change of pace without being unapproachable.

--------

Alexandria (Level 100 Dungeon)

https://i.imgur.com/PWXpsV4.png

It is unfortunately necessary context that many players are not going to want to be here, due to moral disagreements with the MSQ's requirements of their character. If they stick around for endgame, they will likely see this again in Expert Roulette, but the first-time experience is still important.

Mobs Before First Boss

One Pack

A dull forced single pull. This was annoying enough in leveling dungeons, and is outright egregious in endgame dungeons that will appear in Expert Roulette.

You want the player to move through the Alexandrian environments in a very staggered way, fine. Make the single pulls mechanically interesting. Give us multiple AOEs, or minibosses that interact with the environments, or something.

Two More Packs

Nothing mechanically interesting here, but can be double pulled.

Immunization Protocol B9

This is literally one mob, a robot elephant that does nothing interesting, and the party has no option to fight it alongside other enemies. (It does spawn directly after finishing the last pack, but that is hardly resource depletion when it does little damage.) Boring!

First Boss - Antivirus X

https://i.imgur.com/Wuyhtms.png

Immune Response

Cleaves either the front or the sides of the boss, with telegraph and glowing front beam / side tentacles. Simple on its own, interesting when later combined with the sequence of donut and cross AOEs.

(no castbar) Five AOE Sequence

This mechanic has good cues to what is happening, between the red text "Foreign entity removal initiated.", the circles and pluses appearing in green one at a time over the five purple dots that start the mechanic as the lightning flows through them, and the telegraphs and audio cue as they go off. The pace is also good, giving the player time to plan out their movements.

What would make this mechanic really great is adding a visual indicator for how long each AOE has until it goes off. (A falling object over the circles would look too much like a tower, but a second color could fill in the circles and pluses one piece at a time, or something like that.) That would help the player time their movement around the cues and their Global Cooldown, rather than partly reacting to when each AOE goes off (until and unless they learn the snapshots from sheer practice). Especially with the overlapping Immune Response casts, this mechanic could have lots of room for skill expression and satisfying resolution - it just does not have the indicator to enable that.

This is the core concept of the boss, repeated often, and it would go from decent to great with that visual change.

Quarantine

An AOE tankbuster and a stack for everyone else. Not hard, but it is different, pushing the tank to stay out of the stack.

Overall Impression (Boss)

A good boss that could be a great boss, just needs that visual representation of the sequence timing to enable and encourage mastery. When I was checking a video of this, the party in the video took about three party KOs on this boss, and they kept going because they understood their mistakes and saw improvement each time. That is what encouraging failure looks like.

Mobs Between First and Second Boss

Two Packs

Nothing mechanically interesting here, but can be double pulled.

One More Pack

A dull forced single pull with a short downtime before the next pack spawns.

One More Pack

Still not interesting, and a series of only two packs is not long enough to create resource depletion the way Holminster Switch does - especially in Expert Roulette, where despite the tighter Item Level sync, once the player gets their gear going they will come in with 30 Item Levels above the minimum.

To make matters worse, after this pull, the party spends a long time waiting at the pile of wood and stone rubble before being able to reach the second boss. While the dialogue there is story-relevant, in repeat completions the placement is boring, and it negates any resource depletion by giving time for cooldowns to come back. This being Expert Roulette, that is a lot of boredom and frustration.

Instead, I would present the opportunity for a standard double pull and have the storytelling sequence start as the last pack is engaged, playing out during the party's fight against that pack. The rubble would then be shot down just after the last mob dies, allowing the party to move on without waiting. Even if the player has returned to the game for the Dawntrail MSQ, the five preceding leveling dungeons will have refreshed their familiarity with a standard dungeon pull, and the extant voiced dialogue of the sequence would allow them to take it in without diverting their eyes from the pull at all.

Second Boss - Amalgam

https://i.imgur.com/c2qcAQk.png

Disassembly

The boss splits itself into many pieces, matching the name, and slowly creates a triangle enclosing half the arena before hitting that triangle with an AOE. This whole battle is purple on purple, and only the edges of the danger zone are marked until the "there was another indicator for this" telegraph at the very end, but this still works: it has a long lead time, the red text "Amalgam divides into parts.", and the boss model changes dramatically.

Centralized Current / Split Current

Casts with growing purple glows, hitting the front-to-back line of the boss and the sides of the boss respectively. The veteran player may recognize this pairing as Hot Tail / Hot Wing.

While this mechanic is fine, the credit for that mostly goes to the castbar. The glowing areas initially do not give any clue that the AOEs will hit beyond the bright section, and the eventual front-back line on each is too thin to be reliably seen even aside from its lateness. If you want the player to have a good chance to guess this pattern from the visual, rather than relying on the castbar and perhaps getting hit once, the visual needs to be better tied to the area being hit.

(no castbar) Electrope Beams

There are five spaces for circular cannons to drop down at the front of the arena, three of which are filled in the first iteration and four in subsequent iterations. These perform simple line AOEs that combine with Disassembly.

The combination is fine, but very much relies on the cannons being at the front of the arena (thus likely to be on screen already), the lead time of Disassembly, and the red text "Energy accumulates in the electrope!" to be viable. The thin lines extending from the cannons are not suitable to draw the player's eye on their own.

These are followed up by spreads that appear just before the combination goes off, prompting the player to spread out of the combination safespot. The timing is acceptable, but significantly more likely to get the player caught in an adjustment loop than the similar setup from Firearms in The Skydeep Cenote.

Ternary Charge

Easy but fun, just like Shore Shaker from the Prime Punutiy in Ihuykatumu. This is the same sequence of concentric circle AOEs.

Overall Impression (Boss)

An okay boss, nothing special despite the potential in the theme of a boss that splits and recombines. Scraping by visually; this is about the worst viable visual design before it gets annoying.

Mobs Between Second and Third Boss

Four Packs

Nothing mechanically interesting here, but can be double pulled.

Third Boss - Eliminator

https://i.imgur.com/5KaDS0S.png

Partition

A simple side cleave that sets up for Reconfigured Partition.

Reconfigured Partition

The cast name is an important clue that the sword will move to and hit the other side, making this a fair mechanic instead of a reaction to the animation.

Subroutine

The boss splits one or two parts off, so that those parts can later perform distinct AOEs. (The claw does a line AOE and the orb does a donut AOE.) In theory, this is a fun way to build on and connect to the preceding boss, Amalgam.

However, the initial positions of the parts on the boss's back are so high up that they are not properly visible when the player character is close to the boss at the default camera tilt or even slightly higher, which makes this functionally only a warning that the boss set something up. Since in the story this duty is taking place in a computer, the cast name "Subroutine" could set up just about anything, and that is all the player gets if they do not see the parts separate.

(I literally did not know that this boss splits itself until I went into this dungeon with Duty Support for this review - the comments from Krile and G'raha Tia clued me in. On my first attempt with a friend and Duty Finder matchmaking, I did not see the parts separate due to the combination of camera tilt, camera position, and rapidly moving my eyes trying to learn the duty. On subsequent completions, I had learned to anticipate the potential positions for the part mechanics and look there at their scripted times, which meant I was not looking at the boss to see the parts separate.)

Since it is not practical to expect the player to adjust their camera tilt in combat - APM aside, they may not have the keybind flexibility to do so - and camera tilt is important to the player's HUD Layout and seeing around their character, I would fix this by moving the detachable parts of this boss model down and having their removal significantly change the boss's silhouette. That would make the first-time player much more likely to see the separation without Duty Support commentary, while also not changing the boss's unified color scheme.

(cast by Elimination Claw) Terminate

A wide line AOE cast at the same time as another mechanic from the boss, with the only indicators being the cast on the enmity list and the purple lines extending part-way from the claw at the side of the arena.

If the player reliably saw the claw leave the boss and move to the side of the arena from the preceding Subroutine, this mechanic would be okay but on the fast side for processing the combination. Since that is not reliable, the purple glow is not sufficient to reliably draw the player's eye when combined with the other nearby purple from the Reconfigured Partition when this first appears. And since this does not have a long cast, the result is annoying "gotcha" potential where the player does not see what hit them. (Remember, the player has a lot to see their first time in a duty, especially for casual players.)

(cast by Elimbit) Halo of Destruction

A donut cast at the same time as another mechanic from the boss, with the indicators being the cast on the enmity list, the darkening of the unsafe areas of the arena, and purple vertical lines clamping down on the "spotlight" of the safe zone. The donut is followed up by spreads, or later by a Terminate also asking the player to move their character out of the donut safe zone.

Again, if the player reliably saw the orb detach, this would be okay. Since that is not reliable, the indicators are somewhat lacking, though less badly than Terminate; at least something is shown over essentially the whole arena.

(started by Subroutine detaching both parts) Add Phase

The six spread Lightning Generators have the potential to be interesting, as different jobs can rotationally benefit from multiple targets in different ways. However, in the context of a dungeon boss where an add phase must be very short, the lack of party coordination on who will target which adds first limits the skill expression from these adds in Duty Finder matchmaking.

(cast by Elimbit) Light of Salvation (Line Proteans)

These come with telegraphs, as is necessary to get the player to spread by reacting.

Here, the initial presentation of proteans is fine but not good; the Elimbit has moved after its donut and before the first appearance of this, so between the movement and the adds the player may well not even see the Elimbit and purely react to the telegraphs. (Compare to Ktisis Hyperborea having Hermes cast True Aero, which is also a line protean - and even has a baited follow-up - but comes directly from the boss, which the player is reliably tracking.)

After the add phase, this is sequenced after a knockback circle (no castbar), with the protean telegraphs appearing just before the knockback resolves and leaving several seconds to spread between the knockback and protean resolutions. While that sequence is in itself reasonable, the need to react to fellow players' choices necessarily draws the player's eye away from the Partition / Reconfigured Partition that the boss starts casting during the spreading.

That last part can easily hand the first-time player an individual KO - remember, their character just took damage from the knockback and the protean even if they avoided being knocked into the wall - without giving them the opportunity to see that any avoidable damage was coming at all, let alone what it was. And even on repeat completions, the visual demands are too much for a mechanic that Duty Finder can only do reactively: Partition / Reconfigured Partition require looking at the castbar and the sword, while reacting to fellow players' choices effectively for the proteans requires looking at one's character's feet the whole time. As a result, the most practical method is for the player to try to take their protean close to the front-to-back line of the boss as prepositioning to react to the Partition / Reconfigured Partition, and resolve that on reaction once the proteans resolve. This further pushes reaction within the sequence, making it unfun.

Any mechanic that is going to push the player into reacting to fellow players' choices using vision must be designed around the assumption that it completely occupies the player's eyes for the duration, due to the inherent pressure on reaction time pushing them to keep looking at the relevant locations. Surrounding mechanics must also be chosen and designed on that basis.

Elimination

The boss slashes the arena several times, dealing light damage each time and leaving a set of lines slashing the ground. These lines then expand and explode in the same sequence, requiring multiple movements to avoid further damage as the pattern is quite dense.

If the visuals were clearer about where each set of AOEs will hit, I could enjoy looking for areas that will not be hit often and plan to dodge accordingly. But the visuals given focus more on glow and flash than clarity, and there are no floor references to work with, so this is unpleasant to look at and annoying to analyze.

Overall Impression (Boss)

Would have a much better first-time experience if the player could reliably see the pieces of the boss separating and reattaching. Manageable on repeat completions, but still noticeably annoying.

Overall Impression (Dungeon)

Slow and boring, not allowing the concept of the first boss to shine. When it comes to slowness, the tighter Item Level sync compared to the rest of Expert Roulette at least has a reason to be there - the single pulls and forced wait do not.

--------

The Interphos (Sphene, Level 100 Normal Trial)

https://i.imgur.com/pyYpedR.png

As with Alexandria, it is unfortunately necessary context that many players are not going to want to be here due to moral disagreements with the MSQ's requirements of their character. As this is a story trial, despite the lack of Duty Support the player has little reason to return after their own MSQ completion, so they are statistically quite likely to be here for their MSQ.

As a general visual note, the NPC illustration on the Queen Eternal's battle dialogue popups extends a little too high, obscuring target buffs and debuffs in the default HUD Layout position of the target bar. Rather than expecting even the most casual player to move their target bar in order to track their own applied debuffs, I would move the NPC illustration down - or better yet, allow the position of these popups to be adjusted in HUD Layout.

Legitimate Force

A fine one-two punch sequence, with visuals on the boss's four arms that allow the same side to be hit twice while displaying both parts of the sequence. Clear enough, and the first glow naturally draws the player's attention.

Aethertithe

The initial many-hit raidwide is fine.

The subsequent puddle AOEs and final hit on each section are completely sunk by their visuals. The red text "The absorption of aether accelerates!" is a fine warning and not the issue. The issue is the distorted grid visual for the danger zones. For me, I manage, but it looks like it should be a draw-in mechanic even if not caught by the final hit on each section. However, some players got so motion sick from it that they vomited, and I can understand why; the appearance that the floor is moving in a distorted pattern can throw off visual balance feedback quite strongly.

In theory, using the motion of high-contrast objects is a way to provide a visual cue that will reliably be seen without resorting to glowing or standard orange telegraphs, working with the motion sensitivity and contrast requirements of peripheral vision. In practice, the results of this have made it clear that such technique should not be used to distort a floor - and if you want to experiment with it in other ways, run tests with players who suffer from various forms of motion sickness before putting those visuals in the live game.

Coronation (Circle AOEs)

The way these targeted AOEs leap towards the selected player character is spooky and not helpful, since it reduces the player's lead on the AOE and often does not help the player figure out who is being targeted. (The party is likely stacked from a previous mechanic.)

If this instead came with a Prey targeting marker, like its other version, it would be a reasonable mechanic with an appropriate lead time.

Virtual Shift

Transforms the arena into a different environment.

Hourglass Arena Mini-Phase

A very restricted arena that makes for some fun and approachable challenges.

Coronation (Double-Targeted Line AOE)

This is a really cool mechanic. The two targets can work together and pull the cannon around with their tethers so that no one gets hit by the beam, and they have a long time to see the telegraph and how the cannon responds to them. The crisp visuals are also excellent, helping the player see the pieces of the mechanic regardless of whether their character is tethered.

While Duty Finder does not coordinate, and this mechanic seems to be here because of the Extreme version coming in Patch 7.1, this still functions in Duty Finder due to the long lead time and only taking a single Vulnerability Up for getting hit. The player has long enough that they have a good shot at sorting it out without ever feeling pressured on their reaction time, and if they fail they are still able to continue fighting without even waiting for a raise.

If you want to have some complex mechanics where the player must react to another player's choice, this is how you do it. Such mechanics are still inherently limited to regular duties, but this is how you do it.

(This mechanic also appears in later rounds of the "main arena" mini-phase, combined with Aethertithe.)

Downburst

A simple, familiar knockback from a corner or the center of the arena, complicated by the restrictive arena shape. It is not hard, but it is fun, with enough time for the player to fairly reason about the complication their first time.

(no castbar) Donut AOE from Drone Circle

Nothing special, but a fine mechanic. The visual cues of the drones swirling in to form the circle and then lighting up the edge of the danger zone is what makes this work.

Powerful Gust

Another simple, familiar knockback, this time from a side of the arena. Has exactly the same design results as Downburst.

Levitation Mini-Phase

The way this phase opens by moving the player character to make sure there is no floor under them is an important trick, pushing the player to notice the levitation without having to read their debuff bar. After all, this is Normal Mode. The forcible move is not ideal, especially out of melee range, but doing it during an arena change makes it less jarring.

Castellation

Giving the player ample warning of every height change is what makes this work. The opening showing the levitation and the red text showing the changes with "Gravity increases, making you fall." and "Gravity decreases, making you float." are crucial.

I must also compliment the attention to visual clarity while having the wall still look like part of a castle. Not only do the edges of the holes glow, but some light shines through the holes into the safe zones to help the player line up with them.

Since failing this mechanic is a one-hit KO (the wall sweeps the player character out of the arena), I would lean even further into the visual clarity by making the irrelevant top of the wall less jagged and less glowy. I would also consider making the safe zones a little wider, since this is mandatory for MSQ and very casual players may not have the required movement precision. (Especially for the quicker sequence of two Castellations in a row - the down to up transition ensures that the player character is already close to a safespot, but that does not help a player who lacks the movement precision to land in it with their first movement attempt.)

Legitimate Force (with levitation follow-up)

The same mechanic, but this time the player will fail the impending fall if they keep their character close to the center of the arena.

Now that the player has already worked with the levitation mechanic, this is reasonable, and the timing of showing the red text shortly after the second arm glow is perfect. Since they have already seen Legitimate Force several times and the basic format is already familiar, the red text will appear pretty much exactly when they have likely finished processing the second glow cue, encouraging them to smoothly add "and then move to a platform" to their movement plan if the hits alternate.

The timing of the fall allows for a moment of hesitation, which is necessary since the casual player has no practice in executing movement sequences without hesitating.

Overall, this is a good example of how to set the player up for success with a movement sequence mechanic.

Main Arena Revisited

Absolute Authority

The sheer length of this castbar, the dramatic dialogue, and the 5-second countdown are normally associated with very high damage mechanics, making the player likely to shield and/or tank Limit Break this... which is largely useless, since the cast does only moderate damage right away and instead applies debuffs for a mechanically intense sequence.

Duty Finder does need a breather before and after dealing with all of this - there are flares, "stand with someone" triangles, Acceleration Bombs, and petrifying gazes attached to player characters in quick succession, combined with rounds of circle AOEs on the ground and repeated pulses of damage throughout the sequence. However, this presentation is frustratingly misleading, as the party does need their healing and mitigation tools but not for the opening of the mechanic.

To improve this mechanic, I would split the presentation into two castbars. The first would apply an "inactive form" of all the debuffs and faded "frozen" markers for the flares, showing even the casual player that they should prepare for something to happen later. The second castbar would still be long to give the player time to think, and at the end would "activate" all the debuffs and start the sequence. In this way, even if the player does not read debuff bars, they would spend the long cast preparing for a sequence involving later healing rather than expecting a massive raidwide.

(Yes, my proposal would be gently introducing Duty Finder to the same "this debuff is not active yet" concept that underlies the Code Smells in TOP, but with a later activating cast rather than an expiring timer. It is a good concept for all skill levels, and really does not need to be limited to Ultimate.)

Once the player actually knows what this mechanic is, it is reasonable, as the debuffs are suitably timed with overhead markers and the player need only look under their character's feet to dodge the circle AOEs. The pace feels scary because there is no breathing time during the sequence, but there is enough time to do it.

Divide and Conquer

What a sweet way to get Duty Finder to do line proteans. Awesome mechanic, excellent work.

Unlike in Alexandria, here the player is generally looking at the source of the proteans. Then the hits one at a time nudge the party into spreading out, especially since each hit has crisp telegraphs and the "selected for something" overhead marker - meaning they will probably be fairly spread out for the final hit on every party member at once, so they have a good chance of managing the final protean positions. (They still might get caught in adjusting loops and fail, but there is only so much to be done about that when the player is approaching the duty based on reaction rather than planning.) The cast name is also fitting, but it would not be nearly enough on its own.

More like this, please. Short of retraining the Duty Finder population to do general preparation including protean positions, you would never be able to drop the "guide rails" off of mechanics like this, but you could still gently introduce so many concepts. That would make the regular duties better stepping stones to the high-end duties, without making the high-end duties any less of an accomplishment.

Final Phase

Wuk Lamat's intervention here is demotivating because she steals the spotlight from the player, both in storytelling and by doing so much damage that the party's efforts are largely irrelevant. How can you expect the player to put in effort if you tell them that their effort does not matter? To make matters worse, the cutscene sharply interrupts the second Absolute Authority resolution (unless the party has not yet met the generous boss HP threshold to allow that), viscerally stopping the player from carrying out the duty's scariest mechanical sequence! They are all wound up and going, and then control is suddenly taken from them by Wuk Lamat forcing them to cede the stage.

NPC intervention in a duty can be done well, but it must not take away from the importance of the player's actions. For example, The Final Day (Endsinger Normal) also has an NPC intervention cutscene, but the Scions praying for the Warrior of Light's success makes the player's actions more important rather than less important. That is very fitting for the climax of an expansion's MSQ, while this is not. Likewise, The Final Day's NPC intervention comes when the player is expecting not to have control of their character, because they see the raidwide coming that previously needed a tank Limit Break 3 and they see that they do not have the Limit Break available to them. (As a result, the motivating effect of the Scions' prayers in The Final Day is similar to that of Phoenix's intervention in The Unending Coil of Bahamut (Ultimate) (UCoB).)

Having an NPC intervention related to the power of friendship is all well and good. It can even be based on trying to restore Sphene's memories. Suppose instead that after the second Absolute Authority completed, the Queen Eternal continued to overload and then a cutscene started. In this version, Wuk Lamat would be accompanied by G'raha Tia and Krile in trying to restore Sphene's memories, and for their success Sphene would summon one or more of the bosses from Alexandria. The NPCs would engage those boss(es), allowing the player party to continue fighting Sphene. Since the three NPCs are not enough to form a light party, the player would see the NPCs' resources slowly deplete from their own fight over the course of the final phase, creating a thematic sense of urgency and possibly even a (generous) hard enrage.

Dynastic Diadem

A basic donut with a small safespot, encouraging the player to move their character to the front center of the arena. This would be helpful prepositioning for Royal Banishment, but the shared point of the cones there does not match the center of the circle here, so it is not great.

Royal Banishment

Raidwide spam and an appropriately slow alternating cone sequence. A fine mechanic to convey Sphene's overloaded state, having simple movements and many small hits of unavoidable damage.

Overall Impression

This trial is really brought down by its final phase and its story context. The actual mechanics of the other phases would be very good to great with some polish to a few mechanics.

----------------

Despite many strong concepts and some truly excellent mechanics, every single group MSQ duty has at least one major design flaw, several of which are in visual design. This is not good, both for immediate enjoyment and for getting the player to the endgame with the confidence that they can enjoy it.

I may be an Ultimate raider, but I always worry when doing a new MSQ duty for the first time. I always have to. I nearly got hard blocked from continuing the MSQ once, and any new duty could be the one that stops me... or the one that tells me that the game is moving towards demanding something I cannot improve at. Patch 7.0's MSQ left me with a lot of concerns in that regard, since not a single one of the group duties felt good to play.

I will now continue my review of the design of the Patch 7.0 duties, in this subsection covering the two optional dungeons that are not required by the MSQ. Conceptually I would include Arcadion Light-heavyweight Normal in this subsection as well, but I have split them up for length, since both dungeons have mechanics that deserve extended analysis.

Being Expert Roulette dungeons, these duties can benefit from the player being encouraged to return to them many times after completing their stories, but they must still provide an encouraging first-time experience to the casual player while holding up to the repetition. After all, the rewards for the repetition-based gameplay loop become frustrating if the player is too discouraged to go back to the duty.

Since these are endgame duties, and returning players will often warm up first in the MSQ even if they return in later patches (when the latest MSQ is the patch MSQ not required for these duties), these duties have less need to concern themselves with the very weakest players. However, most active endgame players are still casual players.

(Since Duty Support does not include these dungeons, I had to take the arena pictures in Explorer Mode, so as not to inconvenience other players. Please forgive the lack of bosses in the arenas.)

----------------

Tender Valley (Level 100 Dungeon)

https://i.imgur.com/Sy1OWH5.png

Mobs Before First Boss

Four Packs

Not only can these be double pulled, but each double pull has one to two individually more dangerous mobs with higher HP. This is the same sort of opportunity for skill expression as it is in the leveling dungeons, and it is even more welcome here with the heavy repetition and full job toolkits of Expert Roulette.

First Boss - Barreltender

https://i.imgur.com/NXfubzu.png

Heavyweight Needles

This would be a normal set of cone AOEs originating from the boss, except that the early telegraphs are narrower than the actual AOEs; the telegraphs only expand to show the actual areas to be hit at the last moment. Only the larger open areas actually contain a safespot, while the more numerous small open areas get expanded over and hit twice.

As a result, the first-time player most likely follows the telegraphs and moves their character to either a small open area or the edge of a large open area, thus failing the mechanic. But even though they know what to do next time, they feel baited rather than encouraged.

Why? Because there was no other tell that the early telegraphs do not match the actual AOEs, nor are the expanding telegraphs used in the player's favor. (Fighting Kefka is fun in Sigmascape V4.0 (O8N) and its Savage version (O8S) because there is a tell for when he is being a trickster and the telegraphs do not match the actual AOEs, not just because the player is told by the story that he is a trickster boss. The expanding telegraphs in the circle version of Dispatch (Vanguard Commander R8) are acceptable because they start expanding early and are positioned carefully to guide the player into a true safespot.) There is no skill to be learned here, nor is the player expecting (or willing) to undertake the process of progression for a regular dungeon boss. This is a "gotcha", where the player just has to know / remember that the telegraph system lies here even though 100 levels of gameplay has taught them to trust it. (All the worse if the player remembers that telegraphs are granted by the Echo!)

This is the boss, and by extension the game, punishing the player for doing exactly what the game has taught them to do. That is not an enjoyable experience in any way; if anything, it is treating the player as lesser.

As such, this would be a better mechanic if it used something like the glowing lines from Black Cat Crossing (in AAC Light-heavyweight M1 (M1N)) as needles, even though those are further from the actual AOE shapes - at least it is intuitively obvious that the actual AOEs are not pixel lines, so the player would expect to discover how far they expand and in what shape.

What about the player who already knows how the mechanic works?

Well, "find a large open area" on its own is not interesting, even if the player does not know the hidden pattern. (There are only two possibilities: either the safespots are north and south-east, or north-west and south.) Even for the casual player, this is boring the 90th time, and this is cast twice at the start of this battle before combining it with any other mechanic.

To avoid that boredom, the player must be presented with opportunities for skill expression. This mechanic could permit skill expression by dodging only as far as necessary if the floor geometry had references for the player to work out where the AOEs will expand to. But the floor has no such reference lines, so that opportunity is denied to the player and the mechanic is just boring.

Just because the lore concept is an open desert field rather than a "settled" area, that does not mean you cannot have good floor geometry to facilitate skill expression. This boss jumps to / hits the center of the arena as part of several of its attacks, including this one, and the floor already has cracks to convey the arid desert environment. It would have been perfectly reasonable to have the bigger cracks radiate from the center of the arena due to the repeated impacts of the boss hitting that spot - after all, the battle takes place in the boss's home - and use those to mark the potential edges of the safespots.

As it stands, at no point is this mechanic fun. First it is unfair, and then it is boring.

Tender Drop

Drops many cacti, that will explode shortly after the resolution of the next cast (either Heavyweight Needles or Barrel Breaker). There are two types of cacti: those with small explosions are bare, while those with medium explosions have light brown spines and a pink flower.

I could just correctly say that the tell distinguishing the two types of cacti is inadequate and low-contrast, since the spines are tiny and the pink flower blends easily into the orange floor. But I want to show you what the player can easily wind up seeing while looking around at all these cacti.

Here are some freeze-frame screenshots from a gameplay video without third-party tools:

https://i.imgur.com/1UxQQL5.jpeg

https://i.imgur.com/gaAbRwM.jpeg

https://i.imgur.com/frPYkiB.png

While the video does not load in perfectly, you can see that the player's interface is much clearer than the cacti, because various visual effects are blurring the cacti. In the first screenshot, the blur is exclusively coming from the boss itself, as part of the animation for a subsequent cast that will resolve just before the cacti explode. In the second screenshot, the blur is coming from the point-of-view player's own job actions, which are often left on as an important source of job fantasy and/or gameplay feedback. (Even if the player knows about the option to set those to Limited or Off!) In the third screenshot, both sources of blur are present simultaneously, albeit with a different subsequent cast.

This is already bad for visual clarity. But what the player sees is even worse than this, especially when they do not already know about the flowers and spines indicating different explosion sizes.

Consider the player's time window to look over the cacti. They cannot do so immediately at the end of the Tender Drop cast, since the boss has just thrown the cacti into the air from itself. The cacti land in clouds of dust that obscure any spines and reach up to the flowers, so the player cannot practically use that time to rapidly look between the cacti either. They can only begin checking for spines and flowers once the dust settles, leaving them precious few seconds to look over all the cacti and the subsequent cast before deciding where to move for that cast.

As a result, the first-time player is forced to move their eyes rapidly either trying to locate all the cacti accurately enough to match them up with the subsequent cast, or waiting on the subsequent cast before checking the cacti near its potential safespots. (I say "trying" because there are far too many cacti for the player to succeed with precise locations for every one in the available time. The human eye has limits.) This rapid eye movement effectively creates further blurring, whether the player is using visual information from while their eyes are in motion, is locating some of the cacti purely via peripheral vision in order to have fewer eye movements, or both. (With some disabilities, it is also possible for the player's eye muscles to move their eyes slower than those of an able-bodied person.)

If you blur those screenshots further and look at each part for only a split second... well, the flowers and spines are no longer really discernable on the cacti. What the player actually sees of your beautifully rendered cacti may be politely described as "a blurry heap".

So the player may well not discover that the flowers and spines exist and are relevant from looking at the cacti before the mechanic resolves - to discover that on their own, they have to happen to catch a cactus with them and a cactus without them in their central vision for long enough to notice those elements, in close enough succession to remember what the other type looks like, in the midst of battle with a boss whose core "feature" is that the early telegraphs lie.

What about from looking at the explosions as the mechanic resolves for the first time? Then the player is encouraged to look at their character's feet, both to see how far the early telegraphs expand and to see if they got their character far enough from all the cacti. Unless they stood their character on top of a cactus, that is a terrible place to look to see either cactus type, let alone both. (And the size difference in the explosions is not large enough to be easily visible by looking at the curvature of the circle telegraphs, whether a small part in central vision or the whole circle in the sea of orange in peripheral vision.) So the player is even less likely to discover that the flowers and spines exist and are relevant at that point.

As a result, the player can easily go through the entire mechanic without discovering that two types of cacti exist at all, and be left with the impression that they failed because they did not locate the cacti quickly and precisely enough to locate a safespot. (Or that they succeeded by chance for the same reason.) In turn, this impression drives them to focus even more on locating the cacti next time, which is not just uncomfortable, it drives them further into exactly the rapid eye movement strategies that make it so unlikely for them to notice the different types of cacti!

For a significant subset of players, the visual design of this mechanic takes the effort they put into learning and steers it into a dead end. Once the player is in such a dead end, the only ways out are for a fellow player to tell them or for them to check video footage slowly with the intention to improve at the mechanic. Neither is likely, especially for the casual player! They will generally just conclude that the mechanic is unfair or that they cannot do it, especially in the context of a "gotcha boss" that has already shown them a willingness to be unfair.

And to make matters worse, disabled players are disproportionately likely to wind up in this subset no matter what their disability is. Visually impaired players suffer for the obvious reason. But a player with a non-visual disability will pressure their vision more to compensate for their disability, and has had ample practice doing so from the MSQ necessary to unlock this dungeon - which also drives them to rapid eye movement strategies, especially given the long history of past mechanics in their mechanical vocabulary where precisely and rapidly locating many explosion sources is the necessary method!

(I wound up in exactly this dead end, until a friend told me that the flowers and spines existed. Cue another friend shouting "what flowers and spines?!" There were only three people in that conversation, no bystanders, all raiders that can handle at least final floor Savage... and the two people other than me had not been delayed by the graphical update, so they had the flowers and spines on their screens many times.)

This is completely unacceptable. It is your job to prevent the player from getting stuck in dead ends, not encourage that!

What about the experience of the player who does know about the two different types of cacti, and that the safespots are always near flower-less, spine-less cacti?

Consider the sequence with Barrel Breaker. The player with this knowledge can plan to have their character ride the knockback into a corner, and accordingly look for a corner with a safespot. So for each corner, they naturally look at only the cactus closest to that corner, and the hidden pattern (that there are two safespots in opposite corners) ensures that they will find one after looking at at most two cacti in their central vision. This is a reasonable task for the time given, and thus a reasonable sequence.

(Notice that the hidden pattern does count as helping the player here, because the player does not have to know it exists to benefit from it. The player who does know about the hidden pattern only needs to look at one cactus to find a safespot for this sequence, and that does not count.)

But consider the sequence with Heavyweight Needles. There is not even always a shared safespot between the two mechanics, so the player may have to check the cacti over most of the arena to locate one that will have a small explosion and plan to dodge accordingly. And even if there were always a shared safespot, the player would still have to either check all the cacti near the Heavyweight Needles safespots once the early telegraphs show, or start checking all of the cacti before the early telegraphs show and waste some of their effort. (In either case, most of the cacti they check have flowers and spines, and so will not stop their search.) This is not at all reasonable when each cactus has to be checked in central vision to discern its explosion size.

I understand that your duty design team is neither your 3D art team nor your story writing team, but you must ensure that each visual indicator's necessary standard of clarity is communicated between teams and tested extensively before release. It is the visual design equivalent of marking "this cast name is a mechanic cue, fancy translations cannot be at the expense of clarity at speed" that you learned from Starboard / Larboard Wave Cannon in O11N and O11S. Otherwise the work you put into making each battle express the surrounding story is liable to cause gameplay problems, as it does here.

To salvage this mechanic, obviously the two types of cacti would have to be made visually distinct despite the game-induced blur on the screen and the effective blur from the player's eye movements. Rather than rely purely on high contrast and glowing at the potential expense of the battle's lore concept, I would both heavily differentiate the cacti and present the mechanic in a way that gives the player more time to check them.

Remember how I mentioned changing the floor geometry for Heavyweight Needles? Working from there, tiny green buds of cacti for all possible positions could be present at all times, and then this cast (suitably renamed) would grow the selected cacti from the bud form before they release their spines explosively and shrink back to bud form. This would naturally eliminate the time period before the cacti land and the dust settles, when the player currently cannot check the cacti effectively. More subtly, this would show the player that the set of possible positions is limited but large, and allow them to review the set in "free time" between mechanics or even outside of battle entirely.

While that would also eliminate the spines as a tell between the different types of cacti, the spines are nearly invisible in the midst of battle anyway so that is not much loss. Instead, I would have the cacti with small explosions grow thin and tall with no flowers or branches, while the cacti with medium explosions grow spherical with a glowing blue flower from all the water aether they are taking in to grow. Not only would the completely different shapes and moderate glow actually manage some visibility through the blur for the first-time player, but once the player realizes that the tall cacti have smaller explosions, they can handle future repetitions by looking for the tall cacti in their peripheral vision and then getting the precise position of one tall cactus and its surroundings in their central vision. (Which would stop their visual search much faster than is currently possible, especially for the Heavyweight Needles sequence. This is why I did not suggest simply making the cacti with medium explosions bigger - those are not the ones the player needs to find first.) So their effort put into learning would steer them away from the dead end and towards a reasonable method for finding the necessary visual information.

Barrel Breaker

The boss winds up a punch to the center of the arena, knocking back all player characters unless they use their knockback immunity. Simple on its own - aim for a corner - but is always directly after a Tender Drop cast.

Prickly Left / Prickly Right

While the early telegraphs and the glowing side of the boss show this as a 180 degree hit to that side, they again lie about the actual size of the AOE. But even for the first-time player who applies the "gotcha" from Heavyweight Needles here and tries to dodge just far enough for that, this mechanic still baits them into getting their character hit since the AOE expands much further. (The only safespot is entirely opposite of the indicated side.)

Once the player knows how this mechanic works, they can use the facing arrows on the side of the boss's targeting ring to locate the safespot precisely for optimization, so this is less bad than Heavyweight Needles. But it is still annoying that the early telegraphs lie with no other tell, and the mechanic is just not that interesting once known.

Overall Impression (Boss)

The core concept of this boss is unfair the first time, and poor visual design prevents it from redeeming itself on repeat completions. Even if its visuals were cleaned up, it is not interesting.

Mobs Between First and Second Boss

Four Packs

These can be double pulled, again with a single larger mob in each double pull. Each double pull is also preceded by a water pipe slide, both separating the encounters and showing the story of reaching a once-settled area.

Second Boss - Anthracite

https://i.imgur.com/c78bVc6.png

Anthrabomb (Basic Form)

Tosses three bombs onto the arena floor, with black arrow arcs leading to images of the bombs where they will land. (The bombs then explode with a single point-blank AOE each.) The arcs are a good and important part of the visual indicator, as they allow the player to locate all the bombs by following the arcs out from the boss. The cast length is also appropriate.

Anthrabomb (Line AOE Tutorial)

Tosses a single bomb into a pipe opening in the arena floor, with a yellow arrow arc leading to a yellow arrowhead (and no bomb image) over the pipe opening. (The matching pipes on one side of the arena then do line AOEs that knock back any player characters hit, with the two types of pipe being distinguished by color, shape, and patterns used as decorations by the previous civilization that built the arena.) The cast length is again appropriate, and showing this cast alone is important to have the player notice that this AOE does not hit directly where the bomb landed.

The edges of the line AOEs are supposed to be something the player can work out via floor geometry, specifically the slightly reddened lines of smaller stones across the arena that meet the edges of each side pipe. However, the color difference and the gaps between stones are so small that these lines become invisible as soon as any effective blurring is introduced, whether from job actions, boss animations, the player's eye movements, working in peripheral vision, or the player's camera movements. And even when they are not blurred, they are still interrupted by the ground pipes, which take up much of the space that is near the entrance to the arena and therefore is most likely to be near the first-time player's camera at this point.

I understand that this battle takes place in the ruins of a previous civilization, but you can and must still find a way to have either suitably crisp floors or mechanics that do not require them for visual clarity. (And I know you can do that; despite the restricted color scheme for the floors of Alexander - The Cuff of the Son (A6N), Alexander - The Burden of the Son (A8N), and their Savage versions (A6S and A8S), the raised and lowered panels for Swindler's Height debuffs are very clear.)

For example, you could have had the previous civilization color their floors and pipes using the natural and enduring colors of various stones, rather than fading paint. (Presumably the boss would blow away any dust.) That would give you a much less constrained palette, even within the hard building stones readily available in the western United States of America that Shaaloani is based on. (While it would limit the logical decorative pattern choices of the building civilization to relatively large, blocky options that make sense as being constructed directly out of stones, the current decorative patterns are not reliably getting seen anyway. Finer detail for lore purposes could be provided by carving the stone if you want.) If you are willing to grant the building civilization the ability to preserve softer stones in transparent and durable material, you could even have the edges of these line AOEs marked in ornamental obsidian, which would be very high contrast and available in the geographical area.

Regardless of your choice of justification for contrasting colors, the floor geometry showing the edges of the line AOEs should extend over the floor pipes. As it stands, the red circle pipes leave little to go on even once the player realizes that the floor pipes are each bisected by one edge of the line AOEs.

Anthrabomb (Final Form)

Tosses three bombs combining the initial two forms; two land on the arena floor and one goes into a pipe. Would be entirely reasonable with good visual design for the floor and pipes.

Carniflagration

Three final form Anthrabombs in quick succession, followed up by a basic spread, with no intervening castbars. The time given for each Anthrabomb is not terribly short, but the second and third appear during the placement and detonation of the previous bombs. The spread markers appear as the final set of bombs is thrown.

Here, it becomes even more important to mark the floor pipes and the edges of the line AOEs in high contrast, because the player's view of the indicators for the second and third sets of bombs will be blurred at some points by the movement of their camera as they move their character to dodge the preceding set.

But there is another problem with this mechanic.

With this sort of mechanic, the player knows from long experience that they only need to find one safespot, and that doing so is faster than mapping out all the safe areas in the arena. So especially if the player is used to playing jobs that can do their full damage at range, a natural response to the faster pace is to start by moving their character away from all the bombs - for which they only need rough locations from peripheral vision - and then further move their character to the nearest spot that will not be hit by a pipe. (This second movement is guaranteed to be small by the alternating layout of the pipes, which the player learned from the line AOE tutorial Anthrabomb, and the player can begin locating the yellow arc during the first movement.) For other mechanics of this sort, this is an excellent method to handle lack of practice with the specific mechanic's visual indicators, or simply high speed demands.

However, this method applied here generally drives the player to move their character up against one of the side walls, beside the pipes. (More so due to the lack of contrast in marking the edges of the line AOEs, encouraging the player to use the side pipes themselves as references.) At that point, consider the behavior of the player's camera. If they try to place their camera over the side pipes, it will instead hit the invisible wall over the side pipes and press up against their character, giving them a view of very little. If they instead place their camera away from the side pipes, even at the maximum distance from their character it cannot show them the whole arena for the subsequent sets of bombs. (If the player primarily moves with a mouse, that will incline them to do the latter.)

As a result, the player must fight their camera rather than the boss itself. This is annoying, both from the immersion break of fighting the interface and from the player's prior learning failing them because of the interface. The player winds up having to avoid this method of dodging at this boss, not because it would not be effective for their character but because it does not allow them to see the later bombs.

If the side pipes did not have invisible walls over them, the player would be able to point their camera into the arena from there and have it be useful just like they do with most arenas, which would eliminate the issue. If the attacks after the first were of some other form that did not need a clear view of the whole arena, perhaps something dodgeable by the castbar, the views available to the player after any first dodge would be adequate. If this mechanic was presented in a context that effectively granted the player an extra-zoomed-out camera to see the whole arena with, as Anabaseios: The Twelfth Circle (Savage) (P12S) does in its Gaiaochos mini-phases, that would also eliminate the issue. (Raiders colloquially and lovingly refer to those mini-phases as "UAV", short for "unmanned aerial vehicle", for the distinctive feel that is fully designer-intended but acts like the camera zoomhacks used in the TOP world race.) But none of these options are true here as the boss stands.

Should the player persevere and be strong enough to seek out hidden patterns even in a mechanic they find hectic, the situation does eventually improve for them. Many seemingly possible sets of bombs are actually blocked, creating several locations away from the side walls that always have a safespot nearby. Once the player learns to keep their character away from the side walls, they have the view of the arena needed to notice one or more of these locations via extended repetition. In turn, should they gravitate to one of these locations, the smaller required movements from staying within that area make the mechanic less hectic, finally allowing them the time to reliably reference the low-contrast markings on the current floor during the mechanic. They may even be able to optimize their movements for job uptime. But the path to that point is needlessly discouraging, and struggling players may not be able to follow it at all.

Overall Impression (Boss)

This boss is an unfortunate victim of its visual design and the game's standard camera. There is optimization potential under the surface, but it requires finding hidden patterns to avoid fighting the interface and the visual design.

Mobs Between Second and Third Boss

Two Packs

These can be double pulled, and the second pack includes two Yok Huy Vessels whose auto-attacks cleave an unmarked circle around their target. This is mechanically interesting. Further, it is not unduly burdensome to the casual healer even if their party chooses to take the auto-attacks for better damage output, since every healer job at level 100 has a variety of AOE healing tools that can easily handle this amount of incoming damage.

Indeed, I consider this a good thing, as it encourages less skilled healer players to try using some party healing cooldowns on trash pulls and then notice being able to use those cooldowns again on the next boss. This gently hints to them that AOE healing cooldowns can be effectively used on trash pulls, even if only the tank is taking damage.

After these two packs, there is an arrow slide:

https://i.imgur.com/iqKtz4L.png

This arrow slide is clearly meant to introduce the player to the arrows that the next boss places for Greatest Labyrinth. However, since it is completely straight, the player engages with it in exactly the same way as the water pipe slides in the previous section. This makes it easy for the first-time player to miss the behavior of the arrows, especially if their party members are already familiar with the dungeon and charge ahead. (Even if the whole party is there for the first time, they may well be looking at the treasure coffer or still thinking about the previous trash pull.)

To make this a less missable introduction, I would have the arrow slide take the same amount of time, or at most the amount of time that the physical pipes take, but include some turns without turning the whole hallway. (The exact setup may have to be tuned to avoid motion sickness.) This change would make it clear that the behavior is different from the earlier physical pipes, without requiring the player to already understand the arrows or adding frustrating delays to repeat completions.

One Pack + Miniboss

The second pack after the slide is a single Yok Huy Attestant, standing in this arena:

https://i.imgur.com/yYyAmZH.png

The veteran player who has leveled many jobs through the Shadowbringers range may well recognize this arena from The Qitana Ravel, and expect similar mechanics, depending on how they leveled. And they would be correct; the mechanics of the Yok Huy Attestant are almost identical to those of the Ronkan Dreamer minibosses there, with the only difference being that here the Sun Toss ground AOE dropped on a random player character is targeted immediately as the statues fire their line AOEs from Ancient Wrath. Such a player already knows how the statues work, and can manage this timing by verifying the statue dodge before that animation begins.

However, the player who only subscribes intermittently, or who is newer and so has not had an extended period of being caught up on the MSQ (or nearly so) to level many jobs while being able to access The Qitana Ravel, may only have seen the Ronkan Dreamers in one dungeon completion - and that two expansions of MSQ ago. They are unlikely to remember the miniboss or the statues, and so must learn how the Yok Huy Attestant works without that reference. Therefore, they must watch the animations of the statue AOEs to check their guess at Ancient Wrath, which pulls their eyes away from the Sun Toss telegraph. While they will probably notice the orange telegraph, the delay in looking at it makes it reactive for them until they stop looking at the statue AOE animations.

As a result, even though neither mechanic is reactive once learned, the player's learning processes for the two mechanics interfere with each other. Even veteran players can be in this situation if they do not often repeat leveling dungeons, or if they level each job to the cap each expansion by directly queueing the highest dungeon that job can access, and so have not seen The Qitana Ravel in two expansions. Or if they simply want to re-check their understanding, since this is a different dungeon and this miniboss is the first strong hint of similarities to Ronkan design.

In high-end duty design, such interference is often an acceptable side effect of forcing the player to manage their attention through rapid mechanical sequences. It is acceptable there because the player is expecting and willing to undertake the process of progression, and so allows for getting their character hit by the next attack in the sequence in order to learn the current one. But this is a regular dungeon, where you have none of that patience from the casual player. They may easily require more than the two casts of Ancient Wrath before Sun Toss is added to learn the former mechanic well enough to be comfortable not looking at its resolution, and then judge this section of the dungeon on the resulting poor first impression. (Especially if they had bad experiences with mechanics in the Patch 7.0 MSQ to get here, and/or are being extra careful after the "gotcha" mechanics of the Barreltender.)

Even if the player does know Ancient Wrath well already, it is still a little satisfying to actually get to see the uptime dodges that stay in melee range / Line of Sight without worrying about Sun Toss. Let the player have those moments.

To make this section still strongly reminiscent of The Qitana Ravel without being exactly the same, I would swap out Sun Toss entirely for a different mechanic on the smooth timing of the older miniboss. The replacement mechanic could be a little harder than in The Qitana Ravel, since the player here is more experienced, this dungeon is not strictly mandatory for MSQ, and Expert Roulette dungeons are repeated more than Leveling Roulette dungeons. For example, the Yok Huy Attestant has the same 3D model structure as Lozatl (the first boss of The Qitana Ravel) complete with the crystals on each shoulder that glow to indicate hitting that side of the arena. Those side cleaves would have fit well here, being simple on their own but combining with the ruined wall segments to present new considerations to the player, and would also be a recognizably Ronkan mechanic in the context of the model and the statues.

Other than the one disappointing timing choice, I do like this section. It is a good way to hint at story through mechanics before spelling it out in the post-dungeon cutscene, and being able to pull the preceding pack into the arena is welcome in Expert Roulette. (Even if the second Ronkan Dreamer in The Qitana Ravel offers a harder pull, with an additional pack and a tighter Item Level sync.)

Third Boss - The Greatest Serpent of Tural

https://i.imgur.com/7UzBGhe.png

Bouncy Council

The most repeated mechanic of this boss, spawning a set of smaller serpents that each do one ground AOE (simultaneously) and then leave.

(The first cast is exclusively thin front-to-back lines, the second cast is exclusively point-blank circles with one corner open, and all subsequent casts mix the two types. These subsequent sets always have exactly two circle serpents that occupy adjacent corners, and three line serpents occupying the remaining corners and the middle of the arena. Line serpents may only be pointed north-south, east-west, or directly diagonal.)

I appreciate that these serpents have bright white-blue arrows marking the direction they are bouncing in. This is a crisp, high-contrast cue that does not obscure the lore idea of the serpents creating their AOEs by bigger bounces once they have wound up, nor does it rely on glow effects. It provides exactly what the player needs to work out for themselves what form and direction each AOE will take. Excellent work. I would personally prefer that the arrows appear immediately, but they are structured very well.

Unfortunately, the placement of these AOEs relative to the floor tiles is not good. This floor is composed of square tiles carved with geometric patterns courtesy of the Yok Huy, and each corner serpent is precisely centered on a tile - but the width of the line AOEs is slightly larger than the side length of a tile, denying the player precise reference points even for the AOEs that are oriented along the tile grid. (And since the telegraphs showing this appear for less than a second, the player can easily mistake them for exactly the same width as the tiles. If they then use the tile edges mistakenly as reference points and get their character hit, they will likely blame the hit on snapshotting or lag rather than having any inkling of the true cause.) The diagonal and circle AOEs have nothing resembling reference points on the floor, which avoids being misleading but still denies the player the opportunity to precisely dodge. Such reference points could have been made by incorporating their possible edges into the carving patterns.

If these AOEs and the floor were tweaked to provide the appropriate reference points, this could be a really fun mechanic, especially as the final form creates small safespots. There is enough time for the player to see all the serpents and work out roughly where they will hit, which would move up to precise locations with the floor references. So for their effort, the player would be rewarded with the thrill of seeing all the AOEs miss their character despite the small safespots, plus of course the job optimization opportunities as previously discussed. But as it stands, those opportunities are not really available to the player, and having only a rough location for the safespots results in worries about reacting to the telegraphs to adjust their character's position.

Greatest Labyrinth

This mechanic opens by drawing all party members into the exact center of the arena, which is useful in some circumstances. However, since here the player must immediately undertake a visually demanding task, the draw-in is disruptive to their view of the arena and their personal tracking of their character's location.

If the player moves their own character, they will naturally include that movement in tracking their character and their camera movement relative to the arena. This allows them to partly compensate for the camera movement automatically, using many of the same brain systems that have evolved for movement on Earth. Forced movement mechanics do not permit this, and instead force the player to re-establish that tracking once their character is in the new location. (Especially forced movements that give no prior indication to the player, such as this one. Mechanics that ask the player to plan around a forced movement at least prepare them for the disruption, especially if they show the player character's new location such as in Eden's Verse: Iconoclasm (E7N) and its Savage version (E7S).)

This does not make forced movement mechanics inherently bad. Far from it. Forced movement has even been used to spectacular effect in some mechanics, such as opening the levitation mini-phases of The Interphos, or the ending of each Relativity in E12S.

But the effects of forced movement on the player's visual tracking must be considered when placing forced movement mechanics and visually demanding tasks. In this context, where the visually demanding task of working with the arrow placements is the core of the mechanic, having a forced movement immediately beforehand disrupts the task without being interesting or necessary.

To avoid this disruption without compromising the arena transformation, I would use the same technique that Everkeep does in Dawn of an Age: place a clear ground AOE with ample time over the parts of the arena you are going to change, and have that AOE KO (and/or move) any player characters that remain in it at the end of the cast. There is no problem with extending that AOE somewhat further to prevent player characters from stepping over onto a just-formed arrow, nor with the bind while the arena transforms.

After the arena transforms, the player is presented with the red text "You feel a longing to vacate your mortal coil...". I understand that this is meant to alert the first-time and/or casual player to the Greatest Curse debuff, which KOs their character if not cleansed by the correct corner tile. But this is not a good use of the red text.

First, this text is exclusively focused on the consequence for failure, not what the player must do to succeed. Such a focus would be appropriate for some puzzle mechanics, where saying too much about a route to success would spoil the puzzle. But this is not a puzzle mechanic; it is a rapid visual processing mechanic. Second, this text is vague and flowery about the consequence. Even if the player is well familiar with reading their debuff bar and with various consequences for failing mechanics, they still have to mentally translate this before they arrive at "failure results in KO". Again, time is of the essence here. Flowery text can sometimes be useful for lore puzzles, which this is not.

I would change the text to "Navigate the labyrinth to break the curse!" That is straightforward and focused on the route to success, mentioning the debuff only as needed to alert the player to its presence.

So what must the player do to succeed at this mechanic? (In more detail than "navigate the labyrinth".)

To cleanse the debuff and thus avoid the KO, the player must get their character to the single corner tile of the arena marked by a blue circle. The arrow tiles prevent the player from simply moving their character there normally. So the player must look over the arrows - the labyrinth - to find out which directly reachable tile will move their character to the desired corner. (There is only ever one such tile out of the eight tiles adjacent to the unchanged central area.)

Let us break down the steps of this task. In particular, let us break down the most efficient general method that guarantees a solution for the player who understands what the mechanic is but has no knowledge of any hidden patterns.

First, the player must check the corners of the arena until they find the one tile marked by a blue circle, the conceptual endpoint to the labyrinth. (The corner markings are not prominently visible until Greatest Curse has ticked down to 16 seconds remaining, even though it starts at 18 seconds.) This may require them to check all four corners, depending on where they start looking. The first-time player wil likely need to see the circle to know that that tile is the correct one, requiring them to check the corners individually in central vision. (The same is true for the player who has completed the dungeon before but has any form of colorblindness that makes the blue and the purple difficult to distinguish.) The player who has done this before, barring colorblindness as mentioned, can use their peripheral vision to check multiple corners at a time. In either case, they may require a camera turn, as viewing all four corners of the arena with their character in the exact center is only possible with precise camera tilts and angles, and even then uses the very edges of the screen that are often covered in HUD elements. (Ironically, the camera tilt settings that make this possible are the same ones that put Valigarmanda's beak off the screen during its lightning beak animation tell. Further, the range of viable settings and angles that eliminate the camera turn would be larger if the player moved their character on their own to the edge of the unchanged central area.)

Second, the player must work backwards from the endpoint to the central area. They work tile by tile to discover the path, asking and answering the question "how can I get to this tile?" over and over again on the smallest scale. (Those readers familiar with computer science may recognize this method as a breadth-first search.) So they look at the two tiles adjacent to the endpoint, and check each of those for whether its arrow points into the endpoint. Due to a hidden pattern allowing the player to return their character to the center of the arena for melee uptime, exactly one of those tiles will point into the endpoint. (The player does not necessarily know that, but they still benefit from it after checking both adjacent tiles by not having to look for paths to both.) They set the one tile pointing into the endpoint as their intermediate goal, since any path there will lead them to the endpoint.

Then they look from their intermediate goal to the two tiles adjacent to that (not counting the endpoint itself, since they already checked it) and check each of those for whether its arrow points into their intermediate goal. One or both of those will, and they set their intermediate goal to those new tiles. (Yes, eventually there can be multiple tiles pointing into an intermediate goal tile.) Then they look at all the tiles they have not yet checked that are adjacent to their new intermediate goal(s), again looking for arrows pointing in. (Some intermediate goal tiles will eventually have three unchecked adjacent tiles, being not on any edge of the arena.) They repeat this process until they find an intermediate goal tile that is adjacent to the unchanged central area, since they know they can reach that tile by normal character movement.

Finally, they move their character to the intermediate goal tile they can directly reach, and wait for the arrows to transport their character to the endpoint. While travelling the path requires no further input from the player, they must leave time for it to occur before the debuff expires.

(What about working forwards from the directly reachable tiles? Tempting as that is, the player has no guarantee that doing so will not lead them all around the arena checking long and fruitless paths. (Even if there is a further hidden pattern guaranteeing that, the player has no way to find that out without either risking many failures or consulting many videos. Regular dungeon guides are generally made shortly after the duty's release without such in-depth knowledge, and working backwards keeps the player's eyes away from most of the directly reachable tiles. Plus the Patch 7.0 duties leading up to this point have inspired no confidence in the player that your mechanics will be consistently visually reasonable.) And if they put their character on a wrong path, the forced camera motion effectively blurs their screen for the duration, leaving them less time to try again.)

(What about working backwards but considering each new intermediate goal tile as soon as it is identified, rather than continuing to check each adjacent tile to the current intermediate goal tiles? That would guarantee success, and indeed in computer science that would be known as a depth-first search. But that would place greater demands on the player's working memory - to hang on to each not-yet-checked tile for longer and accumulate a larger list - without improving the worst-case scenario.)

This requires a large number of visual checks overall, most of which are checking arrow directions. And the arrows are green on brown, with the green color ranging from dark to lightly glowing in a way totally unrelated to the direction of the arrow. The glow is visual clutter, not an aid to the player. So in order to discern an arrow's direction reliably, the player must not only put that arrow in central vision, but they must also stop their eye movement on that arrow briefly to pick it out from the surrounding brown tile. (Especially if they use postprocessing to cope with the forced bloom from the graphical update!)

(You may find that you can discern an arrow's direction reliably when looking at an adjacent tile, not the arrow itself. But I cannot, and I have a recent clean eye exam including peripheral vision and colorblindness testing, so clearly that capability is not available to all your players. I can only imagine this is worse for players with visual impairments. As such, I will proceed based on checking an arrow's direction in the manner I described in the previous paragraph, since that is a much more realistic assumption of the player.)

And the tiles the player is checking at each step are not adjacent to each other, resulting in short eye movements in many different directions. As a result, the player is forced into a flick-stop-flick-stop repeating eye movement pattern, which can be physically uncomfortable to sustain even without a disability raising the time pressure.

Let us take an example to see how much eye movement this is in a bad case.

Suppose that the player goes through nine goal tiles in the course of their search, including the endpoint but not the directly reachable tile, and that each of them has two adjacent tiles not yet checked. (This is very possible when dealing with a long path and/or multiple intermediate goal tiles in the same step. I am not convinced that it is even the worst case.) Suppose also that the player checks the tiles in the worst reasonable order, so they must perform a camera turn to find the endpoint and must check both adjacent tiles for each goal tile until the last one (where they can reason out to prioritize the directly reachable tile).

The camera turn acts similarly to one eye movement. Then the player - even if they can identify the endpoint in peripheral vision - must move their eyes to one tile adjacent to the endpoint, then the other. The same goes for each of the remaining goal tiles until the last one. So that is one eye movement from the camera turn, plus two each from the first eight goal tiles, plus a final eye movement to check the directly reachable tile. 18 eye movements, each of which requires a brief stop.

Suddenly 16 seconds is not all that long, even setting aside time for the player to reorient, the player to move their character, their character to travel along the arrow path, and client-server latency. In fact, when including those, this mechanic demands almost two eye movements per second to ensure success.

This is an even higher eye movement frequency than Ultimate mechanics such as TOP's Party Synergy (initial clone AOEs) and Predation Dodge, where the player can reduce character movement time with planned Sprint and expects ample practice to improve their visual processing speed. (Notice that Sprint does especially little here, since part of the character's movement time is along the arrow path where movement speed buffs do not help.) Sure, those include some larger eye movements and that does matter, but both repetitive motion and rapid stopping and starting are their own physical burdens.

(Party Synergy method: Check Playstation markers before the clones appear, so you must preposition your eyes for the markers rather than the clones. Ignore the middle clone, locate the side clones (and the eye) roughly by spawn sound and peripheral vision. Look at the arms of the side clone visually nearest after checking Playstation markers; if they are empty, look down to their feet to check whether they have skates or regular feet. (Skates means Superliminal Steel, regular feet + empty hands is the decoy. There is only one decoy.) If you found the decoy first, look at the arms of the next clone nearby. Upon identifying a weapon, regardless of finding the decoy or not, start moving your character to the general area of the two potential safespots while looking towards the opposite side of the arena for the other clone. Again look at the clone's arms; if they are empty and you have not yet found the decoy, check their feet. Identifying the second weapon stops your search.)

(Worst case to start of character movement: 3 eye movements (arms, feet, arms). Worst case to complete solution: 5 eye movements (arms, feet, arms, feet, arms). Your eyes will only have to cross the screen once. Time between clone spawn and AOE snapshot: 5 seconds. This is slightly over one eye movement per second, with some visual clutter from the decoy but less than in Greatest Labyrinth.)

(Predation Dodge method: Look at the arms of one of the first clones to spawn. (There is no decoy here; empty arms always come with skates.) Then look at the arms of the clone opposite it. Preposition character roughly between the two potential safespots and preposition eyes on one of the remaining intercardinals, where the next clone's arms will be. Once that clone spawns and you identify its weapon, look opposite it at the arms of the last clone. Then look at Final Omega until the glow shows for the hourglass cleaves. Move your character to the first safespot, while looking at the first safespot for precise positioning. Preemptively look at the second safespot, again to aid positioning.)

(7 eye movements every time, in 11 seconds between the first clone spawn and the first AOE snapshot, with no visual clutter. Your eyes will only have to cross the screen twice. If you can do Party Synergy, you can almost certainly do this, with some extra time to handle the Final Omega overlap.)

Some players do choose to use more visually demanding methods for TOP than those I have just described, because they prefer to always start with the same gender of clone in identifying each pair of AOEs. But that is an option. When they do that, they are choosing to do a harder eye movement task in exchange for an easier pattern matching task. If they find themselves struggling with the harder eye movement, they can take the easier eye movement task and learn the pattern matching starting from either clone by practice and memorization, which gives them an actionable route to success.

Greatest Labyrinth has no alternate option. (Other than trial-and-error and following fellow party members, both of which are unreliable and not really solving the mechanic.) The player who finds the repetitive flick-stop-flick-stop eye movement physically uncomfortable, injurious, or outright impossible is just stuck with it.

Why is a regular dungeon mechanic more demanding than an Ultimate in raw eye movement frequency and sustaining that frequency, both of which the player cannot change if they find themselves limited?

This is absurd.

To examine the intensity of this visual task in another way: the time that the human body (including brain) takes to initiate a "flick" eye movement varies quite a bit even testing the same person at different times, but a typical time is around 200 milliseconds, or one-fifth of a second. That is not counting any time the player spends deciding where to look depending on what they see. So by asking the player to make almost two eye movements a second, you are asking them to spend on average almost 40% of the search time literally waiting for their eyes to get moving. Again, this is absurd.

(Some readers may be aware that in some circumstances, such as examining a face, humans regularly make multiple eye movements per second. Presumably that is not mired in delays because the task of looking at a face allows the human brain to plan several movements ahead to look at different facial features - a sort of planning Greatest Labyrinth permits very little of, Party Synergy permits somewhat, and the Predation Dodge completely supports once the first clones appear. And the tasks where such frequent eye movements are common also do not have such hard stops, hence why we regularly miss details such as accidentally repeated words when reading. I can tell you that I find reading much more comfortable than Greatest Labyrinth, despite working far faster to read at my natural speed.)

The only reasons this demand did not create an obvious disaster are that the player can notice their party members' attempts in peripheral vision, and that party KOs do not lose particularly much time on regular dungeon bosses.

Extending the time limit could improve the situation, especially given the hidden pattern. (The hidden pattern is well-designed, allowing faster players to optimize for job uptime without having to worry about estimating the time taken to travel the correct arrow path.) That is why Deltascape V3.0 (O3N, Halicarnassus) does not suffer the same way from its own visual labyrinth, despite relatively low contrast; even if the player cannot use the motion of the tiles to find the tile directions in peripheral vision, they have far longer to search and several still tiles to work towards along the way. It asks the player to do several small searches, which limit the length of the worst-case scenario, with a generous timer.

But I think there is a more elegant solution here: make the glow help the player visually, as would be expected from such a mechanic.

Each arrow is already composed of one large triangle as its head, two chevrons pointing towards the head, and a tiny triangle filling in the middle. I would make the head triangle glow clearly regardless of bloom, and have the rest of each arrow in a single non-glowing medium green. (Including on the arrow slide earlier in the dungeon.) The idea is that when the player is looking at the endpoint or an intermediate goal tile in their central vision, they would reliably be able to see the glow of any adjacent arrows pointing towards it in their not-quite-central vision, clearly differentiated from those pointing in other directions. Making this reliable eliminates the need to look directly at adjacent tiles that do not become intermediate goal tiles. This would cut the required eye movements almost in half and make the directions of motion much smoother, thus making the visual demand far more manageable without changing the idea or the complexity of the visual labyrinth.

The faster the player needs to see and process a visual indicator, the clearer it has to be. The necessary standard here is much higher than it is in O3N because the player must work faster.

Moist Summoning

This would be a simple multi-hit stack, except that each hit leaves a puddle from the serpent add that hits the stack area. As a result, not only does the first-time player have to react to the presence of the puddle, but every time the player must react to their party members' choices of where to move out of the puddle.

While this need to react to fellow players is not a disaster, it introduces just enough of a delay that each puddle's time to activation is annoyingly short. This would be a less reactive mechanic if it overlapped with something that suggested a movement direction to the party, such as a rotating cone AOE, despite the overlap technically making the mechanic harder.

Greatest Flood

A simple knockback from a corner tile, specifically (relative to the center of the arena) slightly further out than the closest corner. This is only complicated by the puddles from Moist Summoning persisting.

A fine mechanic; the puddles present a consideration to the player without being too difficult, and the player can use their knockback immunity for uptime or dealing with bad puddle placement.

Great Torrent

Hey, an actually fun mechanic! The player has the time with the telegraphs to precisely position outside the first AOE in the sequence, and enough time to spread after the sequence completes.

Overall Impression (Boss)

This boss has a perfect reason built in for excellent floor geometry and the mechanics to match, but it does not actually give us the floor geometry that would enable precise dodging. Greatest Labyrinth is also a low point, since it leaves some players with no actionable route to success.

Overall Impression (Dungeon)

The trash mobs are interesting but not too difficult almost across the board, but every single boss has serious visual design issues.

The fact that visual design plagues every single boss in this dungeon, not just one mechanic or even one boss, makes it feel like this is a test of whether lowering visual clarity makes for fun difficulty. Whether or not you intended it as such a test, I will give the answer: no, and that is not something the vast majority of players would give you useful feedback on regardless.

The average Ultimate raider does not consciously think about what sorts of capabilities a mechanic requires, nor how they move their eyes to do a mechanic. Any given casual player is even less likely to think about such things. If they struggle with a mechanic because you obscured the needed visual indicators and references, outside of obvious cases like "this duty is orange on orange" they will generally not perceive that their issue is with a lack of visual clarity.

They may instead notice only that they keep failing the mechanic without any improvement past a certain point, and simply say "this is too hard" on that basis. They may completely miss a core part of the visual indicator, as with the cacti created by Barreltender, and base their assessment on a mistaken idea of what the designer intended. They may notice that they are unable to get all the information they need in time and deem the mechanic or duty "unfair", especially since Barreltender shows a willingness to be unfair from the beginning of the dungeon. They may find themselves delayed in discerning the needed information, correctly perceive that their time to make a decision with that information was very short, but declare the mechanic "reactive" since they do not consciously perceive that the delay came from obscured visuals rather than no visuals. Or they may simply not like it but be unable to pin down why.

All this gets drowned out not only by the angry back-and-forth over "is Dawntrail too hard?", but also by things that are easier to complain about. It is within the reach of the average casual player to identify the Strayborough dolls as annoying, so you get lots of complaints about them. It is not anywhere near the reach of the average casual player - in effort or knowledge - to explain how your visual design choices affect their visual processing of mechanics, the way that I have done here.

Meanwhile, the player who can handle the obscured visuals is totally unaffected. The mechanic is not any more exciting for them, because the mechanic underneath is the same. The arena and animations do not fit the lore and story any better, because an immense variety of lore concepts can be presented with clear visuals for needed mechanical indicators. I described how I would change the visual design of this dungeon to demonstrate that lore fit and visual clarity are not mutually exclusive. And of course, visual elements that are not relevant to mechanics can be as subtle as the artist and/or lore writer desires.

It is exactly as I wrote in the design principles section of this review: making needed visual indicators difficult to see is anti-fun.

I have taken the time to give this detailed breakdown because I want you to adopt - and follow - visual clarity and accessibility standards that suit the real-time nature of this game. No more orange on orange, as players told you about Eden's Verse: Furor (E6N, Ifrit and Garuda), Asphodelos: The Third Circle (P3N, Phoinix), and their Savage versions (E6S and P3S). Those were two raid tiers apart, you had the time to act on the feedback from Eden's Verse in your production pipeline for Asphodelos. No more indicators that require color vision or are heavily obscured without it. And nothing the player needs to see should ever wind up as a blurry heap!

Build visual clarity and accessibility standards into your design process. Make them reflect the requirements of the mechanics that use each visual element. Seriously.

--------

The Strayborough Deadwalk (Level 100 Dungeon)

https://i.imgur.com/XhhwIG1.png

Mobs Before First Boss

Four Packs

Nothing mechanically interesting here, but can be double pulled.

First Boss - His Royal Headness Leonogg I

https://i.imgur.com/TRnfomJ.png

Falling Nightmare

Introduces part of the dolls' punishment for failure by putting the falling heads over each player character, but with the opportunity to dodge since the player character has not been bound by a doll. This would be a good introduction, except that the heads have barely any contrast with the floor and the dolls are awful to deal with.

(The heads inflict 10 seconds of the Benoggined debuff, which has a "ha ha you messed up" visual as well as being equivalent to Out of the Action combined with Heavy.)

The heads do have green sections with more contrast, but the player cannot readily learn to use those sections, since their 2D shapes on the screen vary greatly depending on the boss's choice of the head's facing relative to the player's camera.

Team Spirit

A brief warning cast for the dolls lining up around the edge of the arena, preparing to march. In itself, this is a good thing.

Spirited Charge

Has the dolls begin marching across the arena.

Since these particular dolls are one of the most complained-about mechanics of Patch 7.0, let us go over them in detail. Why are these so annoying, when players have previously enjoyed moving object mechanics such as the brooms in The Grand Cosmos?

First, note that the dolls march quite fast, even faster than Honey B. Lovely's moving hearts. Their speed feels comparable to the bullet hell bubbles created by Feather Ray in The Skydeep Cenote. So just like those bubbles, the player's screen tells them that they got hit by a doll while also showing positions that would have resulted in a successful dodge if those positions were accurate. The resulting "I was out of that!" is all the more unpleasant when encountering the dolls over and over again in Expert Roulette.

Second, the time available after the dolls line up is not enough to check all of the dolls' starting positions. There are only four seconds between when the dolls first stand still in their starting positions on the player's screen and when the dolls begin moving visually, which means even less time between when the player first has the opportunity to see the dolls' starting positions and when the dolls have actually started moving on the server. This is not remotely enough to check what seems to be 16 dolls, arranged in roughly a square with four to each side. (And the player needs to see all the dolls, not just those aimed to march near their character or a single potential safe spot, due to the overlapping mechanics.) To make matters worse, the arena is a circle whose markings do not correspond to any "lanes" for the dolls, making it more difficult to visually trace their future paths. And since the dolls are relatively densely packed, the player needs precise positions to locate safe areas, which they must obtain from central vision.

Third, the starting positions are not sufficient on their own to know the dolls' entire paths, nor the timing of the dolls traversing those paths. Different groups within the square of dolls start moving at different times, which are completely unmarked other than by watching the dolls move. (Even if the order is static, the player would have to spend a long time watching the dolls - without worrying about their job or dodging - to find that out. I think it is a static order, with the east group always moving first.) And when a doll reaches roughly the edge of the arena (perhaps an edge of the hidden square) after marching from the line-up in Team Spirit, it aims at a party member's position (seemingly at random) and marches a second time in that direction. (In one section of the battle, the dolls even repeat this process of picking targets and marching multiple times, as the boss is too busy casting other mechanics to present another Team Spirit or Spirited Charge.) As a result, even in theory, the player does not have enough information to know the dolls' paths and timings until after each doll has not only turned around, but moved far enough to show which player character it targeted / direction it chose for its final march. They could try to determine the dolls' aim from facing, but seeing each doll's precise facing requires central vision and they cannot put every doll in a group in their central vision at once. (Again, the doll heads have barely any contrast with the floor, and the doll dresses are not much better.)

Fourth, the player does not have just the dolls to deal with. Even the first Spirited Charge has another mechanic begin resolving before the dolls finish their march. Naturally, this places greater demands on the player in every aspect of mechanic handling, as they are contending with more than one mechanic simultaneously. On top of that, the nature of the overlapping mechanics blocks the most comfortable way of dealing with the dolls.

The preferred strategy for the dolls alone would be to stand still in the center of the arena while as many dolls as possible pick their targets for their second march, then move out of the center between two dolls' paths. (The center of the arena maximizes the minimum distance between the player character and any point from which a doll might target them, thus providing the longest guaranteed lead time for the player between a doll's choice of target and needing to move their character or fail the mechanic.) But this strategy will fail every single one of the overlapping mechanics. Technically, the Looming Nightmare chasing AOEs allow the half of the party whose characters are not targeted by those AOEs to adopt this strategy, but that is not sufficient to reliably cover every player who would want to use it. Pushing some but not all players into a task they find unpleasant is still bad game design.

Fifth, this is a reactive mechanic. The player does not strictly need to react to their fellow players' choices, only to the dolls' paths, but they have very little time to react to the dolls' paths considering what information they practically have and when they receive it.

Even to check the safety of a single spot for the dolls' first march, the player must check all four sides of the hidden square for the precise positions of dolls that would pass by, involving at least eight dolls due to the tight spacing. Setting aside the question of the player's visual path tracing capability, that is eight eye movements just to get those dolls' locations in central vision, so doing that before the dolls begin moving requires at least two eye movements a second with stops on each doll. That speed is even higher and more burdensome than Greatest Labyrinth in Tender Valley, which was already a serious problem, making the only realistic assumption that the player cannot do it.

Instead, the player checks only some of the sides of the hidden square before the dolls begin moving. (For example, the most comfortable way to do this is to pick a corner of the hidden square and move one's character towards it while looking at it, refining the precise choice of position so that the nearby dolls pass by one's character. All the dolls that pose immediate threats are visually close to each other, their closeness to the potential safespots allows the player to look at spaces between dolls, the player need not look elsewhere to place their character in their chosen safespot, and a corner grants the longest time before the dolls on non-adjacent sides reach the vicinity.) But this means that the player must check the other sides of the square while the dolls are already moving on their screen.

A doll visually crosses the arena in about seven seconds, which sounds like a lot of time. But the player has less time than that to complete their character's movement, both because the true server position of the dolls is ahead of their client display and because putting their character in the edge of the arena is a mechanic failure of its own. (A less punishing mechanic failure than being caught by a doll, as the wall inflicts a DOT rather than the usual instant KO, but still a failure.) Moving their character takes time of its own, as does the decision-making aspect of reacting to the remaining dolls' paths. Their character's movement takes time to reach and be processed by the server. If they are playing a melee DPS (or a tank with a melee DPS in the party), they will be disinclined to move their character to a corner. Even if they picked a corner, they may well have picked the corner beside the dolls that start last, not first - remember, even if there are hidden patterns, the first-time player does not know that. (And the last dolls start about 2.5 seconds after the first dolls, so that in itself would buy testers who knew a hidden pattern quite a bit of extra time.) The player can look at the first dolls far away from their character while waiting for the last dolls near their character to move... but during their first time, they do not know when the dolls near their character will start moving, and they need that information to know when they can safely move their character for later dolls and overlapping mechanics. (Likewise, they need to know if the doll hitboxes are bigger than the visual models.) So they will likely look at their character or the nearby dolls instead, making the time pressure even worse until they know the timing well enough to stop doing that.

Let us generously not assume the worst case, the first-time player who has been trained by their job to stay in melee range. Instead let us consider the player who is confident that their chosen initial character position in a corner will evade the dolls that start nearby, and so moves smoothly to looking for the dolls from the far sides that will eventually be near that initial position.

Let us say that those dolls will reach their chosen initial character position in six seconds from the start of the dolls' motion, since it is close to the edge of the arena but not quite there. Take out half a second for the client-server round trip and all of the latency internal to the player's client and the server. Say they have to hold down their movement key / joystick / mouse buttons for a quarter of a second to get the movement distance they eventually decide on, and leave another quarter of a second for their raw, physiological, non-decision-making reaction time. So they have five seconds to get all the visual information they need and make their decision on where to move.

They have four dolls left to check from the two sides they were forced to skip earlier, in two groups that are almost across the arena from each other. They must locate each of those dolls precisely to determine its path precisely, or else risk having to look again later on to update their path prediction. Since the dolls are already in motion, they do not naturally know exactly where to look, and have to get that from their peripheral vision. Oh, and there is another overlapping mechanic starting during this time, the details of which will inform their decision even if they know the battle timeline completely. Since the vast majority of players never attain perfect timeline knowledge of Expert Roulette dungeons, and even for those who do their earlier experiences are important, we must take it that they do not have that timeline knowledge. So allot at least two eye movements for the overlapping mechanic: one to the castbar for a general idea of when the mechanic will begin, and one for either the cast name or an arena tell that draws their eye. That makes for six eye movements before the player can make their decision.

Suppose that the player can make 1.5 eye movements a second, including the two larger movements to the other group of dolls and then to the castbar, while stopping on each object for long enough to draw the needed information from it. (Recall from my analysis of Greatest Labyrinth that this speed is comfortably above the requirements of Ultimate.) This would allow them to complete the six eye movements in four seconds, leaving them one second to make their decision on where to move their character. That is tight for a mechanical overlap, and will draw heavily on the player's previously learned tactics for general dodging, but it is manageable. Indeed, this case is how I handle the first part of this mechanic.

But that is not the worst case.

If the player is here for their first-time completion and looking around their character for when the nearby dolls pass by, that can easily consume the entire second of decision-making time. They are then forced to either try less reliable methods (such as looking for spaces between the moving dolls or tracking them in peripheral vision) for the sake of speed, knowingly work off of insufficient information, or not even notice some component of the ongoing mechanics and get caught out by it. If the player does not choose to start with their character near a corner, and instead tries to work in the middle of the arena for melee uptime, that more than consumes the entire second of decision-making time, leaving them the same set of options as the first-time player. If the player reflexively looks at the overlapping mechanic when it is presented, that may throw off their entire plan for checking the dolls.

And that is just for the first march. The later marches are worse.

Since the overlapping mechanic will be failed by the player keeping their character in the center of the arena for the doll baits, to succeed at the combination the player must have their character at least moderately close to the edge of the arena during that period. Because the dolls' turning points are fairly evenly spaced along the edge of the arena, this guarantees that some of the dolls will be near the player's character when they choose player characters to target. And the closer the player's character is to the center, buying them time to check the closest dolls, the less time they have for the furthest dolls.

But for the later marches, we do not need to consider many cases on how far the player character is from the center of the arena and how the player distributes their eye movements accordingly. To succeed they must check all 16 dolls, and to look at each doll individually with the necessary stops at 1.5 eye movements a second would take 10 and two-thirds seconds - long enough for all of the dolls to more than cross the arena. Regardless of the player's chosen order, it is not a realistic expectation for them to get detailed path information on all the dolls.

So the player is forced into tracking the broad movements of the dolls using the motion sensitivity of their peripheral vision, only looking directly at dolls that they think will come close to a specific spot they care about, usually their character's feet. (This is a natural strategy for bullet hell games / levels that do not provide strong patterns for the player to plan around.) In turn, this means that they select a doll to look at only after either it has moved far enough from its turning point for them to pick up its motion towards their chosen spot in their peripheral vision, or it has moved close enough to where they are looking that they see it specifically in their not-quite-central vision. The former case is reactive enough, since that means the doll has already covered a substantial part of the distance towards the player's chosen spot, but the density of the dolls in the arena prevents the player from reliably selecting a safespot based purely on their peripheral vision. They must either wait for the latter case or move their eyes away from their chosen spot to look directly at the doll.

If the player continues looking at their character's feet and waits for the doll to enter their not-quite-central vision, they do not have enough time to decide where to move their character on that basis - remember, each doll is visually behind its true server position. All of the components of a successful dodge that I described for the first march still apply, with much less time and the other dolls demoted to peripheral tracking. If the player looks at the doll before that, they give themselves enough time to make a decision regarding that doll, at the expense of looking away from their character's feet and potentially missing another doll entirely in that time.

Oh, and some part of the arena will be visually blocked by the boss itself. And the player has to handle the overlapping mechanic at the same time, which is a further visual and decision-making task.

The dolls really are reactive. There is not enough time for the player to get the visual information they need and make their decision without pressuring their reaction time, if they get that information at all. Break down the time taken for each part of a mechanic before you lock in the design, let alone release it, so that you discover these things in the design stage when you can fix them easily. Err on the side of breaking down even things that seem very similar to but not exactly the same as existing mechanics. Have a testing art library with varying levels of contrast and visual clarity, so you can have early internal playtests that flag where major changes are needed.

Sixth, the falling head dropped one player character being hit by a doll can also hit other nearby player characters. So not only does the player have to keep track of the dolls and the overlapping mechanic to reliably succeed, but they also have to keep track of any nearby player characters or the falling heads generated by other players' mechanic failures. Not only is this a further visual burden, but the falling heads blur into invisibility against the arena floor under the visual burden of tracking the dolls. So the player gets unfairly penalized by a falling head they did not see coming from a doll their character did not get hit by.

In The Copied Factory, the pink orbs also explode onto all nearby player characters when touched, but at least the player has the chance to see them coming and stay far away from all the orbs. (Which is visually easier than tracking all player characters near each nearby orb.) And even there, it is still annoying to successfully make a close dodge for uptime reasons and be penalized by someone else exploding the orb.

Seventh, the penalty for the player's character getting hit by a doll is particularly attention-getting and unpleasant, disrupting the player's participation in the battle. The initial bind forces the player to watch as their character receives further damage and Benoggined, then the Out of the Action and Heavy restrictions prevent them from doing anything but slowed character movement for 10 seconds, and on top of that they are shamed with a doll head and potentially battle dialogue. (And remember, the player can wind up with the latter two penalties even if their character avoided all the dolls!) The Vulnerability Up may last longer, but it is not the annoying part of the penalty even for the healer.

If this mechanic were reliably doable, the sharp penalty would be fine. One-hit KOs are similarly disruptive, and have been used without issue in regular duties for relatively easy mechanics. (Along with their use in high-end duties, gradually rising from the moderately easy mechanics to just about any mechanic as the duty difficulty increases.) But because the mechanic is not reliably doable, such a harsh penalty adds to the annoyance of it. You would have gotten a fair amount of complaint even about The Copied Factory's orbs if their penalty was this harsh.

Eighth, the nature of this penalty easily cascades into further mechanical failures and difficulties for the party. The bind often results in the player's character getting hit by the overlapping mechanic, taking further damage and Vulnerability Up stacks even if they are not KOed. The Heavy from Benoggined makes it more difficult to dodge further hits. And the player cannot help their party members with mitigation, healing, and/or raises while their character is Benoggined, denying them the opportunity to help a struggling fellow player.

(Yes, Benoggined does not stack, but that does not really mitigate the cascading nature of the penalty - one-hit KOs would result in later dolls running right over the player character too.)

Ninth, the nature of the penalty makes the battle drag on, since Benoggined directly prevents the player character from damaging the boss and cascading failures can indirectly reduce the party's damage output. As a result, the more the player struggles with this mechanic, the longer the battle will take for them on average, and the longer they will have to deal with the annoyance. Further, this encourages their party members to become annoyed with them for extending the battle, which causes social friction over the player's physical limits they can do nothing about.

Salvaging this mechanic would require solving all these problems, which in turn would require a complete redesign of the dolls and the whole boss, as the doll marches are the core mechanic of this boss as the doll king. As I have little to go on for what the design intention of this mechanic is, I will instead compare to existing moving object mechanics that function well, and remark upon why they do so.

In The Grand Cosmos, the five brooms move slowly, making the player's screen relatively accurate to their actual positions. They are large with moderate contrast, combined with tiles providing light lane markings on the arena floor, allowing the player to use peripheral vision to locate each one and match it to its lane. (The arena is exactly 20 tiles deep, divided into five lanes. And since the brooms also alternate, the player can count the number of brooms to a side once they notice that. Indeed, the brooms remain on the enmity list for the entire battle, reinforcing their permanence.) The player is given ample time to see both the brooms and the dirt puddles (which cause brooms passing over them to pause and sweep up) before the broom movement begins, which is all the information the player needs to know the brooms' paths and timing. The player can react if they so choose, but they have the opportunity to plan - if they get caught by surprise, it is a fair surprise arising from their lack of planning, and they can change that next time. (Indeed, working out uptime strategies around the brooms' movement serves as an enjoyable opportunity for skill expression. The whole arena is covered each time the brooms move, so the player must move their character to succeed, but can plan around when to do so.) Overlapping mechanics do not create unreasonable combined visual demands. And while the penalty for failing a broom is an attention-getting sweep, it is brief and allows the player to continue fully participating in the battle. That is a fun mechanic that held up to the repetition of Expert Roulette while being gentle enough for a dungeon required by MSQ.

In all iterations of The Jade Stoa - Normal, Extreme, and now Unreal - Byakko uses Unrelenting Anguish, where he stands in the middle of the arena and emits moving orbs from himself. There are no lanes on the floor or permanent brooms to aid the player, and the orbs move at a moderate speed. But his orbs are in a fixed spiral pattern with widely spaced arms, even using two arms rather than three in the first phase of the Normal to gently introduce the mechanic. Combined with the high contrast of the orbs against the floor, this allows the player to dodge entirely by peripheral vision in the wide-open spaces if they please, which is the melee uptime strategy. For casting uptime, the player may instead keep their character near the edge of the arena, buying them time to look directly at only the oncoming spiral arm and identify a gap to slidecast their character into. Both strategies are fun and remain visually reasonable when combined with overlapping mechanics. And again, the penalty for failure is just damage and a Vulnerability Up stack.

I notice that both these successful examples only require the player to find a moving object's position to know its path, without use of the moving object's facing. Having to look carefully at an object for facing or path information adds a lot to the visual demand. I expect this is why bullet hell games favor high-contrast bullets whose facing is clear from their shape alone, which the dolls are distinctly not, to allow the player to rapidly combine the rough shape with the rough motion in their peripheral vision.

Evil Scheme

A standard "flower" of exaflares from the center of the arena, having all cardinals start before all intercardinals or vice versa. Has been a fine mechanic in other places, but here it is combined with dolls, and that makes it just another annoyance to the player.

Looming Nightmare

Another standard mechanic, chasing AOEs.

These are particularly annoying in combination with the dolls, for three reasons. First, the requirement to keep moving makes it more difficult to visually check for dolls, as the player must check around a path rather than a single spot. Second, the natural way to do this without the AOEs getting in the party's way is to work at the edge of the arena, putting the player's character very close to doll turning points. Third, if the player's character does get hit by a doll, the chasing AOE will catch up during the bind and hit them repeatedly.

Scream

Another standard mechanic, a cone AOE sequence that covers the whole floor in three groups of two cones each. This is the simplest of the mechanics overlapping with the dolls, and is once again ruined by the presence of the dolls. This also starts the sequence of many doll marches from a single Spirited Charge.

Falling Nightmare (Repeat)

This is almost the same mechanic as before, with two rounds and some extra heads randomly placed each round. If the heads had good contrast, the mechanic would be okay but not good for casting planning.

But it comes with the battle dialogue "Apologies for breaking character, but... perhaps it's time you stopped resisting? There are limits to my patience." If the mechanics were enjoyable, this dialogue would be funny. Since at this point the player's patience has been tested repeatedly by the dolls, it is grating.

Overall Impression (Boss)

The use of the dolls is a reactive disaster, and all the other mechanics do not express the lore concept of the doll king. Having a boss express its concept via a complication to otherwise standard mechanics is fine, but the whole combination needs to be reasonable; here, the dolls are not reasonable alone.

Mobs Between First and Second Boss

Two Packs

Nothing mechanically interesting here, but can be double pulled.

Two More Packs

Between these packs, there is a Ferris wheel that falls as the party passes by:

https://i.imgur.com/M3gIIVV.png

Specifically, it falls shortly after the first time a player character gets close enough to it, with just slightly more than enough time for that player character to run by if they have Sprint active.

To the first-time player, this is a minor "gotcha", as they have no real reason to look to the side of the path while focused on the upcoming mobs, and so most likely will not notice the Ferris wheel's presence until the briefly flashed telegraph at the end.

And then once the player knows about this AOE, they are faced with a choice: try to keep their character at the front of the party's charge to get ahead of the Ferris wheel, deliberately wait until it falls, or take the hit on purpose. None of these options are good. Trying to get ahead of the Ferris wheel involves judging the party's relative positions, for which the client-server round trip making the player's screen inaccurate is highly relevant. Waiting draws the player out of the battle, both from the waiting time itself and the delay in their character reaching whichever party members charged ahead. And taking the hit on purpose just feels bad.

Environment features affecting trash pulls can be fun and interesting, but this implementation is not.

Suppose instead that the spinning Ferris wheel was behind the last pack of this section, and slowly destabilized with a sound effect during the party fighting the mobs. Both of these would encourage the player to see the Ferris wheel well before its AOE. And then rather than the AOE covering the whole area, it would leave safespots between the spokes, shown by a rotating telegraph that slowed down as the Ferris wheel slowed and fell over. (This is physically accurate as well; gyroscopes are more stable when they spin faster.) That way, the player would be presented with a combination between the mobs' basic AOEs and the environment feature AOE, which is not too difficult but would help make the pull interesting.

Second Boss - Jack-in-the-Pot

https://i.imgur.com/xwIbuD8.png

Troubling Teacups

Sets up teacups around the arena and ghosts entering them to create large point-blank AOEs. The ghosts' targets are nice and clear, since they each tether to their chosen teacup and add white smoke to the teacup's purple smoke while tethered.

(The first cast has only the outer ring of teacups and one ghost. Subsequent casts have two rings of teapots and two ghosts. All rings have four evenly spaced teacups.)

This is an example of a good setup cast that does not have to resort to active glowing. The tethers are effective despite the multicolored arena floor, and the white smoke contrasts the purple smoke.

Tea Awhirl

Starts the teacups spinning around the arena, with the AOEs coming shortly after the teacups stop. (If two rings of teacups are present, this includes the rings swapping with each other.) Since the ghosts are now concealed in the purple smoke filling the teacups, this is a visual tracking mechanic where the player must track the teacups' movement. When dealing with multiple ghosts, there is also a memory component, remembering the shape of the combined AOE relative to the teacup / point the player chooses to visually track.

In purely angular motion, all the teacups rotate at the same speed, and they all swap in/out the same number of times. These facts are apparent from watching the teacups in peripheral vision a few times while keeping one teacup or one fixed location in central vision, which are the natural tactics for the first-time player. In turn, knowing these facts confirms for the player that either of those visual methods yields an effective solution. They can follow one teacup around the arena and remember the combined AOE shape relative to that, or they can watch one point in the arena to count the teacups that pass through it and the ring swaps. I appreciate this flexibility.

I can do this mechanic just fine, but I do not like the pulses of smoke over the arena as the teacups start spinning. They are not fun or interesting - in fact, I did not really notice them until I looked at videos for this review - and instead feel like part of a neurological test, purely a means to add stress to the player's visual processing. (There is a reason that you use similar techniques to make arena transformations less jarring to the player.) Neurological tests do not generally make good gameplay.

I also wonder if the relatively even spin of the teacups could induce motion sickness in some players. I do not get motion sick from these, but I would not be surprised if other players did. In contrast, the jumping dogs in Eden's Promise: Litany (E10N) and its Savage version (E10S) make it very clear from the animations and the individual jumps that they are moving, not the arena floor.

Other than those two things, this mechanic is fine. Which is good, since it is repeated very often by this boss.

Toiling Teapots

A sequence of expanding circular AOE puddles, themed as teapots filling static cups to overflowing and leaving tea behind. The relative slowness makes the puddles work, since dodging back into the first AOEs results in taking the puddle DOT.

This is also a good choice of mechanic to express the lore concept of the boss.

Overall Impression (Boss)

Nothing special, but it is okay. Just take out the masking smoke pulses.

Mobs Between Second and Third Boss

Two Packs

Nothing mechanically interesting here, but can be double pulled.

Two More Packs, with Large Dolls

This is an interesting double pull, as the two large dolls that form the last pack perform light staggered raidwides on top of the usual amount of basic AOEs. This gently encourages the player to consider their mitigation tools, AOE healing tools, and potentially even strategies such as "kill one doll first to halve the raidwides".

I would not go harsher than this in regular dungeons, since fairly weak healers do have to be able to complete the duty at minimum Item Level without drawing ire from their party members.

Third Boss - Traumerei

https://i.imgur.com/nDcRCOr.png

Bitter Regret

Hits either the middle or the sides of the arena, with a crisp glowing animation tell in Hades's style. Simple on its own, but comes back in sequences later. The edges match lines on the floor nicely for casting optimization.

Poltergeist

A short pure warning cast for the poltergeist tether setup, good to cue the player to prepare for something. Only appears before the first Memorial March.

Memorial March (Walls + Poltergeists)

Places walls in the arena that can be passed through in ghost form. The poltergeists choose their positions, one to each player character and close to an edge of the arena, and appear as the second wall begins to visually drop. The narrow AOEs with clear telegraphs function effectively to nudge the player character definitively into a quadrant of the arena, to make use of the ghost form necessary.

After the poltergeists appear, they begin casting their proximity tether blasts. Even if the player has not picked up on the battle dialogue, they will quickly realize that the walls do not allow them to move their character far enough to stretch the tether by normal means. This is perfect timing for the activation of the ghost/flesh platforms and the (correct) blue text "A section of the floor begins to glow!" The player may not get it right the first time, as going to one of the platforms after this realization leaves them not much time to double back through and well past a wall, but they are pushed towards the information to figure out how to succeed.

(Nice touch that this text is blue rather than red, since it talks about an aid to the player character rather than a warning of danger. Of course you cannot rely on the player seeing the color difference, but they do not need to here so it is just nice.)

This is also the sort of mechanic where one player doing it correctly automatically demonstrates the solution to their party members, which helps the first-time player catch on even though the necessary information is not where they would look when running their character away normally. Such situations allow you to have slightly counterintuitive mechanics in regular duties, though not full puzzles.

It is good that the tethers do not do terribly much damage if not sufficiently stretched, using only moderately increased damage and a Vulnerability Up stack to indicate failure. This ensures that the first-time player does not feel discouraged by the likely failure or two when first learning the mechanic, and allows the later overlaps and combinations to not be too punishing as a whole.

Ghostduster / Fleshbuster

Visually these are spread AOEs, but the cast names and the red texts show that spreading has nothing to do with it. (The texts are "Traumerei prepares a ward against spirits!" and "Traumerei prepares a ward against the living!" respectively. They appear every time.)

If the player character is in the opposite form at the end of the cast, they take no damage at all, complete with the "DODGE" flying text, green checkmark over their head, and "ding" sound effect for the player. If the player character is in the same form at the end of the cast, they are immediately KOed, complete with the "9999999" (true damage) flying text, red X over their head, and another "ding" sound effect for the player. This is a good use of your feedback tools to help the player learn the mechanic, as both possible outcomes are relatively extreme compared to most other mechanics.

(The first of these is always Ghostduster, to teach the player that ghost form is not always preferred. Eventually they randomize.)

The length of the casts is suitably long, allowing the player to move their character to a transformation platform even if their initial position is bad. This is important given the harsh penalty for failure, and done correctly.

The one thing I would change about this mechanic is to use the generic "targeted for something" marker instead of a spread marker. Spreading is not at all relevant to the mechanic - the whole party can stack up and take no damage - and the generic targeting marker is plenty obvious considering its recurring form and its own sound effect.

Memorial March (AOE Sequence)

The boss spawns two sets of ghosts at the front of the arena, both tethered, with the rows alternating between sets. These create line AOEs that resolve in the same order while the boss is casting Bitter Regret. Showing the tethers and giving the player time to see the ghosts during this cast is what makes the sequence work, since it allows them to be looking at the Bitter Regret animation tell instead of the ghost AOEs as they resolve.

Memorial March (Combination)

Combines the walls and poltergeists with something else. Can be a set of ghosts at the front that will hit two of the quadrants at the same time as the proximity tethers, can be Bitter Regret, can be Ghostduster / Fleshbuster. Once the player understands using the ghost form, all of these are manageable and offer different optimization opportunities. (For example, with the ghosts hitting two quadrants, the player can keep their character in melee range and succeed at the combination by starting their character in front of the ghosts, so that their poltergeist spawns near the ghosts and they can run their character away from both.) This is very welcome.

The pace would be fast if not for introducing each element individually first, but it works since they are introduced that way.

Ghostcrusher

A line stack while the walls are up, which encourages the party to run into the walls to stack. This works, which is good. Oddly given the name, it does not penalize characters in ghost form, but the length of the cast is not long enough to do that without being annoying.

Overall Impression (Boss)

A fun boss that could easily have been in a Stormblood or Shadowbringers dungeon. The opportunity to strategize around the ghost/flesh transformation is very welcome in Expert Roulette.

Overall Impression (Dungeon)

The dolls ruin this dungeon. This is the prime example of "it only takes one bad mechanic to ruin a duty", as there really is only one bad mechanic in here - but the one mechanic is so bad that even young and able-bodied players get annoyed by it easily.

----------------

Both of your "experimental" dungeons are plagued by unworkable visual demands. Please for the sweet love of Hydaelyn, adopt and follow visual clarity and accessibility standards that account for the time pressure of mechanics.

I will now continue my review of the design of the Patch 7.0 duties, in this subsection covering the Arcadion Light-heavyweight Normal Raids.

The same considerations apply here as for the optional Expert Roulette dungeons in the previous subsection. These are still endgame duties that must welcome the casual player. They should be a touch harder, which works out because the party is guaranteed to have at least two members who can raise, but the biggest difference design-wise is that the longer time with a single boss allows for building up a concept over more mechanics.

----------------

AAC Light-heavyweight M1 (M1N, Black Cat Normal)

https://i.imgur.com/pRHutey.png

Before getting into specific mechanics, this is a very well-patterned floor. There are major grid marks that get used later even in Normal, minor grid marks, and diamonds in the center of each floor section.

The only tweak I would make is turning the brightness down slightly, and not even because of the floor itself - because waymarks and the Savage clone tethers are forced to be very bright by the floor brightness, and that makes them uncomfortable to look at for the extended periods that come from progression and less-than-ideal weekly reclear parties. (Sometimes even after changing hardware settings. My monitor is at 0% brightness already. Only this duty and its Savage have this issue out of every duty in Patch 7.0, so it is clearly brighter than the rest of the game.)

Since it is easier as a developer to make one visual asset and reuse it in both the regular and high-end versions of a duty than to make two, and it is more welcoming and immersive as a player to see familiar visuals from the regular version when learning the high-end version, in cases like this it makes more sense for me to comment on the Normal version about changes that are for the Savage version.

Bloody Scratch

First used as a simple opening raidwide, unlike in the Savage version. Pitching the casual player directly into a dodging mechanic could be fine if it were a very simple one, but the standard opening raidwide serves as a gentle warm-up here.

One-two Paw

The cast name cues the player to expect a sequence, the cast length is good, and the glowing elongated claws are very clear about the order of which sides will be hit. (Unlike the various one-two punches in the MSQ duties, this one always alternates sides.)

I would tweak this to hit with the Savage timing, with the first hit right as the castbar ends, rather than the current Normal timing that hits a couple of seconds after the castbar ends. While that would make the mechanic harder in theory, it would be more consistent with the general snapshotting rules of the game, making it easier in practice both here and for stepping up to Savage. The casual player would still have plenty of time to see the claw animations, especially since the boss shows it twice before doing anything else.

Black Cat Crossing

A fine sequence dodge, which works thematically even for the casual player who never sees its Savage upgrade (Quadruple Crossing). The glowing lines do not tell the first-time player exactly where will be hit, but that is completely fine - the natural guess is "these extend from the boss" and the cones confirm that as the first cast resolves. On later casts, strong players can use the floor grid to dodge precisely.

Mouser Mini-Phases

Mouser

A big flashy sequence that draws attention to the gradual transformation of the arena, which is excellent to prepare the player for manipulating the ring themselves during Elevate and Eviscerate. Since the player can avoid all damage here by planning one movement or finding one square that does not get hit, this is an acceptable memory mechanic with plenty of time in the resolution for the player to see the effects on the arena.

The "My riing!" voiced reaction from Metem is a nice touch for story and immersion, though it does not do much to make the mechanic work since it occurs after the player has most likely seen the arena transformation.

Elevate and Eviscerate

Here, the "Going up!" voice line comes at the start of the first cast, so the player can use it to guess at what the mechanic does.

Incredible attention to detail in the visualization of this mechanic. The visual feedback during the cast is what makes such a complex task doable in Normal. (To avoid a KO, the player must manage the direction and the distance of the knock-up to land on a full-strength tile. To know to do that, they must understand that the knock-up will also move their character horizontally and that their character's landing will affect the floor. That is a lot to do, especially since they cannot be expected to learn the distance of the knock-up in Duty Finder.) All of the necessary pieces are incorporated in the visualization, and in a very clear way for the first-time player.

The arc of arrows in the visualization shows what the direction and distance of the knock-up will be if the targeted player character stands still, along with the fact that it is a knock-up rather than a knockback, and even includes a generic humanoid model travelling along the arc. If the player character would land safely on a full-strength tile if they stand still, that whole tile is highlighted green with a vertical "bounce" indicator that signals the impact on the floor. If the player character would not land safely on a tile if they stand still, the whole arc of arrows turns red and ends on either no tile or a partly-broken tile with a pure red highlight, no "bounce". (The fact that there is more than just a color indicator for safe / unsafe is awesome; so much to show and still colorblind-friendly.) And while the source of the knock-up varies - in the first of these mini-phases it is the boss, in later iterations it is a clone - there is always a tether between the targeted player character and the boss / clone. That tether leads to the boss / clone continuously facing that character, showing the player where the knock-up is coming from and therefore how to manipulate the direction. Plus all of this is completely matched up with the player character's movement on the player's screen, so it has to be client-side for rapid feedback. So many visual elements that come together perfectly, enabling the targeted player to control their character's landing, the rest of the party to react to where that character will land before the cast finishes, and the whole party to learn the mechanic.

The only minor pain point of this mechanic is that since the boss's position for the next cast in sequence is set by it dashing to the previously targeted player character, occasionally the boss winds up positioned relative to the remaining full-strength tiles such that there is no place (or only an impractically small place) to move their character and avoid the KO. It is not common, but it is very frustrating when it does happen, since the player who takes the penalty has nothing actionable to do about it. I would fix this by making it more similar to the Savage version, where the direction of the knock-up is the targeted character's facing and therefore fully under the targeted player's control. (This would not require the clone to appear in the first of these mini-phases, just having the boss dash behind the targeted player character.)

Future arena manipulation mechanics like this would still need to only target one player character at a time in Duty Finder, to avoid adjusting loops and to allow the casual player time to learn from their party and breathe. Since this is only one player character at a time, it works well. The selection order even has every player character be targeted once over the first two occurences of this mini-phase, just like Savage, where reasonably possible.

There are small forced downtimes from the knock-up itself, which are an optimization opportunity for stronger players since the boss gives ample warning when the downtime is coming. There are also melee downtimes in the first iteration from the boss turning and dashing, which are not really optimization opportunities because of the reaction to the targeted player's choices.

Overshadow

This simple line stack only appears in the second and later Mouser mini-phases, with the boss casting it as the clone casts Elevate and Eviscerate. This placement is good, being just the right amount of added mechanic for the player to handle once they have seen a few rounds of Elevate and Eviscerate. In particular, the player whose character is targeted for Elevate and Eviscerate can still handle that by looking at the visualization, and the rest of the party can react to each other by looking at the stack. (Spreads would not work here, being a little too much reacting to each other when combined with the Elevate and Eviscerate targeting.)

Shockwave

It is correct and important that this knockback can be negated by Arm's Length / Surecast in Normal, as broken tiles may well leave no safe place to ride the knockback and a party KO from that would be an annoying surprise.

Regular Arena Mini-Phases

Predaceous Pounce

This mechanic is fine but nothing special. It is thematically appropriate, with the animations of the resolution being very fast, but on a purely mechanical level it feels like the sort of thing that is put in a Normal to replace a Savage mechanic that cannot be usefully adapted for casual play. (Being so fast, following the boss around is a dangerous pain in the first usage, which is done by the boss rather than a clone.)

The follow-up spreads (no castbar) ask the party to not stay clumped up in their final safespots or around the boss, which is again fine. They give lots of time.

Leaping One-two Paw

A good introduction to the "boss jumps and immediately does a mechanic" concept of the Leaping casts. The tether visually leads the player to the glowing pin, which provides not just a clear indicator of direction to the first-time player but also a precise reference for exactly where the boss will jump. In turn, that reference enables the player to precisely dodge when seeing this mechanic again, including for the spread follow-up on the second and subsequent casts.

Just like the base version, I would give this the Savage timing.

Leaping Black Cat Crossing

Another good Leaping cast, which is correctly the second use of the concept as dodging its pattern requires the player to have a better grasp of the boss's landing position. The second and subsequent casts of this again have the spread follow-up, which is more complicated here due to the narrower safespots.

Overall Impression

Fantastic core mechanics, made possible and elevated by incredible visual design. Just turn the brightness down a little and make Elevate and Eviscerate use character facing. That I got bored quickly here is entirely the fault of job design, not duty design; any duty that could challenge me in a fun way once I know it would be functionally impossible for the Duty Finder population, and the current job design overall is not up to the task of filling in the gap.

--------

AAC Light-heavyweight M2 (M2N, Honey B. Lovely Normal)

https://i.imgur.com/i7ludwW.png

This arena also has an excellent floor, but the mechanics only utilize it in Savage so it is better discussed there.

Tempting Twist, Honey Beeline (Basic Forms)

The visual effect for each is important to give the first-time player a rough idea of the danger zone. (Tempting Twist is always first.) These are fine.

Honeyed Breeze

I find it very strange that this double cone tankbuster is not given the Savage cast name for the almost the same thing, Stinging Slash. (The only difference between Honeyed Breeze and Stinging Slash, aside from the obviously higher damage in Savage, is that Stinging Slash has much wider cones of roughly 90 degrees each. Which would be perfectly fine for Duty Finder, since the player can learn to give the tankbuster a wide berth from seeing the animation.)

Even if the design intention here is to not present Killer Sting (the tank stack from Savage) to prevent one tank from taking it alone with an invulnerability, matching cast names between regular duties and their high-end versions help the player looking to learn the high-end version.

Beat Mini-Phases (Charm Gauges)

Honey B. Live

The longer cast and arena-wide animation of this raidwide are an effective warning that the boss is setting up something. A good primer for the player to notice the overhead charm gauges when they appear.

(The "Beat" name for these mini-phases is from Savage, where each of these casts is numbered in the cast name.)

The heart stack debuffs shown by the charm gauge do come with a light DOT per heart, which is healable by even a single casual healer so it is fine.

Love Me Tender (Spreads)

The very first cast of Beat 1, showing the player that attacks with hearts on the telegraphs also inflict a heart on their character when taken. It is correct and important to do this, since the first-time player does not know how the charm gauge works. The red text after this hits for the first time, "You are growing infatuated with Honey B. Lovely...", helps reinforce the lesson.

(no castbar, follow-up effect of spread / stack Love Me Tender) Heart Towers

(In Beat 1, four towers spawn. In subsequent Beats, six towers spawn.)

One of the worst mechanics of Patch 7.0. Let us examine why.

Now that the player has seen the heart spreads and their effects, they understand that taking an attack with a heart on it adds a heart to their charm gauge. They are trying to further learn how the charm gauges work, and most likely infer correctly both that taking a tower will give their character a heart and that the party leaving a tower alone will cause it to explode and inflict a heart on every party member. From there, they understand that the party as a whole will get the best results by placing exactly one player character in each tower.

Even the very casual player understands both what the heart towers are and the intended resolution, since the necessary information has been put right in front of them. (Plus the casual first-time player almost always has the benefit of the story up to that point warning them to watch out for charm-related mechanics - cutscene skippers have no need to be in this duty other than wanting the gameplay of the Normal Raids, the unlock for the Savage counterparts, or gear for other duties, so if they are in this duty they usually have decent gameplay skills.) Player understanding is not a problem here.

From that understanding, how does the player try to do the mechanic?

This is a regular duty, with the party probably being formed through Duty Finder matchmaking. The player is operating on reaction. When they see a tower and their character is not already standing in it, their first reaction is to move their character towards the tower unless they have a reason to keep their character elsewhere (such as already standing in another tower). Their second reaction is to try to redistribute between towers if any are left empty after the party first moves. Accordingly, the Duty Finder population handles towers by reacting to their fellow players' choices.

With normal towers, this is not a significant issue, with only the occasional failed tower. Multiple players will put their characters in the same towers, and leave them there because that resolution of the mechanic does no real harm to the party's chances of success. With the heart towers, multiple players seeing their characters all in the same tower will react by moving their characters out, trying to get to exactly one character in each. That makes the tower empty, so they react by putting their characters back in...

The first problem with the heart towers is that they actively create adjustment loops. Not just invite adjustment loops, actively create them. If the player cautiously waits for a party member to take a heart tower, they are still reacting to their fellow players' choices, and that strategy is equally prone to adjustment loops. If the player prepositions their character near a potential heart tower location to indicate intent to take that tower, that gives them longer to try to break out of an adjustment loop but does not make the loops any less common.

As a result, the Duty Finder population has a very high failure rate for this mechanic, and will never be able to master it short of spending tightly limited general preparation time on a priority order of taking towers. (Which I would advise against, since there are much more flexible and interesting things to use that time on if you were going to teach Duty Finder to do general preparation again.) What happens when the party fails and a tower explodes?

Well, the tower explosion inflicts a heart on every player character who is not KOed or charmed (the Fatal Attraction debuff). And since the towers always come after a spread or a pair of stacks in the same Beat, every player character already has at least one heart unless they actively avoided a stack earlier. So those player characters who have not yet taken a tower go to two hearts, and are able to continue fighting. Those player characters who have already taken a tower are forced to three hearts, inflicting Fatal Attraction followed by the Sustained Damage debuff (a DOT that may very well KO them when combined with the damage of a failed mechanic), so they cannot continue fighting immediately at a minimum. And there are Vulnerability Up stacks added on top of that!

The second problem with the heart towers is that failure inflicts the heaviest penalties on those players who did a service to the party by taking earlier towers. This discourages the player from taking a heart tower at all, compounding the issue of adjustment loops.

Once one tower explodes, the only player characters who can take the remaining towers without being charmed are those who either were already charmed or avoided a stack earlier in that Beat. (In particular, those about to resolve a tower can be pushed to two hearts by the explosion and then three hearts by the tower they were already standing in.) This makes the mechanic create cascading KOs very easily, which is not inherently a fatal flaw but compounds the annoyance and anger with the mechanic. The serious penalties for lack of preparation are much more than occasional here.

And if the party makes it through the series of towers, they are then presented with a reaction dodging sequence of AOEs that all also inflict hearts. After seeing that, the player is further discouraged from taking a heart tower (or the stacks preceding the heart towers in subsequent Beats) if they have any concern about their ability to manage the reaction sequence later in that Beat - especially since being charmed during that time often results in taking further AOEs while unable to act.

The third problem with the heart towers is that the surrounding mechanics encourage selfish play, while the towers make the party's charm gauges a communal resource. If there were no stacks, towers, or even spreads, and avoiding a filled charm gauge was purely a matter of individual dodging, that would be one thing. (In fact, using a similar concept to present reaction dodging sequences without seriously penalizing the player until they reach three failures in the sequence would make such sequences a lot more tolerable, functioning as a gentler Thrice-Come Ruin since party members could heal a struggling player's character through it.) But those are not the mechanics here. The party's charm gauges are clearly a party resource, while the mechanics encourage the party to fight each other over it.

The first problem alone is enough to completely scupper the mechanic. All three together are a disaster. The Normal Beats in their current forms were knowably doomed to be anti-fun before a single party saw them.

Surprisingly for a mechanic that drew so much complaint and has so many fatal flaws, part of the underlying notion of heart towers is usable even in Duty Finder. But making it work well with the current Duty Finder population requires changing the whole structure of the duty timeline. For that, see my redesign of this duty and its Savage version.

Love Me Tender (Reaction Sequence)

Moving hearts emitted from the boss and from the edges of the arena all at once, combined with both small and large circle AOEs rapidly appearing, creating a bullet hell section with the added danger of filling the charm gauge. (The moving hearts from the boss are in a hidden fixed pattern, but the first-time player does not know that, and the sequence as a whole is so chaotic that even a strong player who completes the Savage may not notice it.) The hearts from the edges of the arena start in random places and each pick a random player character to aim their (straight line) path through, practically guaranteeing that the player will have to dodge at least one in a panic. And the pink circle AOEs very effectively hide the pink moving hearts from peripheral vision.

On its own, this is visually overwhelming chaos. Combined with the heart towers, it is an invitation to anger. The only thing the player can do about struggling with this is try to avoid hearts from the towers and stacks.

Blow Kiss

A wide cone AOE targeted on a random player character's position at the start of the cast. Annoying when there are still moving hearts in the arena, acceptable otherwise. The cast name is odd, since blown kisses are not avoidable on Earth, but that is relatively minor.

Love Me Tender (Fixed Circle Pattern)

This is acceptable, since even if the player does not remember the "Shiva circles" pattern from previous duties, they can see the sequence of telegraphs appear and have enough time to plan their movement.

Love Me Tender (Heart Stacks, used instead of the spreads after Beat 1)

Each of these two stacks inflicts a minimum of four hearts distributed among the player characters within, but also will inflict at least one heart on each player character within - so each stack will give more than four hearts every time that more than four player characters stand in it.

The arrows around the stack circle are yellow just like normal stacks and they have the same sound cue, being distinguished only by surrounding the arrow pointing to the targeted player character with four floating hearts. So the first-time player might not even notice the difference from a normal stack, being so familiar with the standard stack marker as to not need to look at every part of the visual, and having long experience that keeping track of the stack locations is important when dealing with more than one simultaneous stack. Even if the player does see the hearts and count out the four of them on a stack, whether on sight or in later attempts at the duty, that says nothing about the extra hearts generated by having more than four player characters in a stack.

(Even if the stacks each gave exactly four hearts, that would be a difficult concept to visually convey, since too much difference from the normal stack marker would stop the player from reliably recognizing it as a stack. The added hearts for extra player characters makes it even worse to convey. You could consider having the four hearts instead be part of the four arrows around the stack circle and drawing the entire telegraph in pink, but that still would not reliably have the player see all four arrows.)

The player most likely understands the preexisting concept of multiple simultaneous stacks; not only have they seen two or more simultaneous stacks in prior duties, but it is the logical guess that they are meant to be separated, since if you wanted the whole party to stack together you could present one stack marker. (Though Syrcus Tower does not help matters now that it has the standard stack marker, since those ones can and should be stacked.) From there, it is natural to try to split the whole party evenly between the presented stack markers.

But the Duty Finder population has not mastered that either, as evidenced by the regular individual KOs and even party KOs to two simultaneous stacks in duties such as E7N. Just like the basic spread, splitting the party into two even (or roughly even) groups by reaction requires reacting to fellow players' choices - so the Duty Finder population can never master it so long as they are operating on reaction. (If their general preparation included light parties and there was a community standard flex order within each light party, they could, but that is a far cry from the current player expectations.) Even separating the two stacks is the same way.

Using hearts instead of short-term damage type vulnerabilities to enforce splitting the two stacks does not change the fundamental nature of the task. The Duty Finder population still can never master it so long as they are operating on reaction.

If this were the only mechanic in a Beat that caused inherent problems to Duty Finder, the use of hearts could be a benefit to the mechanic, allowing the party some "free hits" from imperfect coordination without taking any severe penalties. (Indeed, the sequence of these two stacks into six heart towers only mandates that the party take 14 hearts out of their collective maximum of 16 hearts before player characters begin being charmed, allowing for two "free hits".) But in the context of the duty, this is always followed by the heart towers and then the reaction sequence, so the party must both stretch that lenience across the whole Beat and compete over it for the reaction sequence. As a result, the presence of these stacks sets the party up for annoying failure, both in encouraging selfish play and in potentially having too many collective hearts to avoid charms from the towers without making any real mistakes.

Honey B. Finale

A raidwide that transforms the arena back and cleanses all hearts. The damage is only scary because of the high chance of failed heart mechanics earlier.

Regular Arena Mini-Phases

Splash of Venom

This stores a spread for after the upgraded random ground AOEs, though the player does not need to remember that here since it comes with telegraphs. It serves a fine dual purpose of ramping up the duty in a way the casual player can handle while showing what the cast will do in Savage.

Tempting Twist, Honey Beeline (Upgraded Forms)

Now these ground AOEs leave Splinter honey blobs that explode afterwards. Oddly, these are not in the Savage patterns, which is another missed opportunity to help the player learn the Savage version should they choose to. The Savage patterns are still just ground AOEs that the player can learn by seeing them.

Drop of Venom

This stores a stack for after the upgraded random ground AOEs, again with telegraphs. This cast does not and cannot teach its Savage version, since that is partner stacks and partnering by reaction to fellow players is nowhere near reliable.

Alarm Pheromones

This mini-phase is better discussed by the patterns that appear in its different iterations, rather than by cast names, since it has the same cast name every time.

Over all the variations, the visual design of this mini-phase is a miss in both thematic appeal and mechanical fairness. Surely the Groupbees' line tells should be black instead of the current green, perfectly fitting the bee theme and providing high contrast against the yellow-brown honeycomb floor? And the line tell for each Groupbee should extend across its entire path, clarifying that they dash across the arena rather than stopping within it.

Helix Pattern (Alarm Pheromones 1)

A good introduction to the Groupbees' behavior, allowing the first-time player to see them dash one by one while presenting the strong player with an uptime optimization opportunity for tanks, melee DPS, and jobs with frequent cast times.

Bowtie Pattern (Alarm Pheromones 1 and 3)

In Alarm Pheromones 1, this directly follows the helix pattern on its own. In Alarm Pheromones 3, this goes off while the boss is casting the upgraded Honey Beeline, most likely creating a sequence dodge from the main tank dodging the Groupbees and thus pointing the front-to-back line of the boss into the safespot of this pattern. (There is no repeated cast separating Alarm Pheromones 2 and 3, just a Tempting Twist into spread resolution into Drop of Venom.) Both are fine.

If there was another use of simultaneous Groupbee dashes, the first occurence of this pattern would serve as a good introduction to it, but this is the only simultaneous Groupbee pattern in the Normal. In Savage, such introductions are less necessary, as the player has already seen the basic concept of the Groupbees from the Normal and is inclined to study specific mechanics for their patterns anyway.

Random Baits (Alarm Pheromones 2)

The annoying pattern, where the Groupbees appear one at a time in random positions and each set their paths to go through random player characters, with no advance indication of the spawn order or who is baiting which Groupbee.

Here, the player does not strictly need to react to their fellow players' choices, only to the Groupbee paths. So why is this such a troublesome mechanic?

Consider where the player must look on their screen to find out whether a particular spot, such as under their character's feet, is safe.

If they work in the middle of the arena, the Groupbee lines do extend that far but they are heavily exposed to most of the Groupbee dashes, and must react frequently - including potentially being forced out of the middle by multiple dashes in that area. Thus they must also contend with the other case below.

If they work near the edge of the arena, they cannot just look at the spot they are working with, since the Groupbee lines do not extend that far. Nor can they rely on the baiting of a Groupbee towards their intended spot to draw their eye, because the green on yellow is low contrast even for those without colorblindness. Instead they must check a substantial radius around their intended location, and they must do so by moving their eyes around that radius so as to catch the Groupbee lines in the rather small area of their central vision. Even though the Groupbees' Blinding Love cast is 6.5 seconds long, subtracting out the time for the Groupbee line to appear, the player to make a decision about where to move their character accordingly, their physical reaction time, and of course client-server latency leaves substantially less time than that - in which they must scan that entire radius, requiring uncomfortably rapid eye movement even if they succeed. And just like in Tender Valley, that rapid eye movement creates effective blurring.

(Does the player need to stop their eyes on each Groupbee line? It depends. A relatively isolated line will generally be discernable without a stop, since the shape shows the direction. If two or more lines cross near the player's scan circle, they may need to stop and pick out whether either one will pass through their chosen spot.)

In either case, if their character does get hit by a Groupbee, their character is knocked along that path, putting their character in a new location that they presumably have not been able to scan the appropriate radius around. (Or into the death wall at the edge of the arena.) Since such a failure necessarily occurs during the Groupbee dashes, they have even less time to try to scan the appropriate radius around their character's new location, so failures cascade rapidly on an individual level. In turn, this leaves fewer player characters for the remaining Groupbees to target, putting more pressure on the remaining players and thus making failures more likely to cascade on a party level.

And on top of it all, the pace of the Groupbee spawns makes this a reactive mechanic given the nature of all the dashes requiring the player to consider where to move their character.

If the visual design for the Groupbee lines allowed the player to get the strictly necessary information by looking just at their character's feet, this mechanic would still be reactive and unpleasant, but it would be a lot more manageable. The lacking visual design turns the mechanic from merely reactive into a fiasco.

Overall Impression

Where The Strayborough Deadwalk is ruined by one bad mechanic, M2N is ruined by fundamental misunderstandings of how to use randomness and what the player can do by operating on reaction. It is a shame that such a strong lore concept led to such an annoying duty, which is one of the reasons that I chose this and its Savage version for my redesign examples.

--------

AAC Light-heavyweight M3 (M3N, Brute Bomber Normal)

https://i.imgur.com/cd7jYpP.png

You clearly put in effort to give this floor markings, even though this is meant to be a wrestling ring and most wrestling rings have no such internal markers. You did a good job of making the markings look thematically appropriate. Yet the floor geometry does not match any of the mechanics that would benefit from them, such as Lariat Combo and the exploding bombs. (It does not match any mechanic here in Normal, and the one Savage mechanic it matches cleanly has no need of floor geometry, since it comes with full telegraphs.) What an odd choice, and one that makes multiple mechanics needlessly annoying.

Opening Phase

Everything before Doping Draught ramps up the attacks, though the first-time player does not know that yet.

Brutal Impact

Initially a two-hit raidwide, it ramps up to six hits after Doping Draught. The initial form is so light that I would be inclined to give it four hits like the Savage version, especially since its only use is right at the beginning of the duty and so allows for prepull shields.

Knuckle Sandwich

A tank stack, like in Savage, but with only a single heavy hit.

Oddly, this does not ramp up after Doping Draught at all; I would be inclined to turn down the power of each hit both before and after, then give it just one extra hit when ramped up, to gently hint at the Savage multi-hit forms without being too deadly if the off tank is unavailable or runs away after the first hit.

Brutal Lariat

The boss dashing to a side of the arena, raised arm, and fire animation make it clear that the attack will hit that side of the arena, even for the player who does not know that a lariat is a real category of wrestling moves on Earth that all involve charging and swinging an arm at one's opponent. The name is a nice touch, very thematically appropriate.

Unfortunately, even once the player has seen the attack resolve, they have no good floor references for how far they must move their character to be safe. They can try to estimate it based on the circles, but that is not reliable when the pace starts ramping up.

Murderous Mist

The green poison mist and bubbles hint at the 270 degree AOE shape where only the back of the boss is safe, sometimes called Sweep the Leg from its appearance in The Jade Stoa (all difficulties) or Conflag Strike from E6N and E6S. But the mist and bubbles do not do a good job of showing the shape.

First, they are mostly high up, so on the player's camera they will show over a significantly different part of the floor than the attack actually hits. (A true top-down view is not possible from the game's camera in battle without third-party tools.) Second, their sparseness and high variation makes them mostly visual clutter, especially since they do not show an edge.

The first-time player could easily think this is a DOT raidwide, since that is what the mist and bubbles look like. Even the boss leaning back with arms spread could equally be interpreted that way. The player will learn the AOE shape, but by the brief telegraph and animation of the attack resolution rather than the visual effects during its cast.

Because of this, the combined penalty of high damage, presumably heavy DOT, and Vulnerability Up stack are a bit harsh even though the DOT can be cleansed with Esuna. It is not inherently a one-hit KO, but it becomes one quickly unless the party notices and applies healing resources, which is too much for a mechanic where the player can correctly prepare for what they see and still fail.

Since the DOT nature is the core concept of this attack, and important for Savage, I would change the visual effects. Have the effects be further down and show the edge with moving mist, or something like that. (This would be especially useful for those with frequent cast times judging if they have to move, since the boss can be tilted for this attack, unlike in Savage.)

Barbarous Barrage

These knockback towers have excellent visual design, making it clear to the first-time player both that they are towers and that they inflict knockbacks when taken. This makes the long knockback distance completely fair for a regular duty, since the first-time player expects the knockback even if they do not know the distance, and so can position conservatively to keep their character from being knocked into the fire DOT wall at the edge of the arena. (It would be a bit much if the wall produced an instant KO, but that is only in Savage.)

These towers also have the same reference markers for knockback angles around their edge that are helpful in Savage, even though the player has much less use for them here. This is good, helping the player step up to Savage should they choose to.

Plus there is skill expression here in whether and when to use one's knockback immunity, especially when this is sequenced into other mechanics like Fuses of Fury.

More like this in regular duties, please! This is fair, fun, and encouraging.

Fire Spin

A simple spinning cone that hits each spot in the arena exactly once. Nothing special, but fine and here to be ramped up later.

Fuses of Fury

Places eight bombs in a square within the arena, leaving the center free. While this is conceptually fine, it has two visual issues.

First, the short fuse bombs (which explode first) and the long fuse bombs (which explode second) are distinguished only by the length of their fuses. Both have the same swirling pattern on their fuses, stretched out on the longer fuses. This is not enough of an indicator, especially for the first-time player who does not yet know that the bombs explode at different times despite being placed all at once. They will pick up the bombs in their peripheral vision, but not necessarily the fuses, and the directions of the fuses vary as well. The red text "The fuses are ignited!" also does not tell them to look at the fuses, only that the bombs have been set off.

To improve this, I would put the swirling pattern on only one of the fuse lengths, and have the other fuse length be a solid color. I would also avoid using black in either of the unlit fuse designs, to improve the existing teaching visual of the fuses burning to black as the fires travel along them to the bombs.

Second, there are no floor markings corresponding to the edges of the bomb explosions, even though there are circles marked on the floor. Since the explosions are fairly large for the density of the bombs, this is annoying right away, and even more so when these are sequenced with other mechanics such as Lariat Combo and Barbarous Barrage.

Doping Draught

A very thematic spectacle cast, giving the player an extended time to both enjoy the NPC interactions and think about what more dangerous attacks the boss might present. Even for the first-time player who knows nothing about stage wrestling, the dialogue and persistent buff on the boss make it clear to expect a difficulty spike.

For the sake of longevity through repeat completions and increasing average Item Levels in the Duty Finder population, I would want this to come slightly earlier in the timeline. In particular, I would cut the third Brutal Lariat, since two is enough to show that he can use either arm. The second Fuses of Fury before this is important to give the first-time player a second chance to look at the different fuse lengths.

Lariat Combo

Two Brutal Lariats back to back. Unfortunately, the pacing makes the second one reactive, especially since the lack of floor references prevents the player from reliably prepositioning their character just outside the first AOE. (So the player must be assumed to have their character out of position when reacting to the second hit, similar to when fighting Herpekaris in Origenics.)

As before, I would add floor references for this. But rather than only changing the floor geometry, I would enable an even faster pace by giving the player information about both hits in advance. This mechanic could show the boss charging up his arms for each hit, just like Maulskull does in The Skydeep Cenote. Alternatively, we already have the "Limit Cut" number markers (which you even included in your marker guide with the updated Hall of the Novice) that could be shown over his shoulders with the different shapes used in Alphascape V4.0 (Savage) (O12S) Fundamental Synergy. Either addition would allow the player to stand just outside the first AOE and start their character running / stay still as soon as that first AOE resolves, making the mechanic a small taste of executing a plan quickly rather than a test of reaction time.

(And yes, he would still turn after the first charge. Starboard / Larboard again, but this time with animation cues for the sides.)

Infernal Spin

The upgraded Fire Spin, this time combined with concentric circle AOEs and eventually also multiple sets of baited circle AOE drops.

Even though the Fire spin cone itself still only hits each spot in the arena once, the baited AOEs and the practicality of doing this with Duty Finder matchmaking are a hassle to those with frequent cast times. The player has a lot less to work with to plan their movement tools here in Normal compared to high-end duties, so use mechanics like this sparingly. Too much becomes annoying.

I would personally remove the baited AOE drops, but not due to the movement requirements. The timing of the baited AOEs means that the player who wants to preposition for an upcoming mechanic can find themselves blocked by not only moving too early themselves, but also by another player's character movement timing. This is a recipe for strong players getting annoyed at their party members, and is not happening much only because the strongest players have either stopped repeating the duty or regularly KO the boss before this gets sequenced into other mechanics. (Sure, the party can also get in each other's way in the Savage version (Fuse or Foe), but there they can correct each other effectively.)

Overall Impression

Whoever did the visual design for Barbarous Barrage should make more like that. Whatever process led to this floor having tons of useless markings and no useful references should be re-examined.

--------

AAC Light-heavyweight M4 (M4N, Wicked Thunder Normal)

https://i.imgur.com/14X03EQ.png

This floor is in an odd state, being useful for some mechanics and useless to actively unhelpful for others. But in this case the mechanics that it does not fit have their own problems, such that the mechanics are ill-designed rather than the floor being ill-fitting.

Sidewise Spark

The lightning visual effect is by far the primary indicator for which side is being hit. In fact, when the boss has wings out, it is the only indicator, since she does not change stance in that form. Since this visual effect is very cluttered, quite high up, and ramps up gradually, it is not a good primary indicator. Imagine what it looks like for visually impaired players!

Unlike Murderous Mist in M3N, here at least the cast name clues the first-time player in that this only hits one side of the boss. There is no reasonably mistaking it for a raidwide.

But if the player happens to have their camera pointed along the side-to-side line of the boss, rather than the front-to-back line, it is unreasonably difficult to pick her stance shift out of the wall of sparks to find out which side is being hit - and the Savage version does actively make the player do that (during Electrope Edge 2), with another visual cue mixed in! At that point, her silhouette is the same either way, so the player has to pick colors out of the wall of sparks, and that is not an acceptably clear indicator.

You can have the thematic lightning sparks in a clearer way. This could have been animated with the boss pointing her hands slightly further down to send the sparks near or onto the floor, rather than creating a visual wall high up.

Shadows' Sabbath (Basic Form)

It would have been easier for you to have each clone's whole body load in on the player's screen all at once, like you do for most appearing entities, than to have the fancy "displayed gradually as the ring passes over them" effect that you created for this duty. You put deliberate technical effort into making the clones, and by extension the visual indicators of their attacks, both difficult to see and late-appearing.

That was a terrible use of your development time. The obscured visuals benefit no one and hinder a variety of player groups, as I covered in the design principles section. The format of the late-appearing indicators merely teaches the player to learn when the first indicator they can reliably discern will appear and wait until then in boredom, again as I covered in the design principles section. For visually impaired players who cannot rely on the clone's mane direction as the first-appearing subtle indicator, or for anyone if the timings are tight enough, the deliberately late appearance of the more obvious indicators makes their Sidewise Sparks needlessly reactive. And to top it all off, this design encourages the player to use third-party tools that instruct them on mechanics without waiting for the visuals to appear on their screen.

Using the falling ring around each clone to show when its attack will resolve is a good thing, as it clues the player in to the timing without relying on them looking at their enmity list for untargetable adds. It also allows for faster sequences where the player may not see each clone as it first appears. However, for some later clone types, both here and in Savage, it is not tied to the clone's attack timing at all - which breaks the consistency of the rule the player has been using since A Realm Reborn. This seems like a further effect of tying the movement of the ring to the visual display of the clone.

If each clone displayed fully when it appeared, this would be a good introductory mechanic leading into potentially interesting sequences with the clones. But the choice to deliberately obscure the visuals and make them appear late spoils the clone mechanics.

Ion Cannon Mini-Phases

(The name is from Savage, where the cannon hits after Stampeding Thunder must be taken to cleanse debuffs first introduced by the Ion Cluster cast.)

(no castbar) Stampeding Thunder (Platform Break)

The concept of the boss breaking part of the floor with her attacks is cool, and mechanically reinforces that Wicked Thunder and Black Cat are sisters.

But here, unlike Mouser, the player is not given enough warning of which side the boss will choose to reliably avoid all the hits; even in Savage with Sprint already active and lots of practice, it is not possible at moderate ping without the use of third-party tools. In Normal, without the benefit of timeline knowledge, practice, or even getting party members to stand away from the boss's targeting ring, it is even worse.

In fairness, the individual hits are light here and inflict no penalty other than the damage itself. But it still feels bad, because the game has taught the player for so long that getting their character hit is always a failure unless the attack was obviously unavoidable. It feels like a reactive mechanic as a result, and is to a degree, even though on close inspection it is clearly not meant to be fully avoided.

Plus it is an enticement for those players on melee DPS and tank jobs to take additional damage for the sake of melee uptime and expect the healers in their party to compensate for that, which is surely not ideal. (In Savage, the tanks can handle the extra hits themselves and any other role will be KOed by them, which makes it less of an issue.)

It is difficult to eliminate the extra hits entirely without making the floor feel relatively fragile, which would be a detriment here unlike in Mouser. But the animation could tell the player which side will be safe earlier, such as by having the boss point the cannon earlier and briefly charge it, removing the felt and actual reactive nature.

Threefold Blast

The introduction to how the cannon hits work when not pointed at the platform. The hits always alternate sides, and clearly are not meant to be taken since they inflict a Sustained Damage DOT.

For the upgraded forms this leads to, it is a good introduction. The name cues the player to expect a sequence, and it keeps the same time between hits as the upgraded forms. (The hits are 2.5 seconds apart, making them slidecastable with the right Global Cooldown alignment, but the heavy damage and Sustained Damage for each failure leads to many quick KOs if the player tries to learn that alignment technique here.)

As a memory mechanic, this asks the first-time player to remember two items: put their character in the first safespot, remember the other two. For the player who knows the pattern, this is not a memory mechanic at all, since they know that the hits always alternate and that is enough to succeed after putting their character in the first safespot.

Fourfold Blast

This introduces two consecutive hits that do not alternate sides, while the rest alternate. The non-alternating pair can be at any point in the sequence.

As a memory mechanic, this asks the first-time player to remember three items: each hit after the first. For the player who knows the pattern, they need only remember one item: where the "stay" is in the movement sequence.

At the end of the sequence, this also drops circle AOEs under random player characters. This is clearly meant to be an adaptation of the Savage debuffs, and "here are some baited AOEs, move" is useful sometimes. But here it feels mildly irritating, since the same circle AOEs are used so often in this duty.

Fivefold Blast

The final form of this mechanic, and the only Blast used after the first Ion Cannon mini-phase. Always has exactly one non-alternating pair, just like Fourfold Blast. Ends with a spread in the smaller arena.

As a memory mechanic, this asks the first-time player to remember four items: each hit after the first. For the player who knows the pattern, they still need only remember one item.

This is a regular duty. Yes, the hidden pattern is not hidden that deeply, but you still cannot rely on the player knowing it in order to fit memory mechanics within the three item limit. I picked it up quickly, but that does not mean everyone is going to. Please imagine the first-time player who is already tired after a long day of work and caregiving; their working memory is already impaired, and they are not likely to pick up on the pattern from a single Threefold Blast and Fourfold Blast. They do not need to succeed at every mechanic every time, but they should still have an encouraging experience, which "ugh I forgot the sequence" is not.

I see this whole mini-phase as a missed opportunity to express Wicked Thunder's fighting style and hint at the Savage version without spoiling its puzzle.

The Positron and Negatron debuffs already have clear visuals around the affected player character from the Savage version, and "restore balance in your character, with imbalance shown by visual on or over your character" is already a recurring concept in optional regular duties. (Complete with one-hit KOs for failure.) You could have had Ion Cluster itself here, along with the cannon hitting each side in a different polarity like the charging visual hints at and the Savage version does. Just remove the added debuffs that create the Savage puzzle of arranging AOEs in a small space, and lighten the damage of the cannon hits (when taken with the opposite debuff) to make the healing requirement mostly look scary rather than being too difficult for Normal Mode.

That way, the double cannon charges before each hit would remove the memory mechanic escalation, and would show her use of the cannon to force opponents to take damage before getting to the duty's final phase.

Regular Arena Mini-Phases

Wicked Bolt

A multi-hit stack shortly after the arena floor is restored. Can cascade rapidly if multiple party members are KOed at the end of an Ion Cannon phase, but that is not in itself an issue - it would be perfectly fine if not for further punishing players who struggle Fivefold Blast.

(And yes, if the Ion Cannon phases were changed as I suggested, this would make for a more extended healing requirement. A little taste of extended healing is okay for Normal Mode, it does not have to hit very hard.)

Bewitching Flight

As cool as this looks, it is difficult to see where this attack will hit precisely, and that is clearly due to your design choices rather than the lore idea of the mechanic. Since the safespots are quite small, this is annoying even to the casual player who is simply trying not to get their character hit.

The electrope "feathers" of the wings would not look right if they pointed straight down, but you could have made the lightning emitters on the "feather" tips point straight down; they even already have a spherical glow that would make the orientation change look less jarring. Because they do not, the player must choose between looking at the emitters that are offset from where they hit, or the moving spray of lines from each that does not even consistently show the whole area to be hit. Likewise, you could have made the spray of lines stay still for visual clarity, but you put in effort to animate it in a way that obscures the danger zones.

As for the floor, for some reason you chose not to match the spacing of these line AOEs with the width of the floor tiles. There are five line AOEs on a 4x4 grid, when you could have used four line AOEs without really changing any other mechanics. (You could also have made the floor a 5x5 grid, which would have saved you some work for Electrope Edge 2 in Savage, but that would require a different pattern inside the tiles to provide references for many other mechanics.) In fact, four line AOEs would look more natural and balanced, putting two emitters on each wing.

This is annoying enough the first time, when just the boss is doing this attack. It is even worse the second time (and presumably subsequent times), when there is also a clone doing the same thing on an adjacent side, giving the player less time to look at each one. (This is also one of the places where you decoupled the clone's ring from its attack resolution, because the player needs time to see the full visual of the clone.) Clearly someone on the duty design team knew this on some level, since the Savage version of this mechanic removes the clone in favor of using electrope cubes whose spacing does match the floor grid.

This really should have had four line AOEs spaced to match the floor grid, and then it would have been cool instead of irritating. At least this only hits once and does not create cascading failures.

(no castbar, name inferred from Savage) Electrifying Witch Hunt

The lines on the floor left behind by Bewitching Flight expand and explode one by one at random, with circle AOEs dropped under four random player characters after each line but the last explodes. (When there are lines left behind by a clone, a north-south line and an east-west line go off at the same time.)

This is quite a long individual mechanic, since nothing else happens until all but the last line is gone. It is dull, since the circle AOEs prevent strategizing around the areas that explode first, instead requiring extended movement. It is lacking visual clarity, since the expansion and glow of the line that is about to explode are quite mild until late in the process. And the misalignment of Bewitching Flight with the floor grid compounds the issue, since the only consistent visual reference for where a line will expand to is the lines beside it, which may already have disappeared. (Especially if the Sidewise Spark cast during the last line's expansion does not cover the last line, leaving a small safe zone.)

Sure, I can do this, but it is boring. And the length of the mechanic makes the boredom grating. This is not the "use sparingly" I mentioned when reviewing Infernal Spin in M3N.

With a clear visual indicator, the casual player would have been able to comfortably handle the Savage line explosion order; the only reason it is not great to ask them to look so far afield here is that you made the indicator spend most of its time being relatively indistinct. That explosion order would have made this a lot faster, freeing up time for more interesting mechanics and preventing it from being extended movement boredom.

Shadows' Sabbath (Clone Sequence)

(This is very similar to Twilight Sabbath in the Savage version.)

This mechanic does allow the player to optimize by placing their character in the center of the arena to make the required movements for each Sidewise Spark very small. Unfortunately, the deliberately obscured clones mean that I wait for each clone's mane to appear in boredom and annoyance, rather than enjoying the casting-friendly sidesteps.

And for players who cannot reliably discern the tilt of each clone's mane, this is a reactive mechanic! If you do not give them the visual information they need until each clone is about to resolve, or they have to spend longer looking at a clone to find out which side it will hit, of course the mechanic becomes reactive for them. And both of those compound in their own way through the sequence, which can make the final Sidewise Spark from the boss reactive as well.

Witch Hunt

This bears no resemblance to either of the Witch Hunts in Savage, so it seems like a filler mechanic that for some reason is given the same name.

The pacing is only reasonable on the outside of the square spiral, since the player cannot be expected to memorize or "feel out" the snapshots instead of going off the attack animations in a regular duty.

Soaring Soulpress

This basic stack would be completely unremarkable, except that for some reason it uses the Sunrise Sabbath tower jump animation for its own resolution.

Final Phase

The HP% threshold to trigger this phase really suffers from the arena transformation being both long and exactly the same as the earlier Ion Cannon phases. The first-time player who enters this duty well after its launch may barely get to see that this is not the same as those phases, especially if a party member uses a melee DPS Limit Break to cut the extended healing requirement short.

Wicked Hypercannon

(The "temporarily no positional requirements" targeting ring indicator and boss debuff are good, but that is more about system visual design than about the specific mechanic.)

This is thematically cool, and I think it should have amplified that theme by introducing the Duty Finder population to the Wild Charge concept. (That is, a line stack that hits the player character closest to the source of the attack harder, traditionally requiring a tank to take the brunt of the attack in high-end duties.) That would have further expressed Wicked Thunder's anger at the party's survival, provided a stronger opportunity for the tank power fantasy of protecting the party, and gently added to the casual player's mechanical vocabulary.

The smaller platform forces all party members into the line stack already. The whole point of the mechanic is that she is blasting the entire party as a last resort. This already has an enlarged line stack marker, showing that the whole platform will be hit, and the line stack marker already has an arrow pointing downwards, normally to the target player character. You could have made a new Wild Charge marker by replacing that arrow with a red orb (from the tank stack marker) that is always over the closest player character to the source of the attack, and making the stack arrows between that orb and the source spikier and red. (Naturally, then, a double Wild Charge that requires both tanks would use two tank stack orbs.) By leveraging markers that the player already knows, the intuitive guess would be the correct one.

That being a concept introduction, I would not be too harsh with it. I would only have it hit the closest player character harder, and not so much harder as to instantly KO a non-tank; just enough harder that a non-tank will require healing to stay up for more than a few seconds, so that the casual player has the feedback to confirm their guess and tanks are encouraged to use their defensive cooldowns.

(Also, I would remove the circle AOEs dropped under player characters. They are overused in this duty as it stands, and they would be a distraction to teaching a new mechanical concept.)

Overall Impression

Making indicators deliberately difficult to see is anti-fun!

Aside from that, this duty is brought down by not being daring enough, missing multiple opportunities to present exciting and casual-friendly mechanics. As the final floor of a tier, casual players have the weapon blades as an incentive to repeat it for longer than the other floors, so it can afford to be slightly more complex than the rest of the tier. (In things that the player can actually improve at, not hoping they notice the hidden pattern in Fivefold Blast.)

On a broader scale, I find it strange that this duty's design draws so little from the first major phase of its Savage version, which historically is strongly tied to the Normal, and so much from the second major phase after the Savage-only arena transition. Historically, the Savage-only phase of the final floor of each raid tier is the main opportunity for challenge-seeking players to take on the difficulty of learning mechanics without inherent help and expectations from the Normal version, and conversely a major opportunity for duty designers to present an entire major phase without being constrained by the need for an adaptation that works in the Duty Finder context. And that historical position works well, since the player who reaches it is no longer brand new to raiding, so they have less need of help from the Normal.

Drawing mechanics from that final phase into the Normal version reduces the former's challenge and spoils some of the surprise, while diluting the first phase mechanics beyond easy recognition is strange and unwelcoming. Heavily diluted mechanics often wind up as dull filler even by the standards of regular duties.

Sure, Duty Finder does not do mechanic-specific planning and so will not manage the proximity baits of the Savage Witch Hunts, but Narrowing Witch Hunt / Widening Witch Hunt without those is literally a sequence of alternating donut and point-blank AOEs. You could have put that in as a sequence of two AOEs instead of four, followed by a basic spread, and the casual player would have figured it out. (Circle AOEs dropped under player characters wind up replacing a lot of party coordination elements when adapting mechanics, but there is such a thing as too much of them and there are other usable mechanic components.) I covered some other potential adaptations when discussing their diluted Normal versions. Just about the only mechanic in the first major phase of the Savage version that does not adapt well to a regular duty is Electrope Edge 2.

----------------

I do see why so many players initially enjoyed this Normal Raid tier. It has a variety of lore concepts across the duties, each duty has its own music for the first time in years, and M1N is a very strong start to the tier. But M2N is an unmitigated disaster, M3N has its visual annoyances build up over time, and M4N gets extremely boring with repetition as the player almost literally sees through the clutter and obscured visuals.

I will now continue my review of the design of the Patch 7.0 duties, in this subsection covering the high-end duties. As each of these has a corresponding Normal Mode, readers may wish to compare to my reviews of the corresponding regular duties, particularly for visual design which is often shared.

These duties are simply not for casual players. This is the planner's world.

----------------

Worqor Lar Dor (Extreme) (Valigarmanda EX)

https://i.imgur.com/tTB8em2.png

(no castbar) Animation Tells

Unlike in the Normal, here the player largely overcomes the poor teaching of the animation tells, and even two of them being overly similar, by repetition and learning when to look for the animation tells. But the duty would still be improved by the same teaching and visual clarity suggestions I made for the Normal.

Cataclysms (Phase Additions to Animation Tells, Charring / Crackling / Chilling)

The fact that each point in the timeline with an animation tell also comes with one of these makes the duty harder, but also reinforces looking for the animation tells at those specific points.

Opening Phase

Spikesicle

For this to be good rather than have the icicles be irritating, the player has to know the second hidden pattern: that the center base of the boss's targeting ring arrow does not get hit by any icicles after the first two. This is a bit much to expect the player to notice in an Extreme, and fellow players can easily wind up not passing it along if they do not themselves struggle with dancing around the icicles.

Skyruin (Fire Phase)

Triscourge

Though it is not apparent early in the duty, the way that each phase empowers its own elemental debuff is thematically cool. The long debuff timers are also welcome, both to aid the player in learning the timeline and to allow the debuff mechanics to be staggered throughout each phase.

(no castbar) Volcanic Drop

The red text "A volcanic eruption is imminent..." to cue the player to look for a volcano certainly helps with progression, especially "blind progression" without guides or other outside information.

Despite the volcanoes being far from the arena, this mechanic is fine visually, since the player only needs to look at one volcano and can then infer the state of the other.

The two sets of small baited puddles are also okay, since the party can plan to move as a group, though many parties wind up dropping them on reaction instead. It only takes one party member who forgoes the plan to make the party react to their puddle, and the open space means they are not penalized for it. Indeed, parties often do not bother making a plan for it.

I would prefer if the party was forced to make a plan for this rather than operating on reaction, not just for my personal enjoyment but also as skill-building for those new to high-end duties.

Scourge of Ice (Fire Variant)

The player can do this by reacting to the overhead marker, like they do in Duty Finder, so this is more a thematic mechanic than anything else. Its only real contribution to the high-end nature of the duty is that stronger players can use the debuff from Triscourge to know which of the two random timings they have in advance.

Mountain Fire

Thematically, this is a very cool tankbuster that satisfies the tank power fantasy of protecting the party. The spectacle of it helps with that too.

But despite looking right at the spectacle, I do not enjoy it except in videos, because in the duty I am too busy worrying about whether my character is correctly aligned with the tower blocking the boss. If there were floor geometry references for each of the three positions, this would not be an issue. (Most line-of-sight mechanics do not ask for rapid judgement of the attack source's location, and so there is time to get the alignment right comfortably.)

The movement is almost exactly every other Global Cooldown, which is fine and even suits RDM's Dualcast, but the player has to wait to see Valigarmanda's movement before starting their own character's movement.

Scourge of Fire (Fire Variant)

A good mechanic; the heal check is appropriate for a scary Extreme mechanic, this is always before the permanent DOT of Tulidisaster, and the puddles combined with light party stacks force the party to make a simple plan.

Scourge of Thunder (Fire Variant)

There is so much room here that parties often wind up spreading on reaction rather than making a plan. In this case, that is difficult to fix, since the mechanic is also constrained by making the other phase variants feel appropriate.

Ruin Foretold (Add Phase)

Unlike in the Normal, here the party can plan out where to take the stacks, and the light party Wild Charge nature requires them to. The heal check nature of this phase encourages healers to think about how they can reach both light parties, which is a good thing, without the heal check being too difficult.

I do not like the alternating cones being so fast and frequent in any duty that is supposed to be repeated often, but I will cover that where they come back in Savage. There they are much worse. Other than that, this phase is fine.

Tulidisaster

A big, splashy, thematic end to the add phase heal check, culminating in the same permanent DOT from Normal. The trickle of persistent damage changes up the style of the heal checks for the rest of the attempt, which is a good thing.

Skyruin (Thunder / Ice Phase)

The random phase order is totally fine. The mechanics to deal with before completing the duty are not changed, only the order.

Thunder Phase

Scourge of Fire (Thunder Variant)

This is still a planning mechanic in its weaker single-hit form, and the basic light party stack is a classic for those new to high-end duties for a reason. The fact that it cares about levitating or not in the thunder phase is an amusing reinforcement of keeping the light party together.

Hail of Feathers

Being able to see the feather order coming makes this a lot more fun than it is in Normal, even though the mechanic is harder.

Having the party create a safespot by killing an add is also nice to sprinkle in every once in a while, since it creates different rotation considerations. (There is such a thing as too much, since short-lived adds too often can break the structure of a job's rotation entirely, and long-lived adds eventually stop feeling like adds.)

Scourge of Thunder + Scourge of Ice (Thunder Variant)

The fact that planned spread spots require two different plans for this definitely encourages parties to just do all of the Scourge of Thunder spreads by reaction. Players new to high-end duties are not yet used to holding two plans in mind, but they have no trouble bolting for an apparent safespot like they have done often in regular duties. Since the floor does half the work of spreading for them and there are four extra spaces, they are not often penalized for this, making it easier for parties to go along with that rather than building their plan execution skills. (Or to not bother making a plan in the first place.)

The two plans are not out of the reach of Extreme at all. As with the Volcanic Drop, this would be a better mechanic if it gave the party less space to encourage remembering the plans.

Thunderous Breath (with Arcane Spheres)

This is the Normal mechanic overlap. If there was more party planning in other mechanics, that would be fine as the thematic breather, but as it stands this feels a bit easy.

Ruinfall

The visual design is good, just like in the Normal, but little is added here.

Ice Phase

Scourge of Fire + Scourge of Ice (Ice Variant)

Having the tanks stay out of their light parties for their spread AOEs is not a difficult concept, so it does a good job of gently introducing the idea of variations on the light party.

(no castbar) Avalanche

Here the player does not even need to turn their camera, which is good to focus on the spatial reasoning of combined AOEs instead.

(no castbar, snowflake AOEs) Arcane Spheres

There are so many snowflakes here that parties resort to using their waymarks to create floor references for these, since the arena itself does not provide them. Getting the waymarks right for that purpose is quite the hassle without third-party tools, since they cannot be checked at all until well into the duty.

The spatial reasoning aspect of this is really diminished by the annoyance of needing to either estimate where they go or use waymarks for it.

Scourge of Thunder (Ice Variant)

There is even more room here than in the fire phase!

Ice Talon

Teaching the off tank to make sure they are second in enmity is a good thing.

Final Phase

Wrath Unfurled

The Damage Up on Valigarmanda is a thematic warning before the final Mountain Fire and the enrage.

Tulidisaster (Enrage Sequence)

This definitely had to be the enrage. Not only is it on-theme, it would have been a missed opportunity if you had not done it.

Overall Impression

This is a lot more comfortable than the Normal, and overall a mediocre duty that is too light on the planning requirement. It definitely feels like the true form of the elemental phases, especially with each phase empowering certain attacks, but the lack of floor geometry for the ice and "regular" phases is a drag.

Notably, this is a very short duty, at not even seven and a half minutes. If it was relatively intense for an Extreme, that time frame would be fine, but the ease of the mechanics feels like there is not enough material in the duty.

--------

Everkeep (Extreme) (Zoraal Ja EX)

https://i.imgur.com/lyaXyz0.png

Opening Phase

Multidirectional Divide

Despite not having direct floor geometry, the clear spikes for the AOE lines make it workable to find safespots near them, and that is all the player needs here.

Forward Half / Backward Half

This is a fine random variation to use in overlaps, though it is easier to remember as "go where the castbar says, and for Forward go into the sword" than to do the spatial reasoning of the boss turning around with the sword.

Regicidal Rage

Players new to high-end duties have to learn tether taking somewhere, and this gives them lots of open space to do it. Since the tanks need to split them, they also practice planned spread spots.

Dawn of an Age (Distorted Worlds 1)

Vollok

Launching straight into the swords on other platforms is definitely correct here. So is going straight to the combination of two distorted worlds with Half Full, since all the pieces are familiar from the Normal version. Since each tile only has one matching sword across all four distorted worlds, the player can focus on checking the safety of one or two tiles rather than all 16.

Greater Gateway

The fire and wind tethers are clever upgrades to the Normal version. The player can do the spatial reasoning about one direction from each platform, keeping the task from being too difficult for players new to high-end duties.

Chasm of Vollok

It is nice to see tile spreads returning, complete with their marker from E4S and E12S. Standard spreads would not have fit well on such a small platform.

Projection of Triumph (Blade Field)

Both of these mini-phases are a fun way to encourage individual planning.

There is a lot going on for an Extreme once the overlaps begin, but the player has ample time to see the point-blank and donut AOE chains coming and keep them in mind for the overlapping mechanics. The blades on the floor that activate these AOEs also provide the appropriate floor geometry.

Projection of Turmoil (Stack Line)

These mini-phases are a really cool way to introduce the idea of planning to stagger debuff resolution gently, giving the party a lot of time and the ability to do it through staggered positions in space rather than going directly to moving at specific times.

Once it is learned, it becomes a breather mechanic, especially if the party is using the "conga line" plan where those with frequent cast times do not move until near the end. I enjoy the distance heal check of that method, even though it is suited to new raid healers, because it creates the opportunity for skill expression of reaching all the way across the large arena through my fairy.

In the second of these mini-phases, the Half Full casts are a simple overlap while encouraging the "conga line" plan, or at least coordinated movement.

Bitter Whirlwind

Tanks are going to learn to use their invulnerability cooldowns with a plan, rather than holding them in case of emergencies, and that is a good thing.

Dawn of an Age (Distorted Worlds 2)

The two-platform setup of this is good at gently teaching players new to high-end duties to be aware of healing range, since the DPS players have to make a small but active effort to be in range.

Drum of Vollok

This is not difficult so long as the role with the Enumerations stands still well in advance. If they do not, they tend to fling their partner's character off the arena, which is a good motive for antsy Disciple of War and SMN players to learn to stand still.

The downtime is exactly the same for everyone.

Sync

The two platforms both affecting each other is a nice twist, especially the way it includes the tile spreads. In particular, here the tile spreads enforce planning, since there are no extra spaces.

Aero III (Wind Tornado)

Reminiscent of The Cloud Deck (Extreme) (Diamond Weapon EX) in making the party coordinate to swap who moves when, though with role groups rather than light parties.

The angle of entering the tornado really matters here, which would be a problem if there was not both ample time and a convenient line between tiles to aim along. Since the time and the floor geometry is there, this teaches the player to be mindful of what is coming and not panic, which are both good.

Here, the downtimes are different for each role group, and so can mess with the player's personal cooldown usage a bit. I have definitely done the optimization strategy of using free heals to safely delay my Burning Chains tether break so that I can use Aetherflow first.

Actualize

I am not sure why this particular cast of the arena transformation raidwide comes with a knock-up. That seems like an oversight, as there is no mechanical reason for the forced downtime.

Dawn of an Age (Distorted Worlds 3)

The first sword AOEs leading shortly into tile spreads encourages the party to assign tile spread spots rather than reacting to each other. Some parties will wind up doing it on reaction anyway, but having a long time to see each other's positions and spread out during the Sync mitigates the dissolution of the plan.

(Otherwise, this phase is repeated mechanics.)

Actualize (Enrage)

There is nothing wrong with using a powered-up cast of the familiar raidwide as the enrage, but thematic enrages are definitely nicer.

Overall Impression

The best-designed high-end duty in Patch 7.0, with no major flaws and some clever mechanics. A strong contender for best overall duty design in Patch 7.0, competing only with M1N.

That I got bored quickly here is entirely the fault of job design, not duty design; nothing suitable for Extreme can challenge an Ultimate-level healer, and the current healer job design is not up to the task of filling in the gap.

--------

AAC Light-heavyweight M1 (Savage) (M1S, Black Cat Savage)

https://i.imgur.com/pRHutey.png

As with the Normal, this floor is great except for being slightly too bright. In fact, the brightness change I suggested there is primarily for this version.

Opening Phase

Quadruple Crossing

I like this. It turns the Normal version's Black Cat Crossing into a party coordination mechanic, a dodging mechanic the party has some control over, and a part of the heal check all at once.

Here it takes the place of the opening raidwide, both launching right into the duty mechanics and getting the player to practice the AOE baits in this form before it becomes complicated by Leaping, clones, and Nailchipper. This is an appropriate amount of advance practice, enough to keep the later forms from being intimidating but not so much as to create a boring opening. And since the player will be doing their opening burst during this time as well, they are encouraged to learn how the almost three second gap between each hit and the next fits in with their job rotation.

Nine Lives

This warning cast (indicating that the next mechanic will be stored in the next clones, appearing before every Soulshade) is not strictly necessary here in Savage, but it is very thematic, and for a first floor it is not a bad choice. The warning reminds the player to remember the next mechanic for later, creating lenience in the duty's requirements of timeline knowledge and attention control. Some lenience in those things is necessary at this difficulty, though naturally much less than in the regular duties, and this sort of warning cast is one of many tools to create that.

One-two Paw

Finally a timing that feels right for this, rather than the Normal delaying the first snapshot from the end of the castbar. The time between the first and second hits is not so tight that the player must learn the snapshots rather than reacting to the first animation, but is tight enough that the player must be ready to go rather than hesitating, which is the right speed for a very simple concept like this at this difficulty.

One-two Paw (Double Clone)

A fine simple spatial reasoning mechanic, which is not actually a memory mechanic since the player can use their character's current position as a reference for the second safespot of the One-two Paw being released. If the player struggles with doing the spatial reasoning in real time, they can memorize the possible patterns or a condensed rule outside the duty, making this not an unavoidable spatial reasoning test either.

The time given is quite long, including both the clone awakening and a second Nine Lives to warn of the next stored mechanic. For stronger players and/or repeat completions, no longer needing the warning cast, it is up to job design to fill that void.

Double Swipe / Quadruple Swipe

Partner / light party stacks are a classic random variation in high-end duties, and this is a clear and thematically appropriate name that the player can process quickly. This is good, since they have to process it quickly coming out of the double clone One-two Paw.

Leaping Quadruple Crossing

The player has seen Leaping Black Cat Crossing in the Normal, so what this does is immediately quite clear. But they must remember to dodge into their planned positions for the stored Swipe, which is just enough to keep them from treating it the same way as the opening.

Bloody Scratch

This physical raidwide is scary in this form, always coming directly after unavoidable damage and requiring mitigation for Disciples of Magic to avoid KO at full HP and 10 Item Levels above the minimum. Teaching the party that they have to mitigate is welcome, even though it results in some frustration for new raiders and those in their parties. (Especially since the healers themselves are more likely to wind up with KOs from lack of mitigation.)

Mouser (Mouser 1)

The gradual arena transformation itself feels very good here. Going down to four tiles, all of which are already fragile, emphasizes the deadly dance and prepares the player for the lack of extra tiles, while the maximum of two movements to evade all the hits and the long cast make it not too difficult for new raiders.

Once the player is aware of the hidden pattern to start on the south-west central tile and look at the south-east central tile, they only need to remember one item: whether that south-east central tile gets hit once or twice. And here in Savage, you can ask the player to find hidden patterns or learn them from fellow players within reason, so this works out.

Elevate and Eviscerate

The random selection of whether each DPS will be targeted in this phase and each tank or healer in Mouser 2, or vice versa, is perfectly fine. For the task of completing the duty, it largely evens out since each player character is chosen exactly once unless there are KOs or disconnects. For those seeking parse runs, they generally want to skip the downtime entirely by getting the random variation where they are selected late in Mouser 2, then KOing the boss first, but duty design should not be warped around facilitating the perfect parse run every time.

The random order within each role group is also okay, though it can get to be a bit much for players who enjoy precisely mapping out their plans for playing their job in the duty. There are eight possible times where they could take the downtime and two possible clone orders in the Leaping phase, creating 16 possible timelines to work with. Not only would the player finish gearing a role before reaching 16 duty completions, but the downtime can force the player to leave a major personal cooldown unused, thus affecting their timing for subsequent uses for the rest of the duty and encouraging them to make separate plans for that case. (All these downtimes are outside of raid buff windows, which is a necessity given the "2 minute meta" in current job design.) Obviously the knock-up makes it difficult to employ a bind and permit the player to continue their rotation, but the effect is worth noting.

As for the attack itself, I hope your design intention is for the player to realize they do not need to know whether the clone is preparing the knock-up version or the downward punch version. (The "braindead" strategy is to put one's character at one corner of the chosen strong tile and facing inwards, causing both versions to land on that tile.) Party members can leave the clone under the boss from when they were targeted, so the player cannot reliably see the clone's stance and glowing arm, hence the popularity of a strategy that does not require them to look at the clone at all.

One minor pain point for optimization-oriented players and those with bad ping / reaction times is that this mechanic is not hard enough to force players to plan out who will take which strong tile and in what way, as an inherent result of the random selection order and cast variation necessitating time for the selected player to position and face their character. This leads to parties operating on "use eyes" reaction to each other instead, especially during progression, which is annoying when trying to cast, do positionals, or just not get one's character swept off the arena by someone making a last-moment tile change. Not every mechanic can avoid this result, but it is worth trying to keep it from happening too often.

For all its flaws, this is still mostly a fun mechanic.

Shockwave

Since this knockback goes through knockback immunity, unlike in the Normal version, it enforces for the party that they must avoid breaking tiles with Elevate and Eviscerate in order to leave themselves room for this. The visual for which tiles will be restored first is not particularly dramatic, but it does not need to be since the player has plenty of time to look at it and is expected to plan to look at it. It is clear enough for its job.

Grimalkin Gale

Tries to enforce that the party plan out spread spots and split up accordingly using the Shockwave, with moderate success given the light hand on a new-raider-heavy population. Some parties do still spread by reaction after the melee pairs of each light party gapclose in and the ranged pairs stay out. For me, with a full plan and a strong desire to avoid adjustment loops, moving my character to a planned location immediately and stopping there usually works to get my partner to move accordingly.

Memory Phase

Since this phase centers on working memory and spatial reasoning, I will count out items in working memory over the whole phase.

Leaping One-two Paw

The mechanic itself is familiar from Normal.

The player must look at which clone is spawned, then transfer the leap (direction shown by a buff on the boss, though not for long given the spatial reasoning task) to the clone. This creates one item in working memory, the clone's leaping position. It looks like two items initially, since the positions are intercardinals, but the player can either use the intercardinals as a set or place the number waymarks over the possible leaping positions to compress this into one item. (The latter method is very popular, since it also provides a reference for the clone's Leaping Quadruple Crossing resolution point.)

They must also remember which side is safe first, as in/out since all clones in this section face into the center of the arena, creating a second item in working memory.

Leaping Quadruple Crossing

Familiar from earlier, except that the boss turns to the center of the arena and makes a second Leap from its earlier position. This makes checking the direction harder until the player realizes to either re-orient or use the buff on the boss.

Storing this clone's leaping position creates a third item in working memory.

It looks like there should be a fourth item here, to remember which clone has which cast since they activate in a random order. However, since they are stored in a fixed order, the player can use the order of the items in their working memory to reconstruct which clone has which cast. (The north position will always be tied to the north clone, and the south position to the south clone.) As a result, this phase just barely scrapes by the three item rule, which is part of why typing out the stored items into party chat is so popular. (The other reason is that players who are comfortable with the spatial reasoning can help their party members by doing this.)

The two timelines from different clone orders are fine. The party will want to have a different healing and mitigation plan for each, but they have ample warning for which to use once they see the first clone activate.

Tempestuous Tear

These light party stacks force the player to not just remember the stored clone position and first safe side, while also being the only major magical damage in the entire duty. Magic-only mitigation tools get piled on here.

Nailchipper

Again forcing the player to do more than just remember the stored clone position, and also to potentially bait in the opposite role order for Quadruple Crossing. The hardest mechanic in this duty, so it is probably wise that this comes with only a single item retrieval from working memory.

The size of these spreads makes me think that Quadruple Crossing is best baited on the cardinals, leaving the intercardinals safe, to allow the role group with Nailchipper to leave ample room for the other group. Having this sort of strategic consideration is a good thing! (Unfortunately for my own duty completions, that is not what Aether Party Finder does, and the Normal version does not have the flexibility to suggest it.)

Mouser (Mouser 2)

All the considerations are the same as in Mouser 1, except for the addition of Overshadow / Splintering Nails.

Overshadow / Splintering Nails

Party stack or role stack is also a recurring random variation, and this adds a moderate heal check to the phase. There is enough time for healers to set up the party healing between their own Elevate and Eviscerate hits, though it makes them think about doing that, which is good.

Final Phase

Raining Cats

This is the type of mechanic where tank invulnerabilities are always going to be the preferred method unless the duty forces the party not to use that method. It is simply more consistent for the party to involve fewer players and fewer steps in resolving the mechanic, tethers or no tethers, and more consistency results in a greater chance of duty completion on each attempt. Which results in most parties not practicing the "normal way", and therefore having trouble doing it if the tanks use their invulnerabilities earlier to salvage an attempt.

As long as you know that, this mechanic is fine.

Predaceous Pounce

When reviewing this mechanic in the Normal, I wrote "on a purely mechanical level it feels like the sort of thing that is put in a Normal to replace a Savage mechanic that cannot be usefully adapted for casual play." So why is it here in Savage?

This is not interesting to do at a basic level, and offers little to optimize around. (Trying to remember all the jumps would be excessive as a memory mechanic, which removes optimization options, plus the One-two Paw overlapping the end limits the player's choice of safe locations anyway.) It is boring. Its saving grace is that raiders so rarely see it, since it is the last major mechanic of the first floor, and so it is often skipped by combination of gear advantage over the minimum Item Level and better job performance than is required for the duty.

Mouser (Enrage Sequence)

As rare as it is for the player to see this, it is very cool. Not only is it thematic to take the floor out from under the party using the pre-established gradual arena transformation, but it is still worth it for the player to do their best to track the sequence in order to dodge for as long as they can before they run out of tiles.

Overall Impression

A decent first floor, if on the easier side even for the first tier of an expansion.

It is worth noting that the unavoidable damage to non-tanks in this duty is almost entirely physical damage. In such a duty, the difference between mitigation tools that mitigate only magical damage and those that mitigate both physical and magical damage is very large. In duties where that damage is entirely magical or almost so, the difference between those mitigation tools is small or even irrelevant. Just like Anabaseios: The Tenth Circle (Savage) (P10S) Harrowing Hell, the duty design choice to present a largely physical heal check is what made the job design difference in tanks' party mitigations become a problem. (I could write a whole other post about good and bad homogenization, but that would be outside the scope of this review.) With Patch 7.10, you granted DRK and GNB some physical damage mitigation on their party mitigation tools, but that was not available to them in Patch 7.05 when you released this duty.

--------

AAC Light-heavyweight M2 (Savage) (M2S, Honey B. Lovely Savage)

https://i.imgur.com/i7ludwW.png

This arena's floor is very well designed for the mechanics of the duty. There is no grid because there is no need for a grid. The cardinals, intercardinals, and even half-way points between cardinals and intercardinals are all marked with sharp contrast from both the regular honeycomb pattern and the blue Beat stage. The triangles that mark these at the edges of the arena are perfectly sized to provide reference points for dodging every Tempting Twist, Honey Beeline, and their Splinter follow-ups without telegraphs, enabling the player to precisely find their spread and pair spots for their party's strategy. As a bonus, these also mark every possible Groupbee spawn position.

Opening Phase

Call Me Honey

A classic opening raidwide, but also necessary to show the party how hard it hits before Rotten Heart. Due to the Beat raidwides, it does not come back until then.

Splash of Venom / Drop of Venom

Spreads and partner stacks are another classic random variation pairing for memory mechanics, and this is the only type of memory mechanic in this duty. (Including the Beat versions that give hearts instead of high damage, Spread Love / Drop of Love.) Since there is only ever one of these stored at a time, this duty is comfortably under the three item rule.

However, this memory mechanic is overused in the duty. Six times in eight minutes gets boring, especially since the opening is forced to have two of them back-to-back to show the player the Savage forms of the Splinter AOE patterns. It also does not help that the memory mechanic is never combined with anything other than Tempting Twist / Honey Beeline in its resolution, though that would be fine if not for the overuse.

While the cast names here are not blatant about the number of hits, they are still quickly processed so they are fine.

Tempting Twist / Honey Beeline

The follow-up Splinter honey blob patterns here should have been in the Normal. Duty Finder could have done them fine, even though the casual player will generally not appreciate how cleanly they lay out the spread / partner spots and how little movement is necessary for Honey Beeline.

Killer Sting / Stinging Slash

Since there are five points where the boss casts one of these two tankbusters at random, there are 32 different possible timelines for tank mitigation. But the total lack of hidden patterns denies the tanks, and by extension the party, most of the planning possibilities around invulnerability usage. This is a good example of some randomness making the duty more fun, and too much making it less fun.

Honey B. Live: 1st Beat (Beat 1)

You made this Beat only force the party to take 23 hearts, not the maximum safe 24 hearts, because on some level you knew that the party making a plan and executing it correctly is not enough to guarantee that no extra hearts would be taken. As I explained in the design principles section, that means the mechanic is not appropriate for the deadly dance of high-end duties in the first place.

On the other hand, I do enjoy how the Beats have a different style of heal check, with the individual mechanics hitting more lightly but combined with the heart DOTs. This is an interesting way to vary the structure of healing requirements within a given difficulty.

Centerstage Combo / Outerstage Combo

These are good.

The cast name is both thematic and quickly processed for the first hit, and the structure offers optimization for every role that has movement-dependent optimization. (Only the cross AOE that is always the second hit requires leaving melee range, the pacing is fast enough and far enough to require instant casts or healer extended slidecasts but not for annoyingly long, and the donut AOE is a healing opportunity.) Most uses also encourage the player to dodge in a way that prepositions their character for the upcoming mechanic.

Love Me Tender (11 Towers)

The fact that 11 tower spots are used randomly out of more than 11 possible locations prevents the party from fully planning out this mechanic by assigning player characters to tower positions, priority order or otherwise. The party does not have the advance knowledge to work out which remaining locations will be used.

The fact that the initial eight hearts for this Beat are assigned completely at random (outside of not giving any player character more than two hearts) prevents the party from creating a plan that sufficient players are willing to learn and execute by assigning player characters to the timing of tower spawns.

(Planning it that way is theoretically possible, most easily by having the first eight towers be taken by a fixed order of party members and then going back to the top of the list for the priority order to take a second tower, with everyone in the center of the arena when not taking a tower. But the fast debuff checking required for the priority order is more suited to Ultimate, since it must be done either before any of the towers or with the ongoing towers in mind, and with small icons that are not really meant for checking other party members' debuffs. No one really wants to be at the top of the list, since it is more work and it makes them less likely to have personal leeway on the reaction sequence, while their party members at the bottom of the list will never be reached in the priority order. That method does not care about melee range at all, and even encourages leaving melee range early to help the party count towers, so players dislike it for the sake of trying to meet the damage output requirement. And of course, it requires thinking of it at all, which is counterintuitive and well beyond most players at this difficulty level. I thought of it while doing this review, not in my many weeks completing this duty every week.)

Technically this mechanic is plannable, but not within a reasonable effort or in a pleasant way. As a result, parties resort to roughly prepositioning based on their planned clock spots or their Honey Beeline spread spots before the towers begin, and then having each player react to the appearance of towers near their character.

The problem here is not that they are reacting to the towers - the 10 seconds between a tower's appearance and its resolution is plenty of time. The problem here is that they are reacting to each other's choices, both of whether to take a tower and of potentially leaving part of the arena unattended entirely. So exploded towers and towers taken by two player characters still occur without anyone making a mistake, which is not an acceptable result in any high-end duty.

Regardless of whether you intended this to be done using that thoroughly unpleasant plan or by the party reacting to each other's choices, this is a fundamentally bad mechanic. You must ensure that a reasonable and enjoyable plan exists for every single mechanic in every single high-end duty.

Loveseeker

This is such a basic AOE that I see its role as pushing the player characters out of the center of the arena before the reaction sequence, whether that is meant to push them closer to the edge for their own benefit (away from the moving hearts emitted by the boss), to their detriment (towards the random moving hearts from the edge of the arena), or to put an endpoint to the planning method for the preceding towers.

Love Me Tender (Moving Hearts Reaction Dodging + Stack)

This is less of a visually overwhelming mechanic than it is in Normal, since the pink circle puddles are not included and the party has the opportunity to practice and plan. But that does not make it not a reactive mechanic.

Here, in theory, the party can plan to place the non-stack group (with three hearts each) between two of the fixed spiral arms from the boss, the stack group (with less than three hearts each) between a different two of the fixed spiral arms, and have them each follow behind an arm slowly at a slightly smaller radius than the edge of melee range. I have gotten parties to do this, and it significantly reduces the disruption from the randomly baited hearts coming from the edge of the arena. Frankly, in a party full of players at my skill level, and who know that fact, it would almost eliminate that disruption entirely, as the slow stutter-step movement with tight control from the party leaves only a small range of angles from which a heart baited on a group can still block that group's path several seconds later.

But even in that best case, such a plan does not entirely eliminate the disruption. So when it does happen, each player must make their own judgement on which way to dodge a stray heart and how far they have to move - the party cannot make and execute a plan that keeps each group together, because that would rely on exactly matching individual judgements of angle and distance. Those individual judgements are not themselves particularly reactive other than for hearts hidden by the stack, since the player can see each heart coming in peripheral vision and filter out the fixed spiral pattern. But once a group is forced to make those individual judgements, they must then react to each other to bring their characters back together after the stray heart passes by. As before, this is an unacceptable result in any high-end duty.

And in practice, the situation is much worse than this, because most parties do not follow this partial plan.

First, the format of the slow movement required will not occur to most players at this difficulty level naturally. The vast majority of mechanics do not penalize moving fast, so the player defaults to thinking of dodging mechanics at their character's full running speed. There are no puddle drops here, which is the usual way that mechanics enforce and suggest stutter-stepping while also providing the edges of the puddles as references to the player for their next movement. And even at the Ultimate level with those references, parties learning TOP Pantokrator usually make a mess of the puddle drops for quite a while, since one player overshooting affects their entire light party.

Second, the presentation of this mechanic pushes players away from that movement pattern. Since this starts during a raid buff window, and BLM did not have Ley Lines flexibility in 7.05, both BLM and PCT players do not want to move their characters away from their personal area buffs. And the fixed spiral arms are hidden by the randomly baited hearts, pushing the player's attention to reacting - like they did in the Normal - in early progression. Once they are in a reactive mindset, that makes it less likely for them to discover the hidden pattern in future, though it is not a complete dead end unlike Barreltender in Tender Valley.

Third, if the stack group is not already together, they must react to who gets the stack, and then to each other in how they bring their characters together. All this while the nearby hearts are obscured by the stack. (Unlike in Normal, the heart stacks here do not give more than four hearts if the maximum of five eligible player characters stand in it, though who gets the "free heart" for the remaining part of this dodging sequence is random.) This further pushes the player into reacting in early progression, and if the plan is disrupted.

Fourth, even one party member who does not follow that plan will bait several stray hearts on top of the direct confusion, dissolving the plan into every party member reacting individually. In high-end duties, one player's not knowing the plan or execution failure collapsing the plan for the whole party is perfectly acceptable, and often even necessary. But if the duty is well-designed, the feedback from the duty reinforces the importance of the plan to the party. Here, the feedback from the duty drives the party back into reacting and away from the plan. In turn, this compounds by creating more and more stray hearts until it becomes a reaction time pressure mechanic.

Fifth, this mechanic is a recipe for players getting angry at each other or quitting without noticing the possibility of planning. Those players who have low ping, good reaction times, and comfort in bullet hell games progress quickly through this mechanic and the rest of the duty, since the duty is too easy and repetitive for a second floor Savage with the reactive mechanics excluded. They do not want to hang back and make a plan, they are inclined to continue progressing the Savage tier. (Even for the occasional failures that they themselves make, they are often inclined to look at them as a personal failing rather than a structural problem with the mechanic.) The duty is not hard enough to push them into planning regardless, and frankly it would have to be harder than an Ultimate to overcome the counterintuitiveness of the partial plan.

Meanwhile, those players who struggle with this mechanic with no plan are forced into doing it in the fully reactive form by even a single party member who feels that way, resulting in them failing it frequently. And then that party member may well shame them for getting their character hit, since healing the boss and giving it a Damage Up affects the whole party. This is a very discouraging experience whether or not they come to perceive the possibility of a partial plan, and instead encourages them to respond angrily to the party member who shamed them. (Whether over the shaming itself for their reaction time, or for neglecting and/or refusing the partial plan.) Even parties with no significant differences in reaction time can wind up sniping at each other over individual failures, as it takes serious contemplation to realize that the party can never truly master the mechanic.

For all of these reasons, most parties treat this mechanic as largely reactive, with either no plan at all or only a general area for the stack group.

Again, this is a fundamentally bad mechanic.

On a personal note, the reactive nature of doing this without the full party on board with the partial plan puts noticeable physical stress on my arms, due to the party's reliance on my individual success. The towers were less so, since fellow party members would move their characters away from my glamoured-for-visibility character, but my party cannot help me in that way here. Early progression of this duty, doing this over and over again in quick succession, was not pleasant.

Alarm Pheromones (Alarm Pheromones 1)

"Grand Octet, but bad."

That is how a friend put it to me, referencing the UCoB mechanic with eight divebombs, and I completely agree.

This is the same as the Normal version's Alarm Pheromones 2, the Random Baits "pattern", except that in Savage the knockback from getting hit by a Groupbee goes further. (Making it more likely that a single hit will put a character into the death wall, and therefore making the mechanic cascade faster.) The increased harshness of the penalty is not the issue, nor is the mechanic being unsuited for Normal Mode.

The issue is that this mechanic does not fully yield to planning either.

Suppose that the party approaches this mechanic on reaction, just like in the Normal. Then it is reactive and failure-prone for all of the reasons I covered in reviewing the Normal.

Suppose instead that the party adopts the "Mario Kart" strategy, which is running along the edge of the arena as a group to stick together without having to react to each other. (This method is sometimes employed to have baited attacks pass harmlessly behind the party, and other times to create an even spacing of objects dropped on player characters for later usage.) Initially that is an improvement, but true Mario Kart does not work here, since the lack of advance knowledge of the order means that the party can find a Groupbee blocking their path no matter where they start. Once that happens, the player must again make an individual judgement on whether to continue forward or stay back, just like with the moving hearts earlier. As a result, the party becomes separated in the same way, and must either react to each other to bring their characters back together or revert to approaching the remaining divebombs on reaction. In the former case, the party can elect a member to try to stick to, usually the main tank for melee uptime reasons, but then still must react to each other in a more limited form. In the latter case, again the plan dissolves, for analogous reasons to those regarding the moving hearts.

Suppose instead that the party adopts the "box" strategy, splitting the party into groups with widely separated "boxes" that they each start in and try to dodge in. This is an improvement in the average visual processing demand, since most Groupbee baits will wind up only going through one "box" and the lines wind up less cluttered, but it is not an improvement in the worst case. High-end duty plans must be reasonable in the worst case. Further, the judgements on dodging each Groupbee are still individual, so each player must still react to the choices of the other players assigned to their box.

None of the available plans eliminate the reaction time pressure, since they all include the party reacting to each other, and operating entirely on reaction creates reaction time pressure as it does in the Normal. Yet another fundamentally bad mechanic.

That does not mean that you cannot have high randomness or gradually revealed information in divebomb mechanics. Grand Octet does both and is fine. But Grand Octet gives the party enough information in advance to be fully plannable, which is why it works. There, the player only ever needs to react to the boss's choices of random variations, and they have enough time to do so at the Ultimate level of preparation.

Adding to the annoyance of this mechanic, it is not only an uptime mechanic but also right before a raid buff window. As a result, the KOs and Damage Downs sustained from failure here have an outsized impact on the party's damage output, which matters to completing the duty here in Savage. (And moving reactively to avoid the Groupbees often disrupts the player's damage output, since that is inherently not planning.) This makes the party discouraged and annoyed over mechanic failures that are not really mistakes, and even regularly leads to parties deeming the attempt ruined and running their characters into the death wall to try again.

On a personal note, my party cannot help me much here either, so this and the moving hearts create two physically stressful mechanics in close succession. Progressing Beat 2 was really not fun for me, not because of anything about Beat 2 but because of getting there. I had to stretch and eat far more than I should have for the amount of raiding time I was putting in, on top of finding these mechanics annoying.

Honey B. Live: 2nd Beat (Beat 2)

This mechanic is nothing special, but on an individual mechanic level it is okay. Of course, it does contribute to the memory mechanic overuse I mentioned earlier.

Since there is no reason for the party not to put both heart stacks together, they generally use that to have all the 0-heart player characters move as a group and have all the 1-heart player characters move opposite them. This was a predictable result, especially as the gear advantage increases in the Party Finder and Raid Finder populations, since for pure duty completion the benefit to consistency of learning fewer "roles" and avoiding reaction to the puddles and the stacks outweighs the small damage loss of slightly more movement. For stronger players who can handle learning more "roles" and processing information from the boss faster, this can be irritating as it denies them the opportunity to keep their character still for personal optimization.

Alarm Pheromones (Alarm Pheromones 2)

Much like Beat 2, this is an individually okay mechanic that contributes to the memory mechanic overuse of the whole duty.

Having to estimate the space available from the Groupbee lines alone gets a little irritating in the tight space, particularly for the Bee Sting role group stacks at the end. On the other hand, the random selection order here is purely a good thing, since the player can plan out handling the movement for each of the four possibilities and their job without the cooldown-blocking annoyance or timeline profusion of M1S Elevate and Eviscerate.

Honey B. Live: 3rd Beat (Beat 3)

Other than even more use of the memory mechanic, this is actually pretty good.

It is mostly about assigning positions, keeping the party's plan in mind at all times, and working towards one's next needed position during the Combos. Plus stronger players can do some individual optimization even if their party is not up for the more demanding movement of the melee uptime plan, which makes it less irritating to be in a group that prefers the easier plan.

Rotten Heart

In principle, there is nothing wrong with having phases exclusive to high-end duties outside of Ultimates and final Savage floors. (So long as the full story is still available through casual play, as you learned with the Coils of Bahamut, and the phases are designed with care for the player's learning experience.)

But this is 90 seconds of boring in a 10 minute battle, measured from the end of the starting castbar to the enrage. If the player is not currently healing, all they have to do is stand in their clock spot, step into the center of the arena when their debuff has a party-agreed number of seconds left, step back out to their clock spot, and perhaps use their party mitigation. Even if the player is healing, the heal check is not that difficult as soon as the player realizes to stagger their AOE healing and mitigation cooldowns.

For comparison, E12S Terminal Relativity is 63 seconds measured the same way, with ramping damage and more frequent hits, and that in an environment of significantly more difficult jobs and less non-healer healing. Sure, that is a final Savage floor, but even Duty Finder players can handle more than "here is a striking dummy for 90 seconds, except you step into the center of the arena briefly at a specific time, step back afterwards, stand mostly still otherwise, and maybe use a mitigation tool".

(I have heard some players call this "newbie's first Curtain Call", comparing it to the final mechanic of Asphodelos: The Fourth Circle (Savage) (P4S) due to its placement at the end of the duty timeline and its successful resolution necessarily being based on having each debuff not quite expire. I call it "newbie's first Patch" based on the original O12S recurring mechanic, focusing on the partner-based debuff removal and the interaction between the fixed damage timings dictated by the boss and the variable damage timings arising from the party's plan and execution in dealing with the debuffs. Both of those had more going on, and even then Curtain Call was boring for non-healers.)

Exclusive phases should feel special and exciting through their mechanics, even if they are not all that difficult. Rotten Heart misses that mark.

To make matters worse, by the time the player can reach Rotten Heart, this is not the only boring section of the battle for them. The heavy repetition of the memory mechanic has made its teaching section rote and also boring, and that is the entire first minute of each attempt at the duty. As a result, at minimum one-quarter of the battle timeline is taken up by boring mechanics, even if the player does not reach the point of boredom during the annoying mechanics.

Sheer Heart Attack

The thematic enrage of charming the party is a very nice touch.

Overall Impression

That the bad mechanics from M2N made it into Savage is astonishingly awful - the mechanics do not yield to planning either, violating a fundamental expectation of high-end duties! And outside of those mechanics, the duty is mechanically too easy and too repetitive for a second floor Savage. This is another reason that I chose this duty and its Normal version as my redesign examples.

--------

AAC Light-heavyweight M3 (Savage) (M3S, Brute Bomber Savage)

https://i.imgur.com/cd7jYpP.png

Just as in the Normal, the floor markings here are oddly lacking in usefulness. Fuse or Foe is telegraphed the whole time. The only use of this floor geometry is the in/out boundary for Quadruple Lariat / Octuple Lariat, and that only indirectly at certain points where circles meet. This is a needlessly annoying choice after putting in the effort to give this floor internal geometry, and becomes more annoying here due to the higher penalties for failure.

Opening Phase

Brutal Impact

Here in the Savage, the player consistently gets to see this ramp up with each Doping Draught, from four to six to eight hits. Even as a DPS, they get told to mitigate accordingly, which encourages them to notice it rather than tunneling on their job. This is the main mechanic that really sells the boss getting stronger in basic power, not just presenting different and harder mechanics.

Knuckle Sandwich

While the added hits here are mostly noticeable to tanks, the ramp up has the nice effect of encouraging tanks to plan their invulnerability uses around the last two tankbusters instead of purely around maximizing uses. Some tank pairings get more out of this than others, which is an unavoidable result of the differing invulnerability cooldowns.

Quadruple Lariat / Octuple Lariat (Always Out)

Partners / cone proteans is a classic random variation, and the number of hits in this cast name is quickly processed.

The only thing I would change about this is the name; it makes mechanic explanations needlessly confusing due to the overlap with Lariat Combo. Unlike Lariat Combo, this mechanic does not match the real category of wrestling moves, and there are lots of other options.

Quadroboom Dive / Octoboom Dive (Always Proximity)

Partners / spreads is once more a classic random variation, and this is another good cast name. The boss's stance is very well animated here, clearly indicating which side he will dive to for rapid processing without relying on the targeting ring.

I see so many melee DPS get their characters KOed for being too close to the proximity AOE, and that is a good thing at this difficulty. It teaches them to learn not only the plan but the timing to execute it well.

Barbarous Barrage (Barbarous Barrage 1, ends in Murderous Mist)

The excellent visual design for these towers shines even more here than it does in the Normal, using the panels on the outside of each tower to show how many player characters need to take each. All the pieces for this small puzzle mechanic are immediately visible to the party, and it is a good puzzle.

This is almost a satisfying mechanic as well, but I think having to adjust one's character position after each knockback is what prevents that. The player has to focus on aiming the next knockback rather than enjoying the sequence. This is not a bad thing individually in demanding continued focus, but it becomes disappointing since the Savage tier as a whole lacks satisfying mechanics.

Doping Draught (Doping Draught 1)

A long opportunity to raise and heal before an almost completely open raid buff window, with a spectacle to enjoy while the player does what should be a well-practiced burst rotation. The timing of the spectacle is a treat for strong players, as they will use their raid buffs while the boss is distracted by the Refbot. (Literally - the Refbot takes top enmity briefly.) Pressing Chain Stratagem while he is punching out the Refbot is easy, but it is also funny.

Quadruple Lariat / Octuple Lariat (Always In)

Introduces the fire around both arms as an alternate form tell for this and the Divebooms, which is not particularly high contrast but serves well enough here.

Also contributes to the open raid buff window without literally doing nothing for the duration, since there is no leaving melee range and DPS traditionally have the same spots for proteans and role partners. Healers can manage with their extended slidecast windows.

Quadroboom Dive / Octoboom Dive (Always Knockback)

Again introduces the fire around both arms, which makes the clear stance animation all the more important to be seen with them.

There is enough time to move one's character to a planned position after taking the knockback to the right general area, even with a little hesitation on when the knockback ends. This timing is about right for the difficulty; not the longer leeway time of Extreme and first floor Savage, but also not the constant tight control of Ultimate.

Tag Team

Having a separate cast to spawn the clones before each Chain Deathmatch is important and correct, since it gives the player time to turn their camera to put both clones on the screen at once.

Chain Deathmatch (Chain Deathmatch 1)

A good mechanic, that would be made even better by having floor geometry to show the edges of the Lariat Combos from the clones. Being able to use floor references to go exactly as far as needed would be a good opportunity for optimization.

The concept of having to take the hit is a recurring way to ask the player to maintain focus rather than doing what is usual for the mechanic it modifies - we have Beyond Death for a reason - and the presentation here is very on-theme for staged wrestling. Even the chains are clearly magical without being overbearing. (Showing only the chains relevant to one's own player character was correct and necessary.)

The timing is correct for the memory mechanic nature, having to store which clone to be hit by as one item in working memory. Some players may prefer to store it in their camera angle instead, by prepositioning in the middle and then pointing their camera only at the other clone to dodge that clone only, but the camera turn also takes time and that strategy does not work perfectly for Chain Deathmatch 2.

The way that this can be used to practice Chain Deathmatch 2 early in progression reminds me of using Hello World 1 to practice Hello World 2 in O12S progression; it does make the duty easier to progress, but that is a tool rather than inherently bad, and it is especially Party Finder friendly on individual-responsibility mechanics like this one since the whole party does not have to be aware of the later mechanic.

Quadruple Lariat / Octuple Lariat (Random)

Now the player must look at both the castbar and the boss's arms, even once they have fully learned the duty timeline. They have the time to do this, though it can get slightly tight for the difficulty if they have a particularly long movement distance from the random variations of a given attempt. (The Chain Deathmatch directly before the first use of this form will always have the final safespot opposite from someone's assigned position, making the "out" variation have them cross the arena. Again, better floor geometry would help, since it would allow the player to shorten the movement distance by prepositioning their character just inside that final safespot for Chain Deathmatch 1.)

Final Fusedown

Here, the fuse timings being distinguished only by length gets even more annoying than in Normal. The fuses on player characters are very small, and the player must check their own fuse and fuses in the arena rapidly to identify the random variation each attempt. (There is a hidden pattern, but that mostly benefits healers and ranged DPS.) They know that the fuse lengths matter, but seeing them is unnecessarily difficult. If you want to have mechanic indicators that are only visuals around player characters, those visuals need to be appropriately clear. (The gaols in The Weapon's Refrain (Ultimate) (UWU) are not great, but they are large and clear enough for what the player needs to know, which is just who has them. The monitors in TOP are awesome for visual clarity, giving the player everything they need to know including a crisp line for the edge of the AOE they are manipulating. Do more like that.)

Likewise, the lack of floor geometry for the size of the bomb explosions is worse here than in the Normal. The central safespot to wait out the explosion group the player's character is not in is quite small for something that must be estimated every time, but would be fine if there were reference circles with which to work it out.

The concept of this mechanic is very cool, extending the fuse concept from the Normal into something the party has to manipulate. It is just the visuals that get in the way.

Quadroboom Dive / Octoboom Dive (Random)

Now the player must look at three things: the castbar, the boss's arms, and the boss's facing. The clear stance animation is much appreciated, especially given the relatively limited color palette of this duty.

Fusefield

This is "newbie's first Curtain Call", with a priority order for assigning fuses thrown in instead of healers needing to leave healing range for their tether. Unlike Curtain Call, it only occurs once, which is good for keeping it from becoming too boring.

Having a big scary "not too fast, not too slow" debuff management mechanic every so often is a good thing to teach newer raiders that skill. This is slightly too similar to Rotten Heart to be in the same tier, but I would change that by changing Rotten Heart instead since that one gets boring more easily.

Doping Draught (Doping Draught 2)

Again creates an open raid buff window, this time with a potion, though not nearly as long a breather as the first Doping Draught.

Quadroboom Bombarian Special / Octoboom Bombarian Special

Having the breather window between Fusefield and this is important at this difficulty level, since both are multi-hit heal checks and the boss is now at full power.

Fuses of Fury

While this does have the larger safespots of Normal, and is mostly a thing that complicates the opening of Chain Deathmatch 2 (while the party is taking in the needed information), it would again be better with visual clarity for the fuses and the size of the explosions.

Chain Deathmatch (Chain Deathmatch 2)

Much like Chain Deathmatch 1, except that all party members must take the Murderous Mist hit and DOT so mitigation is advised. It is important and correct that the boss face himself for this Murderous Mist, so that the player can rely on its edges lining up with the exact middle and two corners of the arena.

The player still only needs to remember one item, since the other chain is always to the boss.

Fuse or Foe

Players chose to develop strategies that ignore the different fuse timings not just for better uptime, but to avoid the hassle and inconsistency of trying to discern length from very small fuses alone.

Barbarous Barrage (Barbarous Barrage 2, ends in Lariat Combo)

Here, the reactive structure of the random Lariat Combo is more annoying than in the Normal, and not just because it is higher pressure. There is also the inherent lack of precise prepositioning in taking the knockback, combined with having to adjust for that in reorienting one's camera. It gets to the point where I try to skip this mechanic entirely, and I expect others do too, just to avoid the reactiveness and loss of consistency from it.

This sequence is a big part of why I suggested showing both hits in advance in the Normal, not just improving the floor geometry.

Doping Draught (Doping Draught 3)

A thematic warning to the party to use one last potion before the enrage sequence.

Special Bombarian Special (Enrage Sequence)

Thematic enrages are cool. Even though this does not have a true soft enrage before the forced party KO, since it does not ramp up within the sequence, it has a similar effect on the healing plan: requiring healers to keep their focus on both healing and damage output until the very end.

Overall Impression

If this duty had been visually polished and presented as a second floor (with adjusted tuning), it would be excellent. Doping Draught is used to great effect here thematically, which also makes progressing the early mechanics of the duty easier; healers will have comparatively more tools to salvage mechanical failures before the true healing requirements are presented, along with the Doping Draught casts themselves as open windows to raise and heal. Likewise, in attempts to complete the duty for loot, recovering from mechanical failures is easier before the second Doping Draught than it otherwise would be.

As with M1S, this duty inflicts mostly physical unavoidable damage to the party, which creates the same considerations that I covered there.

--------

AAC Light-heavyweight M4 (Savage) (M4S, Wicked Thunder Savage)

https://i.imgur.com/14X03EQ.png

The arena floor is in the same odd state here that it is in the Normal.

Bewitching Flight

Has the same visual clarity issues that I discussed in the Normal, and here it is not practical for the whole party to keep their characters near the boss for a better view of the visual indicator.

Electrifying Witch Hunt

The proximity bait indicator is appropriately clear, and by this point the player should have no trouble learning to already be looking at the boss when it appears. The only annoying part is losing the indicator for the edges of the second safespot when the first lines explode.

Those seeking the perfect parse run will always want the random variation that best suits their job's uptime considerations, but that is okay.

Narrowing Witch Hunt / Widening Witch Hunt

Having half the random variations make the whole party move more would be questionable, in regards to lowering the party's damage output too much to make all possible random variations present a similar difficulty to completing the duty... except that this duty's damage output requirement is only slightly higher than M2S, and job balance is an ongoing subject of concern.

Other than that, this mechanic is a little too easy for a final Savage floor.

Electrope Edge (Electrope Edge 1)

This is primarily a memory mechanic, and something of an odd one to count out. The Sidewise Spark at the end guides the player into realizing the hidden pattern that the two small explosions must have one on the east side and one on the west side fairly easily, since otherwise the mechanic would present either no safespot or extra room.

The oddness of the counting comes from the rapidity of the displayed information. It is not inherently obvious whether the player is better off working in the shapes of the Witchgleam casts or the locations being hit. The speed helps the player use their short-lived sensory memory as a reference, but puts time pressure on converting the shown shapes into location pairs.

Consider the player's ideal categorization of the items to be stored. At the end of the Witchgleam casts, they clearly only need two items: the two safe intercardinals, or alternatively the two danger zones. (Some parties use number waymarks on the intercardinals for easier storage, much like in M1S, though here it depends on their chosen strategy for Sunrise Sabbath.) But how do they get there?

After the first Witchgleam, the player is storing two items: the two intercardinals that were hit.

After the second Witchgleam, the player is storing three items: the one intercardinal that was hit twice, and the two intercardinals that were hit once. They know there are only three sets of Witchgleams in this mechanic, so they can rule in the intercardinal that was not hit at all as a safe zone.

After the third Witchgleam, the player must compare the final hit to their three stored items and identify which of the intercardinals in that final hit was also hit before. The hidden pattern that each of the second and third Witchgleams will reveal one of the danger zones works to their benefit even if they have not noticed it.

That ideal categorization stays within the three item rule, but it requires very fast conversion of the cast shapes into locations being hit. The casts are so fast that I would not be surprised at all by some players memorizing the six possible shapes of Witchgleam casts as an intermediate storage form, or even being entirely unable to do the mechanic short of trying to follow their party.

Personally I can do this, using a mix of shapes and locations. But if someone else cannot do it and cannot reasonably improve at it, that is still a problem.

Sidewise Spark

Here this also comes with four or eight orbs, indicating partners or spreads. These are clear enough through the clutter of the sparks, which have the same problems here as they do in Normal.

Partners / spreads is a classic random variation, and the floor tiles make the tight space work by providing exact references for the edges of the safe zone.

Electrope Edge (Electrope Edge 2)

This is almost a very good mechanic.

The player only needs to remember one item (how many times their character was hit by Witchgleam). Extending the concept of Witchgleam casts expanding the resulting AOEs to be something the party manipulates is cool and thematically appropriate. And the healing requirement is just enough to make the healers think about it while contrasting the light distance heal check here with the heavy short-range heal check of the Wicked Bolt into the Ion Cannon mini-phase. The temporary 5x5 grid is very crisp and clear, the player can learn to work out the orientation of the Lightning Cage AOE pattern quickly by looking at specific sections, and the position within a square not mattering is great for skill expression.

The one catch is that it makes the player do Sidewise Spark with their character off to one side of the boss, denying them a camera angle that allows for seeing her stance in her silhouette. As I mentioned when reviewing the Normal version, this is an unacceptably obscured visual indicator. The sparks should have been much closer to the floor to allow for this to work well.

Ion Cannon Mini-Phase

Ion Cluster

Allows the player to see which debuff their character has nice and early, so they can prepare themselves for the fast pace of resolving the Electron Stream casts.

The visual indicators for Negatron and Positron are well done. The blue and yellow colors could stand to be more distinct and perhaps a little less bright, but the colors are not necessary since they are also distinguished by whether the circles are crossed or parallel. This is the same across the debuffs, the visual effects around player characters, and the cannons, making it well unified despite the three different visual design contexts.

(no castbar) Stampeding Thunder (Platform Break)

It still feels bad to have to take a hit here in Savage, but a lot less so once I realized that whether I used Sprint did not even matter. With a full plan and the party keeping their characters away from the boss's targeting ring arrow for everyone's speed, my character takes exactly one hit from this with or without Sprint.

I do plan extensively to make my practical response time just a little bit faster, not just the prepositioning my character in the middle of the arena that is visible to my party members. I also point my camera either East or West, looking at the boss's targeting ring arrow on one side of my screen rather than the center, so that I can preposition my keyboard hand fully for one movement direction and preposition my mouse hand for the other. Plus, of course, prepositioning my eyes to already be looking at exactly the boss's targeting ring arrow well before she turns.

(Remember all the way back when I was reviewing Herpekaris, the first boss of Origenics, I described splitting up movement along a straight line between both hands for speed? This right here is that usage for speed.)

Is that missing some of the cool animations? Yes, it is. But when you make a mechanic about moving fast to avoid the KO from a second hit, and you do not build in enough advance warning to prevent that little bit faster from mattering, I am going to do it anyway. Even if taking an individual KO here did not cascade into a party KO, it is still not fun for me to be told "not fast enough" by the duty, and this is the sort of thing I can actually do to avoid that. If you want me to perceive fancy animations at all outside videos, do not put mechanics like this with them.

At least party members taking extra hits for the sake of uptime does not burden the healers here.

Electron Stream

This mechanic is almost great.

It has an appropriately toothsome healing and mitigation requirement, complete with a Wild Charge on each side to involve the tanks' personal mitigation and make the whole party respect the positioning. It has a little puzzle in working out how to safely place all the added debuffs, that feels different because the available space is so small. The design within the floor tiles provides reference locations for solutions to the puzzle.

But I do not like that the Electron Stream casts (the cannon hits) apply the debuffs with only five seconds to resolve them. I would have applied them earlier, whether simply with a longer timer or in an inactive form that gets activated by the cannon blasts, and had them resolve at the same time. (As I mentioned when reviewing The Interphos, there is no reason to limit the concept behind Code Smells to Ultimate!)

Five seconds is not too short to just react to something, and the player is supposed to memorize their plan by this point in Savage. I mean, clearly I do a much larger character movement faster for Stampeding Thunder just before this. But this mechanic does not allow for the same prepositioning and pure focus on large-scale movement.

Consider the potential solutions to the debuff puzzle.

The nature of the donut AOEs (Roundhouse Conductor) and the closest / furthest cones (Proximate Current / Remote Current) dictate that the party wants to resolve this on the inner part of the central remaining floor tiles, totally aside from concerns of melee uptime or character movement speed. A party could decide to run the cones and the Collider cleanses outside the donuts, but they really do not want to. (In fact, Aether Party Finder uses a strategy that does not distinguish between Roundhouse Conductor and the point-blank Spinning Conductor at all, in part because those debuffs are not that visually distinct from each other at speed.) As a result, the character positioning precision requirement is inherently going to be quite high, in order to direct the Currents correctly and avoid hitting anyone with the donuts. Such precision strongly encourages looking at the intended destination to aim and avoid overshooting, pretty much requiring it for consistency.

Moreover, look at the arrangement of the three potential positions I could have to move my character to. Not only are they not in a straight line for hand prepositioning, there is no place I can preposition my character to get all of them at 90 degree angles from each other, or even 45 degree angles from each other. (Remember, I effectively cannot use a standard controller, since they do not allow for using one's whole arms to press keys or move joysticks. I use Legacy movement.) As a result, there is no way for me to get all three positions down to one keypress each, or even two keypresses each. At least one of them is going to involve a camera turn, which I cannot preposition my hands for in the same way, or a mix of two different movements.

So I cannot finish moving my character until I have finished moving my eyes from looking at my debuff bar to look at my intended destination. Because the movement distances are so small, I cannot even start moving my character until that point without risking overshooting, even for the potential destinations where I have prepositioned my character and camera to not require a camera turn. For those that do require a camera turn, I can move my eyes to look at roughly the right spot while turning my camera, but properly locking onto it is rough when the camera is also moving.

Let us make a list of steps:

1. The server sends the debuff to my client and my client eventually displays it.
2. See my debuff from my debuff bar.
3. Make a camera turn if necessary.
4. Look at the precise destination dictated by my debuff and the party's plan.
5. Press the keys / mouse buttons to make the appropriate movement, refining aim and length of keypress by visual tracking, while also doing my pre-planned heals and casts.
6. Look "up" to the party list to see if further healing is needed, because I do not have a static and in Party Finder someone might have missed a heal or mitigation tool. (Same with Raid Finder.)
7. Decide on that further healing, and try to get it not only received by the server but rippled out to the whole party before the debuffs expire.

Suddenly five seconds is not that long.

When I was progressing this mechanic, it did not just feel reactive to me, it was physically stressful to the point of hurting my arms to do over and over again. With a consistent cohealer and completing the duty, repeat completions went below the point of physical pain but never really got comfortable - and said cohealer is off to do FRU, so I am not encouraged by the prospect of going back in on my own.

Giving out the debuffs in advance would allow me to move the first four steps on that list ahead of that five-second window, even if they were only given by the previous Electron Stream cast (or Stampeding Thunder for the first one). I can handle getting all three at once, like E12S Intermediate Relativity does, too. Either way, that would be so much less physically demanding without changing the pace of the mechanic or the things the player needs to do at all.

Please count steps really carefully when giving short lead times between the player receiving information and needing to have used it. Or just give longer lead times like you used to.

Electrope Transplant (Phase Transition)

This mechanic is both extremely boring and very physically demanding. It is little more than an encouragement to repetitive strain injury. Why is this almost a minute of the duty?

Literally the only reason that the Duty Finder population cannot do this is the Conduction Point into Forked Fissures "protect your partner" mini-Wild Charges, and for most of the time the player does not have to remember or think about them. The Duty Finder population would figure out the Fulminous Field cones after an attempt or two even with the limited telegraphs given here. Transition mechanics are generally not meant to be particularly difficult, but this is too easy, especially since the same thing is repeated twice. And while the movement timings can be aligned with healer extended slidecasts, I expect that RDM and BLM have a much worse time of it due to Dualcast and Fire IV respectively.

The movement every Global Cooldown requires constant attention, but then it does not do anything interesting with that attention. That is why this mechanic is so boring.

As for the physically demanding nature of it?

Well, the natural way to learn this mechanic is by pointing one's camera at the boss, to learn what Conduction Point looks like and check one's understanding of the plan against the party's movements. Which makes the dodging entirely sidesteps, except for possibly one forward addition (back for melee DPS and tanks) to swap positions for Forked Fissures. Which in turn means that it is pretty much all on my keyboard hand, along with all my job actions. (I am aware of MMO mice, but I cannot offload much to one without injuring myself.) Doing that over and over and over again, every Global Cooldown with a short break for most of a minute... it hurt my hand even when completing the duty once a week for loot. Progressing this mechanic and those after it was awful.

I had to pin down which mechanics I find physically excessive and then check all of them for less demanding methods that still work in Party Finder. This one does have one, but it is only slightly less demanding and only usable after the learning process. Namely, I take the same Legacy movement split method I use for Stampeding Thunder, but here I use it for balancing the physical load between both my hands instead of for speed. And I have to specifically look counterclockwise, since in Aether Party Finder that is the direction of the dodge leading into Forked Fissures, and it is much more practical to aim into the tight space with my mouse if I have to move my character inward.

This was only viable once I made sure that my partner getting Conduction Point would still be visible on my screen and practiced seeing it rapidly while also getting the visual distance feedback to use the floor geometry references for how far to dodge the cones. Further, it can only take half the movements off my keyboard hand, since melee DPS and BLM/PCT want to stay in the same general area - for positionals and personal area buffs respectively - and that leads the whole party to do it by alternating movements. (I could try to explain starting my own character in a different place to my party, but that is confusion-prone and still would not help with the second round.)

I am sure you meant to include some visual spectacle with this phase transition. I am not seeing it, because I have to put the boss entirely off my screen to do this mechanic without hurting my hands. In fact, I have stopped perceiving the excellently done music change during this phase transition too, because I spend more effort and attention on protecting myself from injury than on doing the mechanic itself.

Plus, to avoid having the awful progression experience that I did, I would have had to know that you were willing to create mechanics like this and change how I approach duties accordingly. I would have had to use guides and timelines to screen every single mechanic of every single duty for its movement patterns in depth, flag any particularly demanding movement patterns, check videos to determine if they even have alternate options, all before even committing to the duty. And then I would still have to do the learning process for the mechanic outside the duty from videos, in order to use the physically safe method from the beginning.

That sort of analysis is now required for me to safely commit to a high-end duty, because if you will make one mechanic like this, how do I know that you will not make more? In regular duties, if I discover a problem mechanic and have to leave for my physical health, the discovery is extremely annoying for me at best and injurious at worst, but at least my party can replace me. In Ultimate progression, that is a disaster for both me and my party.

This one mechanic, by itself, would have been enough to stop me from committing to FRU on launch.

Relatively simple mechanics, appropriate for phase transitions, are just not worth this much movement anyway. Every button press or joystick tilt is a tiny little bit of effort that eventually has to give a tiny bit of reward back, and simple mechanics do not have the room to do that for this much movement. (Note that this is much worse than continuous movement in one direction, because that is one press and hold.) If you want a good "overload" phase transition mechanic, make something like TOP Wave Repeater, which requires far more thinking about how it complicates the overlapping mechanics than actual movement.

Conversely, if you are going to make a mechanic that requires this much movement, it had better at the very least be interesting to match, and hopefully naturally load-balanced too. TOP Pantokrator requires moving every Global Cooldown, mostly in stutter-steps, for most of a minute and it is perfectly fine - because it is a proper Ultimate-level mechanic and naturally load-balances itself. Almost all of the movement is in the same rotation direction, with smooth turns and no reason to look at the boss, so I can just do the whole movement with my mouse.

Cross Tail Switch

I hope you meant for parties to use a tank Limit Break 3 on this, because they sure were going to as the natural response to this design. They do not need the damage of a DPS Limit Break to complete the duty, and it is a lot more consistent for either tank to press one button at a specific time than for the healer pair to have a carefully coordinated healing plan and execute it under pressure. (Remember how the Party Finder response to P3S Life's Agonies, the heal-to-full raidwide at the end of Death's Toll, was "Macrocosmos or healer LB3"? Same thing, it can be done without the one button, but the one button is so much more reliable.)

I do enjoy that this needs some healing even through the tank Limit Break 3, and it does look cool. The short downtime is a natural place to use a Global Cooldown shield for it, which is a good nudge for healers new to this difficulty.

Sabertail (Exaflare Field + Light Party Stacks)

This is too easy, but would be fine in a lower floor. It does not even enforce that each light party agree on an initial spot to stand while looking at the exaflares!

The visual density is fine here; the exaflares have clear and familiar markers, and it is okay for the party to take quite a few attempts to notice the hidden pattern that the exaflares away from the east and west walls only ever go north and south. Once they realize that, it confirms their natural guess to split the light parties east and west for melee range access, and each light party only needs to look at the wall furthest from them.

Wicked Special

It is just an in/out, which makes its usage almost entirely a simple demand to look at the boss at a specific time and move. (Healing plans get worked around the potential for the "out" variation, but only when the party is split, and the constrained time is not long.) If the other mechanics in this phase were appropriately hard, that would be fine, but they are not so instead it contributes to boredom and repetitiveness. In context, it feels like "look and react" even though the cast is six seconds long, just because of the boredom.

Even O4S had in/out/Antilight in its second phase.

Mustard Bomb

This mechanic is relatively simple for the difficulty, but not too simple to be used as a "breather" mechanic mentally while still keeping the healing and mitigation requirements up. Between that and its short duration, its two appearances do not get boring.

Aetherial Conversion (Tail Charging, Memory Mechanic)

This requires the player to store exactly one item: which side to move their character to first. The tails retain the color of the stored element, so the player can choose between looking at the boss during the Tail Thrust, storing front / middle as a separate item, or condensing all four possibilities into a set to make one item.

Tail Thrust

Aside from releasing the stored Aetherial Conversion, this also varies the healing and mitigation requirement of the duty; the knockbacks do relevant damage to the party, while the fire AOEs are entirely avoided. I do not mind some variation in where the heal checks are, with long lead times so that the party can use their plan for the timeline, but this is varying the amount rather than the timing. Since heal checks are so inherently tied to the skill floor of the duty, this very much does affect the difficulty of completing the duty on any given attempt.

If the fire AOEs came with a light raidwide too, this would be fine.

Twilight Sabbath

The whole point of this mechanic is making the visual indicators difficult to see to force the player to react. It is not fun or funny. It is both reactive and boring.

If not for the baited puddles dropped before the first clones resolve, Duty Finder could do this mechanic. (They very nearly do, with Shadows' Sabbath in the Normal version.) And all the puddles do is prevent the fun optimizations, like sidestepping with a slidecast, or prepositioning in one quadrant and dashing to another if needed!

This design makes the player work with a bad camera angle to determine which clones will resolve first, compared to the Normal, as they all spawn at once and the height of the falling rings is a lot less clear when the clones are closer in. The player has to do the spatial reasoning for overlapping AOEs on top of watching for the clone manes (or the first part of the visual they personally can use) to appear, which is not at all interesting but adds to the time pressure. And half the time the mechanic is harder, because the south clone resolving second requires a camera turn for the simultaneous Wicked Special.

There is nothing to salvage in this mechanic. Even if the player is not pressured on their reaction time, they do not have anything interesting to do here. And on top of that, here they are more likely to be aware of the third-party tools that the delayed clone visuals encourage, or have someone in their party using those tools to do early callouts.

Midnight Sabbath

This mechanic is only saved by the fact that the wings of the donut AOE clones appear fairly early. Otherwise it would also be annoyingly reactive, on top of being too easy for a major theme mechanic.

Chain Lightning

This is entirely a memory mechanic.

In isolation, it would be fine, since it only asks the player to store three items: move their character to the first safespot, remember the other three. But there is a stored Aetherial Conversion before this, which puts this point in the duty timeline at four items.

Personally I can do this, but it is still unacceptable on principle as a violation of the three item rule. There is no further compression that the player can do here if they find themselves limited by their working memory. Further, this occurs during a raid buff window altered by the earlier downtime, so the player has extra working memory demands from their job at this point in the duty timeline.

This is why so many parties resort to writing down the Aetherial Conversion before this in party chat, since that is shared across the whole party unlike Chain Lightning itself. In turn, that encourages them to write down all the Aetherial Conversions, and possibly Chain Lightning by light parties, which further dulls the duty. (To make matters even worse, all that writing incentivizes parties to look for members who both do not use a controller and have the APM to type in combat.)

Sunrise Sabbath

The closest thing to a satisfying mechanic in the whole raid tier. It is a good idea, with enough lead time for all the random variations, but needs some more meat to it to be a proper final floor setpiece. (Notably, this is the other place you had to decouple the clones' rings from their attack resolution times, to allow the player to see what the clones will do in full.)

Very slightly narrower cannons would also be wise, to reduce the focus on character movement precision to avoid hitting other player characters with cannon baits.

Sword Quiver

The only thing suited to a final Savage floor in here is the heal check. The mental processing and movement would be fine in an Extreme, and without the role partner cones it would be fine in a regular duty.

Wicked Thunder (Enrage)

There is nothing wrong with using a powered-up cast of the familiar raidwide as the enrage, but thematic enrages are definitely nicer.

Overall Impression

"Boring and based on making things difficult to see, the fight."

Too easy for a final floor, still based on making indicators deliberately difficult to see, too memory-heavy, too much movement required without correspondingly interesting mechanics. The exclusive phase is particularly disappointing.

The duty that gave me the biggest push to worry about whether my body can handle FRU.

----------------

To be perfectly honest, even if I knew that my body could handle FRU, I would still have serious concerns about whether FRU would be sufficiently enjoyable to fight to be worth the commitment of an Ultimate. There is not a single truly satisfying mechanic in the entire set of Patch 7.0 high-end duties, and M2S and M4S both lack the necessary care for their difficulty towards what makes high-end duties possible and fun.

Ultimate is fundamentally about the intense challenge, about pushing the player to their limits but not beyond them, and it has to do that by using the duty design to create precise requirements for playing each job in the duty. I enjoyed every moment I spent in TOP because it walked that line. But if current duty design is not delivering high-end duties that walk the broader paths of Savage, how can I be confident that FRU will walk that line?

To be clear, I know that different duties are designed by different people within the duty design team, and that relatively few designers are entrusted internally with Ultimate designs. I still have a great respect for those designers. But it is clear that something within the duty design process for Patch 7.0 has produced duties with major fundamental flaws, whatever it is/was extends to the staples of raiding, and a single major patch is not enough time for feedback requiring major changes to make it through your production pipeline in general. If the length of that feedback loop were to be different for Ultimates, surely it would be longer rather than shorter; much of FRU's design would have to be completed before the release of Patch 7.0, and making design adjustments to the tremendous precision required of an Ultimate would involve careful consideration of how each mechanic contributes to the overall demands of the duty.

For a player who is completely ready for Ultimate, committing to attempt to complete a specific Ultimate upon its launch still means committing roughly 100 hours of their life to effective progression of that single duty. (Even for world progression raiders - they are exceptionally strong players, but they are also working without the benefit of guides and with little time to think about the mechanics outside the duty.) That is before counting time spent finding a group (or multiple groups), time spent ineffectively due to Earth circumstances such as illness, and time spent studying the duty outside the duty.

And they must make that commitment before getting to see the design of the specific Ultimate - purely on trust of the duty design process to deliver a journey that is worth the commitment. The many fundamental flaws in Patch 7.0's duty design, along with the relative lack of enjoyment, do not inspire that trust.

Now for the promised full duty example of the design style I want applied to a Dawntrail concept: redesigning the Honey B. Lovely duties - AAC Light-heavyweight M2 (M2N) and AAC Light-heavyweight M2 (Savage) (M2S). While I cannot promise perfection, as I do not have access to test servers or other design tools, I hope this is good enough to convey my intended style.

The lore and concept of the duties will remain identical. Honey B. Lovely is the queen bee, the idol songstress, the charmer who fights by enthralling even her opponents. One of the design goals is still to express this concept mechanically. It is a great concept!

I will utilize the visual changes I suggested in my mechanic-by-mechanic review earlier: for the Alarm Pheromones mechanics, the Groupbees have their AOE lines extend across their entire path through the arena, and these lines appear in black rather than green to keep with the bee theme while providing high contrast. I will also use a further visual change: having the Groupbees show the telegraphs for their AOEs for the last 2.5 seconds of their cast, to enable precise positioning beside the AOEs for certain Savage mechanics.

----------------

Redesigned Normal Mode Timeline

Even in Normal Mode, I have included some stinging surprises. To emulate the experience of discovery and learning for the first-time player, I will introduce these with the mechanics that include them.

Call Me Honey

Remains as in the original. Raidwides are a staple mechanic, and the cast name fits the duty concept. This makes a classic opener for Honey B. Lovely's performance theme.

Tempting Twist, Honey Beeline

Cast in a random order, one directly after the other, but both will always appear. Safe areas as in the original. This provides a light warm-up with easy dodges that teach these recurring attacks while players finish out their openers, the randomization prevents it from being completely static unlike the original, and strong players can do uptime optimization on repeat runs with Ley Lines placement, slidecasting, and/or melee uptime.

Killer Sting / Stinging Slash

Either one of the two tankbusters, as in the Savage original. (Killer Sting is a tank stack targeted on the player character with the highest enmity, while Stinging Slash is a pair of cone tankbusters targeted on the highest two player characters in enmity. Each cone is roughly 90 degrees wide, significantly wider than the narrow cones of Honeyed Breeze in the original. All targets are selected at the beginning of the cast.) These hit like a Normal Mode; a DPS or healer can be one of the two participants with a mitigation cooldown or a shield, though if they are close to minimum Item Level they will have very low HP afterwards.

There is no reason not to teach the tankbuster names in the Normal Mode, so as to make stepping up to Savage easier. They even keep their markers in Savage!

Alarm Pheromones (Alarm Pheromones 1)

Moved forward significantly compared to the original timeline, but with the same static "helix" pattern of the Groupbees appearing in sequence beside each other and then dashing through the middle of the arena in the same sequence. (Not including the "bowtie" pattern of simultaneous dashes at the end of the original Normal Alarm Pheromones 1.) This serves to introduce the Groupbees for later mechanics to work with. Once again, strong players can perform both melee and casting uptime optimizations based on this pattern.

Honey B. Live (begins Beat 1)

A raidwide with a long cast that transforms the arena, as in the original.

This is where players see the overhead charm gauge for the first time. Unlike the original, the gauge has four heart slots to match the Savage version, providing more leniency while learning the heart mechanics but also allowing the charm DOTs to tick higher. No hearts are initially assigned.

Filling the charm gauge results in five seconds of the original's Fatal Attraction debuff (charmed, swaying, unable to act) followed by 15 seconds of the Devoted Drone debuff - which nullifies all damage outgoing to the boss. (This does get copied to pets, just like Damage Downs and potions do. And it comes with the "Invulnerable" flying text and shield effect.) The heart buff stacks from the original appear on the boss to help clue players in when this happens, and track how many times this has happened, but they have no other effect. This penalty is significantly harsher than the original, but it also takes much more serious mistakes to fill the charm gauge. There are no mechanics where an attentive player might simply not react fast enough, and no Beat requires any individual player character to take more than two hearts - you will see why later.

(Veteran players may recall the Packet Filter debuffs (placed by Firewall) from Alphascape V4.0 (O12N), Alphascape V4.0 (Savage) (O12S), and The Omega Protocol (Ultimate) (TOP), as well as Pall of Rage / Pall of Grief from Cinder Drift (Extreme). Devoted Drone is slightly different in effect, in that it still allows players to apply debuffs to the boss. Obviously, it is also different in being used as a punishing 100% Damage Down rather than a non-punishing party split. But it is not an accident that Omega exerts a degree of control over the party via its total control of the Interdimensional Rift, and Honey B. Lovely is trying to control the party via her pheromones.)

As in the original Normal, the first time (in the current attempt at the duty) that the player character reaches each number of hearts, red text appears to help teach how the charm gauge works. The messages are the same as in the original, simply adding more to match the Savage debuff range:

• One heart: "You are growing infatuated with Honey B. Lovely..."
• Two hearts: "You are head over heels for Honey B. Lovely..."
• Three hearts: "You are hopelessly devoted to Honey B. Lovely..."
• Four hearts: "You are powerless against your undying affection for Honey B. Lovely!"

Spread Love

This is a completely basic spread, with fairly small AOEs, except that each AOE gives a heart. This serves to teach players that any mechanic with a heart on it adds a heart to their charm gauge when taken, in an unavoidable way since every player character present and not KOed gets a spread AOE. To reinforce the lesson, this will always trigger the first red text.

At this point, each player character should have one heart. If players did not spread effectively, some may have two or three hearts.

Outerstage (not a combo, this is not Savage)

This is a completely basic point-blank AOE, except that it has a heart on its telegraph and gives a heart to any player character hit by it. This serves to teach players that avoidable mechanics can also have hearts on them.

I have deliberately chosen not to use the cast name Loveseeker, which the original uses for this same AOE, in order to help the player learn the Combos if they choose to attempt the Savage version.

Love Me Tender (Moving Hearts)

Honey B. Lovely begins emitting moving hearts from herself in the same fixed 3-pronged spiral pattern as the original. (For reference, the prongs are evenly spaced, one always starts directly South, and they always rotate clockwise.) Unlike the original, no other mechanics appear, neither the puddle AOEs baited under player characters nor the baited moving hearts from the edge of the arena.

The point of this mechanic is not to be random and reactive. Indeed, the Outerstage before it is deliberately not random, to push attentive players out of the center and thus give them time to see the moving hearts spawn.

The point of this mechanic is to express "she is literally emitting pheromones to try to charm you" in a way that is consistently dodgeable. For first-time players, this solidifies their understanding of the charm gauge, allowing later mechanics to build on it. This pattern also rewards thinking about whether to slip between hearts in the spiral arms, or follow the arms at the expense of more overall movement, again allowing strong players to optimize based on their jobs.

She stops emitting hearts after the spiral has rotated 120 degrees, so every angle has been passed over by one arm. There is nowhere to stand that avoids engaging with the mechanic at all.

Honey B. Finale (ends Beat 1)

A raidwide that transforms the arena back and cleanses all hearts, as in the original.

Yes, Beat 1 is not all that hard, but it is also pretty short while still teaching first-time players how the charm gauge works. The duty is not done ramping up yet - it has only just introduced Honey B. Lovely's recurring mechanical elements.

Stinging Slash / Killer Sting

Whichever tankbuster was not used earlier. While Duty Finder does not make mechanic-specific plans and so will generally not plan invulnerability uses accordingly, this ensures that the first-time player sees both tankbusters before they can unlock the Savage version.

Alarm Pheromones (Alarm Pheromones 2)

Two Groupbees appear at once in random spots at the edge of the arena, each locking onto a random player character. (They cannot both select the same player character, but it is otherwise completely random. In the event that only one player character is present and not KOed, only one Groupbee will appear.) The selected players are marked, not only with a tether to the Groupbee locked onto them but also with an overhead symbol for which tether they have: the "Playstation markers" that were previously exclusive to Endwalker high-end duties. (Which two markers are used is also completely random.)

https://i.imgur.com/MTS6UjD.jpeg

Duty Finder does not do the sort of planned coordination of multiple tethers that has been historically associated with Playstation markers, nor are they particularly good at remembering which marker they got for later. I do not expect them to do either. But they can learn "you have to do something specific with whoever or whatever has the matching symbol", which is the "real" underlying concept of the markers - and this duty will start teaching that right now, by combining the already-known tethers with the Playstation markers over both the player character and the Groupbee on each end.

Players then have several seconds to figure out who has the tethers - this is Normal Mode, after all, they cannot be expected to pre-spread - and move as they please. After this time, each Groupbee will set their path to go through their selected player character's position, shown by their tether and marker disappearing and being replaced with the Groupbee line. In case players were unsure of what tethered Groupbees do, this shows them and gives them time to dodge, but it is not a difficult guess since they have seen the Groupbees earlier.

As soon as these Groupbees complete their Blinding Love cast and begin moving, two more Groupbees appear and lock onto player characters, again in random spots at the edge of the arena. They still cannot both select the same player character, and this time - if at all possible - they will select player characters who were not selected last time. These two Groupbee tethers will use the other two Playstation markers, ensuring that the party sees the whole set.

Just as before, these two Groupbees will bait onto their tethered player characters, with the same timings.

As soon as these Groupbees complete their cast and begin moving, four more Groupbees appear and lock onto player characters, again in random spots at the edge of the arena. They still cannot have more than one Groupbee select the same player character at once, and this time they will prioritize selecting player characters who have not been selected by either previous set. As this totals eight Groupbees, this ensures that every player must bait a Groupbee (unless their character is KOed for the entire mechanic), forcing them to interact with the Groupbee that has the matching symbol. (If player characters are KOed or disconnected for the entire mechanic, that will just result in someone else baiting twice, it is not a body check.)

By forcing the interaction with the Groupbee, this mechanic as a whole builds the association for players between seeing the Playstation markers and expecting to interact with whoever or whatever has the same symbol. This also shows players that the entire set of Playstation markers can be active at once.

Just as before, these four Groupbees will bait onto their tethered player characters, with the same timings.

As a whole, this mechanic is fairly slow. (Unless some player characters are KOed or disconnected, each player character is only selected once out of the three sets.) This is deliberate, to allow players to focus on and process the new concept of Playstation markers until they are comfortable with it. Even the players not currently selected should still pay attention to the mechanic, to know where the Groupbee divebombs will be and save themselves from any poorly placed baits. (Notice that there are always active Playstation markers until the final set of Groupbees sets their paths.) This also encourages players to learn from each other by observation.

Meanwhile, for players who are comfortable with the concept, the high randomness creates many different possibilities to plan for and try to optimize. Want to bait your Groupbee without leaving melee range or making everyone move? You can do that on an individual level, but your Groupbee can spawn anywhere so it is not going to be the same movement every time, and that is before accounting for the other Groupbee(s) in the same set. Want to bait your Groupbee without losing casting uptime? You can do that, but you are not always going to have your bait at the same time or in the same location, so you have to plan instant casts for each possible time, bait in melee range, or do quite a lot of slidecasting if your Groupbee spawns far away from you. Coordinated raid groups can devise strategies like "always leave the dead center of the arena safe" and have a completely different experience from casual parties who bait anywhere and everywhere.

Honey B. Live (begins Beat 2)

Once again, this raidwide transforms the arena and brings out the charm gauges. No hearts are initially assigned.

Love Me Tender (Towers)

Four player characters are selected at random and given Playstation markers directly above their charm gauge. (Which is why I put Alarm Pheromones 1 before Beat 1 - to introduce and teach each overhead marker separately as they alternate. This also utilizes the fact that the charm gauges sit directly above each player character's nameplate, allowing for a second overhead marker without piling them too high.) If less than four player characters are present and not KOed, all player characters that are not KOed receive Playstation markers.

Each player character with a Playstation marker also simultaneously receives a new debuff, "Command Performance: [shape]", where the shape is that of their Playstation marker. The debuff icon will also show the marker shape. This debuff is cleansed by taking a tower with the matching shape, and KOs the player character if it expires rather than being cleansed. Here, the Playstation markers do not disappear until the debuff expires or is cleansed.

(For those who enjoy linguistic references, here is the definition. "Command performance: (archaic) A dramatic, musical, or similar entertainment performed before a monarch or other head of state, especially in a circumstance where that ruler has requested or ordered the performance." Fitting for the queen bee!)

At the same time, a heart tower spawns matching each active Playstation marker. (This is again not a body check, since it will not spawn more towers than player characters who can take them.) These towers show the heart within the base, like the original, and the Playstation marker floating at about overhead height for the taller player characters. (This positioning is to give clear visibility of both the heart and the Playstation marker.) These towers all spawn in melee range, because at this point the goal is to get players to do their specific assigned mechanic without being literally forced to, for which looking around the arena in a panic to find their tower or worries about losing uptime are counterproductive. Melee DPS may have to press True North if they cannot avoid positionals while in their tower, which is fine.

To reinforce that only the player characters with Playstation markers are meant to take these towers, all player characters without Playstation markers receive heart spread markers that are immediately visible and go off at the same time as the towers. (If a player with a Playstation marker is in the process of disconnecting or has been pulled away by an Earth emergency, a player whose character has a spread marker can step into their tower to prevent it from exploding, but that is clearly not what is suggested by the markers.) Since these spread markers are fairly small as in the original, these players can fit their spreads between the towers without leaving melee range.

Even without having just seen the baited Groupbees in Alarm Pheromones 2, or reading the Command Performance debuffs, a player who sees a marker over their head and the same marker over a tower is quite likely to correctly guess that they should stand in that tower. Intuitively, this further nudges the players with spread markers out of the towers, so as not to hit those taking the towers. (It is still possible for a tower to go off with no one in it, exploding and giving all player characters a heart, but it is quite unlikely barring disconnects and emergencies.)

Which means that here - unlike in the original - Duty Finder is actually capable of putting exactly one player character in each tower, the way the charm gauge encourages! The Playstation markers being assigned solves the coordination problem of who should take each tower for the party, which stops players from messily reacting to each other.

(Veteran players may recall that the Looper debuff appears in Alphascape V3.0 (O11N), not just in Savage and Ultimate, and requires players to take any tower in order to cleanse it and avoid KO. The use of Command Performance here is deliberately similar to how Looper is used in O11N, particularly in the fact that players do not need to notice the debuffs at all - any reasonable attempt to engage with the mechanic will cleanse the debuffs automatically. Command Performance goes one step further with its name and icon, in that players who see the name and/or icon in their flying text will tend to correctly guess what the debuff requires of them without having to hover over it for a description. This is great for controller players, or really anyone who has trouble hovering over debuffs in the middle of combat.)

This mechanic teaches players that the possible interactions when they see a Playstation marker are not limited to divebombs, or even to tethers. In turn, this prevents them from thinking that the Playstation markers are for a concept they already know. (Indeed, even for veteran raiders I have expanded the concept of Playstation markers; its prior uses always involved tethers of some variety.) If I had presented the Playstation markers only with the Groupbee tethers, players might reasonably think that the markers are a replacement for existing divebomb markers or tether types. If I had presented the Playstation markers only with the heart towers, players might reasonably think that the markers are exclusive to towers or charm mechanics. Showing both clarifies that the marker set is a pairing tool, and the player should try to work out the interaction with whoever or whatever they are paired with.

Once the towers and spreads go off, every player character has at least one heart, or is about to be KOed by their Command Performance debuff. If they did the intuitive thing, they will all have exactly one heart.

Centerstage (not a combo, this is not Savage)

This is a completely basic donut AOE, except that it is pink with hearts on its telegraph and gives a heart to any player character hit by it. For players repeating the duty, this encourages not spreading too far in the first place, and for everyone this creates an opportunity for healers to heal the whole party.

Love Me Tender (Towers)

As the previous Love Me Tender, except prioritizing players who did not get Playstation markers then to get them this time. This ensures that every player must take a tower unless their character is KOed for both Love Me Tender casts.

Once this resolves, every player character has at least two hearts, or is about to be KOed by their Command Performance debuff, or died earlier. If there were only three heart slots in the charm gauge, players would have to do it exactly right while still learning a new concept, which is why I put four heart slots in the charm gauge instead. The extra heart allows for each player to save a tower from exploding, or for one tower to explode.

This is not a hard mechanic, but it continues to teach the concept of Playstation markers while expressing Honey B. Lovely's fighting style. It creates a very evocative tension between having to take hearts in order to keep fighting and wanting to avoid filling the charm gauge. Lines like "Who's my obedient little bee?" and "Heed me well or you're as good as dead" have a lot more sting to them when each player is singled out to do a very specific thing - to play the part she dictates in her performance - or take a KO.

Honey B. Finale (ends Beat 2)

A raidwide that transforms the arena back and cleanses all hearts, as in the original.

Splash of Venom

Stores a spread for later, as in the original.

Players might not see this as gear improves, particularly with this being the first tier of an expansion and item level syncs going by level rather than by raid tier. That is okay. This is not that exciting unto Duty Finder, especially since the spreads have to be telegraphed for players to learn it in Normal Mode.

The primary purpose of this section is to be a spacer between the Playstation marker mechanics, so that players have some time to breathe and process what they are learning. The secondary purpose is to teach the AOE patterns to any players who will be trying the Savage later.

Tempting Twist / Honey Beeline

Either AOE at random, with Splinter honey blobs afterwards. Unlike the original, the honey blobs are in the Savage patterns, since both patterns require players to move after dodging the cast and familiar patterns help players make the step up to Savage. This is also more random than the original, which forces Tempting Twist to be first in order to have Honey Beeline overlap with its Alarm Pheromones 3.

As the castbar finishes, the spread markers for Splash of Venom appear. These resolve a few seconds after the honey blobs.

Drop of Venom

Stores a stack for later, as in the original. Duty Finder does not do quick partners. (Yet, anyway.)

Honey Beeline / Tempting Twist

Whichever AOE was not used last time, with Splinter honey blobs afterwards in the Savage patterns.

As the castbar finishes, the stack marker for Drop of Venom appears. This resolves a few seconds after the honey blobs.

Killer Sting / Stinging Slash

Either one of the two tankbusters, as before.

Alarm Pheromones (Alarm Pheromones 3)

As with Alarm Pheromones 2, except with two sets of four Groupbees each. Players have already had the Playstation markers on Groupbee tethers introduced slowly by this point, so they can now handle the faster pace.

Stinging Slash / Killer Sting

Whichever tankbuster was not used earlier.

Honey B. Live (begins Beat 3)

Once again, this raidwide with delayed damage transforms the arena and brings out the charm gauges. No hearts are initially assigned.

Love Me Tender (Towers)

As in Beat 2.

Love Me Tender (Moving Hearts)

As in Beat 1.

Love Me Tender (Towers)

As in Beat 2, once again prioritizing players who did not get a Playstation marker last time to get one this time. Castbar starts as soon as Honey B. Lovely finishes emitting moving hearts, since now players have seen all these mechanics. This encourages players to plan for the moving hearts pattern rather than reacting to it on repeat runs, without being too harsh on first-time players or those who have not noticed the fixed pattern.

Honey B. Finale (ends Beat 3)

A raidwide that transforms the arena back and cleanses all hearts, as in the original.

At this point, the duty loops from Splash of Venom until the boss reaches 20% HP.

Once she reaches 20% HP, she transitions into a new phase. This transition must happen without the charm gauges, so if necessary she will immediately end a Beat (and cleanse any active Command Performance debuffs) with Honey B. Finale.

Rotten Heart

A raidwide that brings with it a new arena transformation, as it does in Savage.

No, Duty Finder cannot do the Savage version of Rotten Heart. Yes, she is still going to drop the act when she gets desperate. (This is not out of any objection to high-end duties having exclusive phases, it is to create a final phase that cements casual players' understanding of Playstation markers.)

Player characters are immediately paired at random with the "get away from each other" blue arrow tethers from Eden's Promise: Litany (Savage) (E10S). Each tether pair gets a Playstation marker, which does not disappear until the tether is removed, along with a light DOT debuff that is weaker the further the tethered player characters are from each other. Casual players will mostly notice the high-contrast arrows on the tethers rather than the DOT, but that suffices to get them to intuitively move away from their tether partner.

After a few seconds, the Playstation markers for all tethers other than one's own disappear visually, to reduce visual clutter while still reinforcing that the markers are a set.

Alarm: Mass Final Sting

Honey B. Lovely then summons a batch of Groupbees, with an elaborate animation lasting several seconds and locking herself in the center of the arena for the duration of their flight. But rather than the Groupbees dashing across the arena, this time she orders them to suicide bomb the arena, stingers first. Their descent creates pairs of widely separated towers, each pair sharing a resolution time and a Playstation marker. (Since there are no hearts here, the Playstation markers are shown in the base of each tower, large enough to be seen through the tethers. They are not also shown over the towers to avoid confusion with the overhead markers on other players.) One pair will spawn at a time, repeating until four pairs have spawned. (This remains true even if some player characters are KOed or disconnected, as there are no hearts to create a body check from extra towers.) As usual for towers, if a tower resolves with no one in it, it explodes dealing raidwide damage.

At this point, each pair of players should already have spread their characters apart before or during the Groupbee summoning animation, making it easy for the players to split up their corresponding pair of towers. (Which is why the animation is so long, unlike in Alarm Pheromones.) Here, the Playstation markers do not completely solve the coordination problem for the party, but they reduce it to coordinating with one other person, and Duty Finder can manage that from previous experience with spreads and towers.

Each tower spawns just close enough for players to stay in melee range by standing in the portion closest to the center of the arena, with its partner tower directly across from it. (Again, melee DPS are rewarded for planning their uses of True North to have it for their tower.) These towers spawn at angles half-way between a cardinal and an adjacent intercardinal, much like the tower placements in the original Savage Beat 3, and all eight of these spots will be filled once all towers spawn. Observant players on non-melee-range jobs are thus encouraged to use their greater range to stretch their tethers further.

A rough diagram of the tower positions:

https://i.imgur.com/mX3zNlB.png

Taking a tower matching one's Playstation marker clears one's tether and DOT (for both tethered player characters). The tethers and DOTs do not otherwise expire. Here there are no Command Performance debuffs because Honey B. Lovely is no longer trying to put on a show, but now only to defeat the party.

(And yes, the cast name is a deliberate reference to BLU's Final Sting. While this version is not single-target in the same way, it is still bees' suicide attacks and ideally one player character matches up with each bee to soften the blow.)

Call Me Honey

Cast only starts once all towers have resolved.

Still a raidwide, which now also reapplies the tethers from Rotten Heart to any players who have not carried them over. The tethers are freshly randomized except for not reusing markers that are still active.

Alarm: Mass Final Sting

As before, except that two pairs of towers spawn at a time. If any player characters are KOed, this is a great big hint to use battle raises in a hurry, since all towers that spawn at once will resolve at once.

Call Me Honey

As with the previous cast, a raidwide that reapplies the Rotten Heart tethers.

Alarm: Mass Final Sting

As before, except that all eight towers spawn at once. Now Honey B. Lovely is quite fed up with the party's persistence, and has finally created a soft body check - avoiding KO from multiple simultaneous tower explosions is possible but requires efficient healing and mitigation. (Plus Normal Modes have even had hard enrages before - Deltascape V2.0 (O2N), for example - just with timers long enough that the party has to have performed very badly to see them.)

At this point, the duty loops from the last Call Me Honey until either side is entirely KOed. 20% might sound like a lot of HP for this phase, but DPS Limit Breaks and gear improvements are likely to shorten it even if the boss does not need to first end a Beat. AAC Light-heavyweight M4 (M4N) has a similar threshold, and strong parties often barely even see the Wicked Hypercannon after the long arena transformation nowadays. I would have put the threshold higher if the transition into Rotten Heart was longer, since I want the party to actually see the tethers between player characters even if they do not see the Groupbee towers.

--------

If I were not aiming to teach a new mechanical concept, I would not set a pace quite this slow for level 100 gameplay, especially outside the MSQ. But Honey B. Lovely's fighting style was an opportunity too good to pass up to teach Playstation markers - she is a performer who fights by enthralling her opponents, of course she wants to single out specific player characters for specific parts in her performance! (Plus, Playstation markers already facilitate high randomness.) So it is thematically appropriate for her to use Playstation markers often enough to guide the casual player into picking up the concept, without needing an allied NPC to instruct or demonstrate. And even a duty this slow can have plenty of sting to it.

Not every duty should have some brand new mechanical concept; even the most dedicated players need time and practice to consolidate their knowledge, and teaching new concepts every duty leaves no room to explore their forms and combinations. But taking opportunities to teach and reinforce new concepts could significantly expand Duty Finder's mechanical vocabulary over the course of several patches. (Especially if casual players in all roles learned to use the debuff bar.) And since mechanical concepts can combine, this would greatly expand the duty designs available for casual play.

Create a new Wild Charge marker and teach it, as I described when discussing the final phase of M4N? Bring the "[number] in Line" debuffs to Duty Finder? Perhaps even teach spinner baits after a few patches, so long as the spins are slow enough that players can dodge through a failed bait? There are lots of possibilities, especially if you continued updating the Hall of the Novice.

Playstation markers are also a good choice to teach casual players because they can then be used to present gentler versions of many mechanical structures that would otherwise necessarily be limited to high-end duties. Fundamentally, in regular duties they can solve coordination problems for the party, allowing the party to tackle the mechanic without having to specifically prepare for it. You could have dungeon bosses where every party member has an additional mob (add) to tank and position, even though non-tanks do not have the tools to get an add's attention if the tank goes for the wrong one - just tell them which add to take! You could bring true enumerations back to Duty Finder, this time without players having to react to each other. Show casual players role partners by giving them their role partners. And so on. That would get more players into high-end duties without making the high-end duties any less of an accomplishment - by instead making the regular duties better stepping stones to the higher difficulties.

----------------

Now for the Savage Mode. As a Savage timeline cannot be properly evaluated without context on the overall tuning of the duty, I shall provide the context for the design I intend first.

This is a second Savage floor. Regardless of job balance, it should not have a damage output requirement almost as high as the final floor of the same tier - especially not since second floors have an Item Level requirement five Item Levels lower than third and fourth floors of the same tier! The tuning I intend is that if all party members are at minimum Item Level, the party can make at most one major mechanical mistake (KO, Damage Down, or Devoted Drone) and still beat the enrage with otherwise perfect play. (Of course, the vast majority of groups will not have otherwise perfect play, thus leaving them no room for major mechanical mistakes at minimum Item Level.) Since crafted gear is five Item Levels above the minimum, parties would not be reliant on raid drops or tomestone gear for further leeway even in Week 1.

As for the healing and mitigation requirements, I intend to require planning but not perfection, with an overall result not that different from the original. Even though new raiders are encouraged to start with Extreme before moving up to Savage, this could be a new raid healer's second high-end duty ever. On top of that, a new raid healer is likely progressing in Party Finder with variable cohealers and mitigation from their parties, so requiring optimized healing cooldown plans is a recipe for quitting in frustration. But at minimum Item Level, mitigation from someone in the party is often required (outside the Beats) to reduce the damage taken from a single cast below player characters' max HP... and having a shield healer Global Cooldown (GCD) heal for every mechanic will not be pleasant.

That is at baseline. Gear will make things easier. But unlike the original, I will not have the heart stacks from filled charm gauges make Honey B. Lovely deal more damage. (Instead, I will add to the baseline healing and mitigation requirements in selected places.) This is for two reasons:

First, because the burden of the increased healing requirements from filled charm gauges falls primarily on the weakest and newest raid healers, who are least able to handle it. Strong players and/or those who already know the mechanics well will tend to seek out other players of similar skill and experience via forming static raid groups, creating Duty Complete parties, and setting higher Item Level requirements. (Especially since Raid Finder matchmaking is not in customary use outside the JP Data Centers.) Thus, the player population that joins parties with weaker and newer raid healers will be more likely to fill their charm gauges than average - and will need more adjustments to the healing as a result, between lower Item Levels and less effective mitigation use! Meanwhile, the strongest raid healers, who can adapt their healing plans easily, rarely see a charm gauge fill (during an attempt worth continuing) even in the original; with the reactive elements of Beat 1 replaced, they would almost never see a charm gauge fill, let alone enough in a single attempt to change their healing plans.

Second, because some players raid to learn a plan and push themselves to execute it as close to perfectly as they can, for the enjoyment of the dance... and some of those players are drawn to healers for a variety of reasons, despite the necessity of compensating for other players' mistakes in the healer role. (Going into those reasons would be a long detour, so I will skip it here.) From that perspective, having to change one's healing plan for the entire rest of the attempt because of how the boss works is an annoyance. (In contrast, Doping Draught is great in AAC Light-heavyweight M3, both Normal (M3N) and Savage (M3S), because it is a planned part of the duty.) Healers who enjoy adapting their healing plan based on how an attempt is going can still do that without the boss requiring it directly - and some recovery scenarios already require it indirectly, by forcing the healer(s) who can act to use resources they were saving for later mechanics.

(There may well be a place somewhere for failed mechanics giving the boss damage up stacks! But it is not here, and it is not where players are learning a new concept for the first time in Normal Mode.)

That context given, here is the timeline.

--------

Redesigned Savage Mode Timeline

0:05 - 0:10 Call Me Honey

Remains as in the original. While not every duty needs an opening single-hit raidwide, it is fine to have one in some duties, and both AAC Light-heavyweight M1 (Savage) (M1S) and M3S have different forms of opening party damage.

At minimum Item Level, this hits just hard enough to reliably KO Disciples of Magic and sometimes (damage variance) even Disciple of War DPS from full HP if it goes completely unmitigated and unshielded. We are in Savage now, time to bring out mitigation tools and prepull shields, or you are going to get sent back to the start 10 seconds in. On its own, this could even hit a little harder, but it comes back in Rotten Heart and I want players to learn how hard it hits here for heal planning there. (It does not come back before Rotten Heart due to the abundance of Beat transition raidwides, as in the original.)

More damage is coming, so ideally the party will use the mitigation tools that are least disruptive to opening burst here (rather than making the healers do more work) and save the rest for later. Mitigation spreading does not need to be perfect, this is only a second floor, but it does need to exist.

0:12 - 0:17 Splash of Venom / Drop of Venom

As in the original, stores a random untelegraphed spread or partner stack for later. When this resolves, it will hit just as hard as Call Me Honey; either the party needs to find some mitigation and efficient healing, or the healers will be scrambling to heal the party up after their openers.

While this is the same memory mechanic as in the original, this design uses it significantly less often - thus making its presence less repetitive and boring while making room for other forms of mechanical difficulty. In particular, the Beats will not use it at all, which helps mechanically differentiate the "enthralling songstress" aspect of the Beats from the "venomous queen bee" aspect of the other mini-phases. This mechanical differentiation in turn helps give the Beats a "wide open stage" feeling, contrasting the restricted spaces that other mechanics create.

0:23 - 0:28 Tempting Twist / Honey Beeline

As in the original, casts a random ground AOE with follow-up Splinter honey blobs that force players to move. This specific cast retains the extended break between casts of the original, to allow players to finish their openers before moving.

The Splash of Venom / Drop of Venom resolves at the same time as the honey blobs (0:33), as in the original.

0:40 - 0:44 Killer Sting / Stinging Slash

Either one of the two tankbusters, as in the original. Now that we are in Savage, these hit like it - at minimum Item Level, a tank participant will be KOed from full HP unless they use personal mitigation.

I have moved this tankbuster ahead in the timeline from the original, so as to create a space between the first and second rounds of the memory mechanic while still ensuring that players learn both possible stored damage casts and Splinter honey blobs early. This encourages healers to use their regens to heal up from the first Splash of Venom / Drop of Venom, which will serve them well in the new Beat 1 where the heart DOTs will tick for longer.

0:45 - 0:50 Drop of Venom / Splash of Venom

Whichever stored damage cast was not used earlier, as in the original.

0:52 - 0:57 Honey Beeline / Tempting Twist

Whichever ground AOE cast was not used earlier, as in the original. Once again, the stored damage resolves at the same time as the honey blobs (1:02).

Here, healers are encouraged to use burst heals before the upcoming Beat.

1:10 - 1:19 Honey B. Live: 1st Beat

A raidwide that transforms the arena, as in the original. Does not hit as hard as Call Me Honey, to prevent immediate KOs from heart DOT ticks. (As a whole, the Beats have a different style of healing and mitigation requirement: lighter big hits, but continuous DOT damage from the heart stacks.)

During each Beat, the boss centers herself facing North and remains there, since she is performing and intends to be the star of the show. This contrasts with her fairly free movement during the Alarm Pheromones mechanics and around the tankbusters, when she will follow the player character with highest enmity so long as she is not actively casting.

Just like the original, the charm gauge has four heart slots. Eight hearts are initially assigned across the party, but with the following rules not present in the original:

• One tank will get two hearts, while the other tank has none. (This encourages attentive tanks to split their incoming damage, with one tank taking the heart DOTs from the beginning and the other tank taking the auto-attacks.)
• The healers will get two hearts between them, either one each or both taken by one healer.
• The DPS will get four hearts between them, in any combination such that no player character has more than two hearts.

All player characters also get a Command Performance debuff with no shape attached; to cleanse this, they must take any heart tower. It is now up to the party to decide who takes which tower(s), but no one can sit out of the mechanic.

Filling the charm gauge has a similar penalty to the redesigned Normal: five seconds of the Fatal Attraction debuff followed by 25 seconds of the Devoted Drone debuff. (The heart buff stacks also appear on the boss just like the redesigned Normal.) This allows players to progress despite being charmed, while punishing mistakes while attempting to clear the duty, as one would expect from a Damage Down - and players care a lot more about their damage output in Savage! The length of the penalty is meant to have a similar effect on damage output to a KO (time taken to receive a raise and in the raise animation, then 25% lower damage and healing for 100 seconds) without draining one's resources or affecting the party's healing and mitigation plan. Indeed, Devoted Drone allows players to debuff the boss precisely so that they can still apply debuff-based mitigation, with the side effect that it also permits debuffs that increase party damage output.

(Filling the charm gauge could have healed the boss like it does in the original; that does not have the same disproportionate effects that the damage up does. But I find being unable to damage the boss more fitting to being enthralled, while the boss healing evokes more of an "aether drain" concept. What if the boss healing was on Wicked Thunder instead, since she is trying to take the player character's soul?)

1:24 - 1:28 Love Me Tender (Towers)

12 Fracture heart towers spawn one by one, separated by the same 2 second intervals (1:28 - 1:50) as in the original. Unlike the original, there are 12 positions for the towers, where each position will be used exactly once in a completely random order. The positions are:

• four "close" towers on the intercardinals, well within melee range, like in the original;
• four "mid" towers on the cardinals, where only the parts closest to the center of the arena are in melee range;
• four "far" towers on the intercardinals, that require leaving melee range to take.

A diagram of these tower locations:

https://i.imgur.com/SjggZdB.png

The fixed set of tower positions allows parties to assign who will take each tower position based on the initial heart assignments, which in turn allows parties to eliminate all reaction to one's party members despite the completely random order. (Recall that this is necessary to prevent pressuring the party's reaction times.) Meanwhile, players trying to optimize their damage output have many possibilities to consider for when they will have to move to take their tower(s), including the possibility of taking consecutive towers.

Here is one solution that the party might agree on, with the tank with zero initial hearts taking the T1 spots:

https://i.imgur.com/JZS3WfH.png

(In this solution, no DPS will ever need to move twice, despite the melee DPS being marked on both priority orders. If either priority reaches the third DPS listed, that is because the first two listed both started with two hearts, and therefore they have all the initial hearts that can be assigned to DPS. That means the other two DPS started with zero hearts, and will pick up one tower each beyond the requirements of Command Performance. (Indeed, the priority ordering in this solution can be read as including all four DPS for each tower, but the fourth player will never be reached.) As a result, each individual player has all the information they need to know which tower(s) they will take as soon as they see the initial heart assignments, which are given 19 seconds before the first tower resolves, and they can take some time working through it without messing up the priority order.)

Each tower resolves 10 seconds after it spawns (1:38 - 2:00). The 10 second period between spawn and resolution will be the same for all heart towers throughout the duty.

The Command Performance debuffs expire two seconds after the last tower resolves (2:02) - slightly longer than needed to ensure that the debuffs are cleansed even by the last tower, just long enough for a player who did not take a tower to realize their mistake.

(As a result, players still engage with Command Performance differently from the existing uses of Looper, despite the former appearing without the Playstation marker shapes. In all of O11N, the first Executable in O11S, and Run: ****mi* (Sigma Version) in TOP, the whole party must cleanse Looper at the same time. In the second Executable in O11S and Program Loop in TOP, the party has the cleanse order dictated to them by the differing lengths of Looper given by the boss. Here, each player character cleanses Command Performance at a different time, but the order is up to the party.)

At this point, the party has a total of 20 hearts if exactly one player character stood in each tower.

2:05 - 2:11 Centerstage Combo / Outerstage Combo

Either of the avoidable ground AOE sequences, as in the original. (Second and third hits at 2:14 and 2:17 respectively, the same three second spacing as the original.)

Unlike the original, most of the heart DOTs have already been applied, so healers should consider taking advantage of the donut hit to heal the party despite this being a burst window. (My thoughts on the 2 minute meta aside, duty design and job design are tightly linked, and the jobs we have in Dawntrail follow synchronized 2 minute bursts.)

2:21 - 2:25 Love Me Tender (Stack)

As in the original, the four-heart Heartsick stack spawns targeting a random player character out of those with the lowest number of heart stacks among the party. It hits lightly at 2:30, with its real danger being that it forces the party up to the maximum of 24 hearts without any player character filling their charm gauge.

There are no "freebies" or "spare hearts" here, unlike the original Beat 1. Since the party can fully plan out the forced hearts and dodge all the avoidable AOEs without reacting rapidly, there is no need for lenience on the mechanics dance. The party has some degree of choice on who will take the stack - anyone who was initially assigned two hearts must get their third heart from fulfilling their Command Performance - but no room for mistakes.

Wise parties will agree on a position to take this stack, and skilled players will start moving towards / away from this position during the preceding Combo depending on whether they need to take or avoid the stack.

2:30 - 2:36 Outerstage Combo / Centerstage Combo

Whichever ground AOE sequence was not used earlier, as in the original. (Second and third hits at 2:39 and 2:42 respectively.)

Like the original, all of the heart DOTs are now ticking, so healers should utilize the donut hit out of burst to heal.

Unlike the original, there is no "spacer" between the Heartsick stack and the start of the Combo castbar, as there are no moving heart AOEs that need time to dissipate. Players can choose to take the stack in a position that leaves them set up for the Combo, especially since they know which one it will be.

2:49 - 2:54 Honey B. Finale (ends Beat 1)

A raidwide that transforms the arena back and cleanses all hearts, as in the original. Again does not hit as hard as Call Me Honey, since shields will be worn down by the heart DOT ticks.

2:57 - 3:02 Splash of Venom / Drop of Venom

As before, stores a random untelegraphed spread or partner stack for later.

This is a good time for healers to use regens to heal up from the end of Beat 1.

3:07 - 3:11 Stinging Slash / Killer Sting

Whichever tankbuster was not used earlier. This allows tanks to have branching personal mitigation plans if they so choose. In particular, a tank can guarantee that they use their invulnerability on Killer Sting either here or the preceding tankbuster, allowing them to take it alone with no further mitigation.

3:16 - 3:21 Tempting Twist / Honey Beeline

As before, casts a random ground AOE with follow-up Splinter honey blobs that force players to move.

Splinter and the stored damage resolve at 3:26. The party has a moment to set up healing before...

3:34 - 3:38 Alarm Pheromones (Alarm Pheromones 1)

Four Groupbees spawn around the edge of the arena, spaced 90 degrees apart. They can appear on either all cardinals, all intercardinals, directly between each cardinal and its clockwise adjacent intercardinal, or directly between each intercardinal and its clockwise adjacent cardinal. (These are the same 16 possible positions as in the original, divided into the sets from the original Alarm Pheromones 2.) At the same time, the closest player character to the boss is targeted for a four-player-character multi-hit Enumeration, visually marked.

2.5 seconds later, another four Groupbees spawn, adjacent to the previous ones but either clockwise or counterclockwise at random. (One randomization that applies to all the Groupbees for this mechanic.) This repeats two more times, so the entire edge of the arena is lined with Groupbees 7.5 seconds after the castbar ends. (Spawn times 3:38, 3:40.5, 3:43, 3:45.5.)

Here is a diagram of the Groupbee spawn groups, each group marked with a different in-game shape sign:

https://i.imgur.com/RVPuxzq.png

10 seconds after each Groupbee spawns, it will lock onto the closest player character to itself and set its path to go through that player character's position, shown by the Groupbee line appearing as it starts its Blinding Love cast. Thus, the first set of Groupbees will lock on 2.5 seconds after the last set of Groupbees appears. (Lock-on times 3:48, 3:50.5, 3:53, 3:55.5. Groupbee casts complete 3:54.5, 3:57, 3:59.5, 4:02.)

As each set of Groupbees locks on, the Enumeration hits once, leaving behind a venom puddle underneath it that lasts for 22 seconds (just like the player-baited puddles in the original Alarm Pheromones 2). At minimum Item Level and even somewhat above it, the four hits of the Enumeration combined will reliably KO any non-tank from full HP, and will be painful even for tanks when combined with auto-attacks resuming after the Alarm Pheromones cast completes. (So unlike the original, there is a healing requirement here even with perfect play.)

Now that I have described what the mechanic does directly, let us go over what it suggests to the party as a solution, and what that solution asks of players.

If the Groupbees appeared by themselves or with a stack, strategies could easily develop that have them baited by half the party (one player character beside each Groupbee), two players (characters positioned opposite each other at the edge of the arena, weaving between the Groupbee lines), or even the whole party (skew from the center and move as a group). But I do not want any players to wind up standing still, nor do I want the whole party moving together - hence the Enumeration and puddles, which make it difficult to leave two player characters out of everything and useless to soak the damage with the whole party. Veteran raiders will recognize that the mechanic suggests having half the party bait the Groupbees, and the Enumeration marker confirms that to them.

(It is still possible to do the other bait strategies, making this mechanic not a body check. But that is not easy or advised.)

In theory, the party can split who baits the Groupbees and who takes the Enumeration however they like. But since the Groupbees are all spread out and the boss has an appropriately small hitbox, parties will intuitively give the Groupbee baits to Disciples of Magic and physical ranged DPS. This has several effects:

• It puts the healers away from the player characters taking heavy damage, creating a distance party heal check even after getting some mitigation / healing from the tanks and melee DPS. Distance healing is a way to make the puzzle of heal planning more interesting, encouraging healers to think about what parts of their toolkit they want to use when. Yet there are lots of ways to solve this particular distance party heal check, making it suitable for new raid healers: any SCH fairy heal (especially Consolation), Holos, Panhaima, Philosophia, any "healer bubble", manual or timeout detonation of Liturgy of the Bell, any GCD party shield, any GCD regen if cast shortly before the baits (utilizing the regen ticks), even Cure 3 will work. If both healers and any tank put something on this, the party will be fine, and they really do not need to put a healer in the Enumeration to find ways to do it. (They still can put a healer in the Enumeration, especially in a single melee party, but it is not that helpful once the healers have thought about what tools work for distance healing.)
• It encourages the Groupbee baiters to think about prepositioning near the edge of the arena, in whatever quadrant the party has assigned them. (If I wanted to prevent the Groupbee baiters from prepositioning towards their quadrants, I would have the Enumeration target the furthest player character from the boss instead. But that is more appropriate for higher Savage floors or Ultimates.) Magical ranged DPS and physical ranged DPS can do this with no downside so long as they do not get hit by Stinging Slash, while healers have to weigh their prepositioning against the range and timing requirements of the healing tools they plan to use.
• Meanwhile, it encourages those taking the Enumeration to think about working in a small space and utilize the ability to move underneath the boss. Parties have the choice of whether they want the puddle baits to be done by a tank or a melee DPS (or another role), but if they want positional uptime they will have to move in lockstep regardless. And since the puddles persist for so long, they do not want to move too far when baiting or afterwards while waiting for the puddles to disappear. (The puddles disappear during the 4 minute burst window, at 4:10, 4:12.5, 4:15, and 4:17.5 respectively.) Skilled puddle baiters will also leave room for BLM/RDM/SGE to come in to melee range after baiting Groupbees.
• Both groups practice recurring movement techniques that are commonly given to their role. The process of baiting the first Groupbee in a quadrant so that it does not get in the way of the later baits - look at it, look around it for the next one, move into bait position tucked in tight, move out of the baited AOE - is very similar to the process of baiting a spinner. (I considered putting spinners here, but the Groupbees create their AOEs by moving rather than firing a beam, so spinners did not fit perfectly. Since Blinding Love is a 6.5 second cast as in the original, the Groupbee baiters still want to tuck in tight so that their first bait cannot cover a neighbor's second last bee (locks in 1.5 seconds before).) The players taking the Enumeration instead practice positioning relative to each other to maintain uptime on small hitboxes, as well as stutter-step puddle baiting.

Despite only having 2.5 seconds for each Groupbee baiter to identify their first Groupbee without working backwards, this mechanic does not pressure the reaction time of anyone who can get to it. This is because the baiters can already be looking towards their quadrants during the initial Alarm Pheromones castbar, and have lots of time to act on that information before their first Groupbee locks on. (I made the initial Alarm Pheromones cast a second longer than in the original for exactly this reason.)

In fact, the rhythm of the Groupbees locking in and the puddles dropping is deliberate, to fit slidecasting and instant casts comfortably for all but the highest amounts of Spell Speed on Disciples of Magic. (Healers' movement windows from 1.5 second base cast times will work. BLM can use a Triplecast for the three movements between their four bait positions, and then has time for a hardcast before leaving their last bait position. RDM can use a Dualcast into Acceleration similarly. You get the idea.) Lining up one's GCD with the movements should be satisfying, and the lock-on times are late in a minute precisely to make that happen for both the popular 2.50s and 2.40s GCDs. (If I wanted a different timing, I would have moved the memory mechanic and/or tankbuster to make it happen.) The point in one's GCD that it lines up with is different for healers than for other roles, but that was an unavoidable result of healers' 1.5 second base cast times compared with other jobs that can precast a spell.

Even before a player gets to that level of mastery, the whole mechanic is meant to be satisfying despite being fairly easy for Savage. It involves using the duty's recurring elements to create a smooth sequence, which players can enjoy watching as it goes off. It feels like a natural upgrade to the redesigned Normal's Alarm Pheromones 2 and 3 - which also have player-baited Groupbees - but this time it is the party's choice who baits what. And it gives players back a measure of control over the battle, contrasting against the boss trying to control their characters in the Beats.

The party then gets a relative breather of a burst window and healing opportunity while the puddles disappear, before...

4:18 - 4:27 Honey B. Live: 2nd Beat

Once again, this raidwide transforms the arena and brings out the charm gauges. As in the original, four hearts are initially assigned. One heart each goes to a random light party (one tank, one healer, two DPS). (This is slightly more structure than the original, which assigns one heart each to two random DPS and two random tanks or healers.) However, another randomly chosen light party also get Command Performance (no shape). The two light party randomizations are completely independent; each party member could get one, both, or neither.

Unlike the original, the party will want to organize quickly, both because their burst should be wrapping up and because there is no stored damage cast to give a longer movement window.

4:31 - 4:35 Love Me Tender (Multiple Effects)

As this cast ends, four heart towers spawn at once, in the "mid" cardinal positions from Beat 1. As before, these resolve 10 seconds after they spawn (4:45).

Five seconds after the towers spawn (4:40), a Heart-struck pink puddle spawns under each player character's position. (Thus players must wait before moving their characters to their towers or planned spread spots, and waiting with a plan is an important recurring skill. For healers, this is also a time to heal the party up from the previous raidwide.) Heartsore spread AOEs also appear on the light party without Command Performance. The puddles, spreads, and towers all go off at once.

Here is a diagram for one solution the party might agree on to assign towers and spread positions, using the Looper icon to represent Command Performance:

https://i.imgur.com/FMkTDTI.png

(The puddle baiting in the original is not bad, but I did not want Beat 2 to feel too much like the redesigned Alarm Pheromones 1 or Alarm Pheromones 2, hence the change to a single puddle each that can be stacked since they all drop at once. This also makes Beat 2 feel more like the redesigned Normal's Beat 2 and Beat 3, despite the redesigned Normal not using puddles at all.)

Once again, the Command Performance debuffs expire two seconds after the towers (4:47). At this point, every player character has at least one heart (from their spread or tower) or is KOed by their Command Performance. The party has a total of 12 hearts if there were no mistakes so far in this Beat.

(And players still engage with Command Performance differently than Looper, because only half the party gets Command Performance here.)

4:44 - 4:48 Love Me Tender (Stack)

Creates another four-heart Heartsick stack, spawning on a random player character with the lowest number of hearts. Brings the party up to a total of 16 hearts when it hits at 4:53.

Players with one heart will want to bring their characters into the stack, while those with two hearts will want to stay out of the stack. Since the light party that initially got 0 hearts is the stack group now, the natural place for players to place the stack is in the middle of the arena, giving the other light party access to melee range if they are careful.

4:55 - 5:00 Blow Kiss

Unlike the original Normal, Blow Kiss is a cone protean with moderate damage that inflicts a heart on each hit. (For those who are not used to old raiding terms: one cone on each player character, do not hit anyone else with yours.) Blown kisses are not exactly avoidable on Earth, and this boss is blowing pheromones around trying to control the party.

Proteans have a classic resolution: everyone stand in their preassigned "clock spot" angles. And since everyone is guaranteed to get hit once and take another heart, this forces the party to distribute the earlier 16 hearts evenly or take KOs / filled charm gauges.

(While Blow Kiss is inherently limited to the Beats, and so does not allow many opportunities to use the protean, that is okay. Parties are likely to assign tower and spread positions related to their chosen clock spots throughout the Beats, helping players remember their spots in that party.)

Since this is a cone protean, not a line, the party can stand quite close to the center of the boss. This allows healers to reach the whole party with AOE heals both before and after - while Blow Kiss itself only deals moderate damage, the whole sequence and the heart DOTs are more than enough to leave non-tanks at low HP and get them KOed by Honey B. Finale.

This is not difficult for Savage as an individual mechanic - nothing in Beat 2 is - but the pacing punishes players for hesitating too much. (In 20 seconds, each player drops their puddle, moves for their spread / tower, moves in/out for the stack, and may have to move again for their clock spot depending on their planning.) It is not an Ultimate pace, or even a final floor Savage pace, but with the heavy randomness it is enough to trip up players who do not know their part well. (The Combos here and the Cycles from Eden's Gate: Descent (Savage) are faster at the same difficulty, but they specify the entire movement pattern from the castbar, so players can execute those sequences without having to think about random variations within them.)

As a whole, the redesigned Beat 2 is the fastest "song" out of all the Beats. And it rewards players who map out their movements, since all necessary information about the randomness for Beat 2 is given at the start of the Beat!

5:05 - 5:10 Honey B. Finale (ends Beat 2)

As before, a raidwide that transforms the arena back and cleanses all hearts. Has a shorter wait after the previous mechanic than Beat 1's Finale, since the party has already had the opportunity to be in healing range and the shorter wait reinforces Beat 2's "fast-paced performance" feel.

5:15 - 5:19 Killer Sting / Stinging Slash

As before, a random one of the two tankbusters.

If a tank used their invulnerability on the immediately preceding tankbuster, it is not available here, as this is less than four minutes later. However, if the first set of tankbusters put Killer Sting first, a WAR player can choose in that branch of their mitigation plan to use Holmgang both at that time and again on a Killer Sting here. (This sort of thing is an inherent result of designing the tank jobs with different invulnerabilty cooldowns.)

Of course, the tanks can choose to tank swap in advance; even if one tank used their invulnerability on the previous Killer Sting, the other tank has the opportunity to use their own invulnerability on a Killer Sting here if they have prepared by having the highest enmity when the cast starts. New raiders will generally not accomplish this immediately, as it requires remembering that the tankbuster is coming before the castbar appears, but it provides an opportunity for stronger players to express their skill by mitigating more effectively.

5:24 - 5:29 Splash of Venom / Drop of Venom

As before, stores a random untelegraphed spread or partner stack for later.

5:31 - 5:35 Alarm Pheromones (Alarm Pheromones 2)

As in the original, except that the Groupbees replay the pattern that the party baited in Alarm Pheromones 1, slowed down to half speed to match the five second pace of the venom puddle drops. (The Alarm Pheromones castbar itself is also one second longer than the original, to match the redesigned Alarm Pheromones 1.)

(As a reminder, one DPS and one tank or healer are selected completely at random for the first set of puddles, while subsequent sets select one DPS and one tank or healer completely randomly out of those who have not yet been selected. If any player characters are KOed, disconnected, or otherwise unavailable, there will be repeat selections after going through all player characters present and able to act, which will KO those selected a second time due to the lingering vulnerabilities. If any player character is KOed after being selected, the puddle bait will remain on them.)

While the original mechanic is fine as an individual mechanic, I made this change to reinforce the idea of the party taking back a degree of control over the battle, even while dealing with being chosen in a random order (within their role group) for the venom puddle drops. Most obviously, the Groupbees have been baited by the players earlier. (And they may have left themselves a mess if they were salvaging Alarm Pheromones 1!) But also, the tight space that the players must work with for the eventual stacks is now created by the players themselves (dropping the puddles) rather than the Groupbees, giving the party more agency in how much space they have and how they arrange it.

As a result, strategy designers seeking to optimize damage uptime will gravitate to having tanks and melee DPS drop their puddles at or close to max melee distance, with the melee DPS working near the "positional walls" between the rear and flanks. The other roles can then drop their puddles well outside melee range, leaving open space to take the role stacks either North and South or East and West.

These choices will become relevant to later mechanics, as the puddles once again persist for 22 seconds each.

Groupbees spawn at 5:35, 5:40, 5:45, and 5:50, along with the overhead markers for the puddle drop selections. They immediately start their Blinding Love casts, finishing at 5:41.5, 5:46.5, 5:51.5, and 5:56.5. These latter times are also when the puddles drop, almost perfectly matching the original timings.

(Here, there is no separate Poison Sting castbar. From a mechanical standpoint this is fine, as players still see the overhead markers for the random selection, and have been primed to expect puddles from the redesigned Alarm Pheromones 1. From a technical translation standpoint, bees are venomous rather than poisonous - the danger is in getting stung, not eating them - and players will get a damage name from the text accompanying the light AOE that drops their puddle anyway.)

The puddles disappear at 6:03.5, 6:09.5, 6:13.5, and 6:19.5 - well into the 6 minute burst window where many players will want to use potions. As in the original, skilled players are encouraged to plan their rotation for every possible time they could be selected to drop their puddle.

6:02 - 6:07 Bee Sting

As in the original, stack markers appear on one random DPS and one random tank or healer at the beginning of this castbar. These role stacks resolve at the end of the castbar, wrapping up Alarm Pheromones 2.

Just like the original, the timing of this heavy damage (hits about as hard as Call Me Honey when taken as two groups of four) encourages healers to think about healing well before it, well after it, or with limited oGCDs, so as to avoid losing damage output during raid buffs. However, the next mechanic will encourage pre-healing even more than the lingering potions alone in the original do.

6:14 - 6:19 Tempting Twist / Honey Beeline

As before, casts a random ground AOE with follow-up Splinter honey blobs that force players to move. This timing dictates that the uptime strategy for Alarm Pheromones 2 leave the open spaces for the role stacks at East and West, so that players can reliably dodge Honey Beeline even though the last puddles have not yet disappeared. I have placed a few extra seconds of breather here, to ensure that the puddles do not get in the way of movement for Splinter while having them matter to positioning for Honey Beeline and the 6 minute buff window. Let players feel the confined space for a moment!

Splinter and the stored damage resolve at 6:24. Yes, this will sometimes mean leaving melee range and moving large distances during raid buffs - thinking about uptime does not always mean it will be easy! Tanks and melee DPS can leave melee range for only part of a GCD and then gap-close back in, and those with frequent cast times can plan out how to get in an instant cast and/or use a movement tool. Or eat the damage loss for one worse GCD; it is not that big a deal for clearing, and I do not think duty design should be warped around always facilitating the perfect parse run with no randomness from the duty or thought about where to put buffs.

Ideally, healers will have set up their healing such that the party is back to full HP from Bee Sting without having to lose damage in raid buffs, so that no one is KOed here. However, if they have not done so, they do have 17 seconds for regens to work between Bee Sting and the stored damage.

6:28 - 6:32 Stinging Slash / Killer Sting

Whichever tankbuster was not used earlier. Again, this allows tanks to have branching personal mitigation plans if they so choose.

If the first set of tankbusters put Killer Sting first, a WAR or DRK player can choose in that branch of their mitigation plan to use Holmgang or Living Dead both at that time and again on a Killer Sting here. If the first set of tankbusters put Killer Sting second, no tank can use their invulnerability both at that time and again here. As before, the tanks can swap in advance, guaranteeing each tank a productive usage of their invulnerability.

This time also allows healers to regen the party between the last stored damage and the start of Beat 3, unlike the original.

6:36 - 6:45 Honey B. Live: 3rd Beat

Once again, this raidwide transforms the arena and brings out the charm gauges. As in the original, no hearts are initially assigned.

Also as in the original, every player character gets a debuff that will eventually explode in a large "defamation" heart AOE that is only slightly smaller than the arena, with the timer based on role groups. (For the sake of technical translation correctness, I am renaming this Venomous Pop, but it is the Poison n' Pop from the original.) At random, either DPS will get the "short" timer and tanks and healers will get the "long" timer, or vice versa, just like the original. The actual durations of "short" and "long" are different, since there is no memory mechanic cast, but the concept is the same.

There are no Command Performance debuffs this time. The party will still want every player character to take a tower - the original Beat 3 is a fine mini-phase and I am only changing its gameplay to replace the memory mechanic - but those who enjoy blind progression get to figure that out rather than being told by the debuffs. In Beat 1 and Beat 2, Command Performance placed restrictions on the party's options, while here it does not so there is no need to use it.

6:43 - 6:49 Centerstage Combo / Outerstage Combo

As before, either of the avoidable ground AOE sequences. (Second and third hits at 6:52 and 6:55 respectively.)

Near the end of this sequence at 6:53, the first set of four heart towers appears half-way between each intercardinal and one cardinal axis: either North and South or East and West at random. (These are the same positions as in the original.)

A reminder of what these positions look like, with the two possibilities in different colors:

https://i.imgur.com/Mp7recR.png

The "short" Venomous Pop timers expire and release their AOEs at 6:59, four seconds after the Combo finishes. The heart towers resolve another four seconds later at 7:03. This is the same relative timing as the original, encouraging but not requiring players with Venomous Pop to preposition towards their chosen spot at the edge of the arena during the Combo.

On a mechanical level, it is up to the party where they will create a safespot from the Venomous Pop AOEs to avoid excess hearts, and whether the towers will be taken by the role group that just had their debuffs expire or the role group that was waiting in the safespot. If they want to put in the effort of finding and using precise positions and Sprint to give tanks and melee DPS uptime while dropping the Venomous Pop AOEs, so be it.

But the healing is easier if the heart towers are taken by the role group that was waiting in the safespot, since that will distribute the necessary 8 hearts evenly across the party for even DOT ticks and AOE healing. (Doing it the other way is possible, but annoying for the healers and not advised unless it is to prevent a missed tower, since the Venomous Pop AOEs deal over half a non-tank's HP at Minimum Item Level and stacking higher DOTs on top of that leaves them not far from KO.) All this is well and good from the original.

7:03 - 7:09 Outerstage Combo / Centerstage Combo

As in the original, whichever ground AOE sequence was not used earlier. (Second and third hits at 7:12 and 7:15 respectively.) As in the original, the reverse of the earlier randomness plays out here with the AOE timers and towers as well. Even the 20 seconds between the first and second sets is just like the original.

The second set of four heart towers appears at 7:13, half-way between each intercardinal and whichever cardinal axis was not used earlier. The "long" Venomous Pop timers expire and release their AOEs at 7:19, followed by the heart towers resolving at 7:23.

At this point, the party is now up to at least 16 hearts, with every player character having ideally taken one Venomous Pop AOE and one tower.

7:28 - 7:35 Take a Bow

When this cast completes, player characters are paired completely at random with the Supporting Cast debuff, which doubles their hearts taken from any source while they are too close to their partner. To aid players in interpreting this debuff, the tethers between paired player characters are in the the "get away from each other" blue arrow style from E10S and come with Playstation markers. (Conveniently, the redesigned Normal ensures that all players who can enter this duty have seen both of these, even if they are new to endgame!) The arrows fade and get smaller whenever a tether pair is sufficiently far from each other to not receive double hearts, providing a further visual cue akin to tethers to an object or NPC fading when sufficiently stretched. The necessary distance is almost the full diameter of the arena.

(Experienced raiders may look at this and think "Remote Glitch but with hearts". Yes, that is what this is. And while TOP Party Synergy is an Ultimate mechanic, "stay away from your partner" is a second floor Savage mechanic as demonstrated by E10S Shackled Apart, and "find your partner via Playstation marker" is already in Storm's Crown (Extreme) - so this is entirely viable for second floor Savage as long as players are not asked to do anything too complex or fast with the tethers on. Do not worry, it is nothing too difficult.)

Seeing this, players will want to not only see their own Playstation marker, but also assign positions to each "side" of a tether to keep them apart for whatever subsequent attack is going to inflict hearts. And the established way to do this is for the party to line up in a "conga line", allowing everyone to see their own Playstation marker and where their partner is. The result looks a lot like the theatrical practice of ending a play by having all the actors line up on stage and bow to the audience... as the name of the cast alludes to. This furthers the evocative tension from Normal in Savage style - in order to avoid filling the charm gauge, players will want to do as she says and find out their specific assigned part!

For healers, the conga line is also an opportunity to heal the whole party up (from the ticking DOTs and the Venomous Pop damage) before everyone must spread back out for the next attack. They can also apply shields for upcoming damage and DOT ticks should they choose, though it is not needed even at minimum Item Level.

7:44 - 7:49 Blow Kiss

As before, Blow Kiss is a cone protean that inflicts moderate damage and a heart on each hit. The Playstation markers disappear at the start of this cast.

This time, the hearts will be doubled (filling the charm gauge) unless players stretch their tethers to manage Supporting Cast. Accordingly, the clock spots must be assigned to tethers and sides, not to player characters. This is not a difficult assignment. Since the tethers must always be stretched, parties will simply take their preferred ordering of the Playstation markers, set that order around one half of the edge of the arena, and repeat it around the other half of the edge of the arena.

Players must remember to go to their new position, and what that position is. But the length of the Take a Bow cast and the time between it and Blow Kiss are deliberately generous, to allow players to scan over the conga line even if they hesitate or put substantial attention on their job. (For a higher Savage floor or an Ultimate, I would set a faster pace to demand that players more tightly control their own attention. But this will very likely be a newer raider's first conga line and first exposure to Playstation markers in the high-end style of "make a plan that accounts for all possible pairings", so a gentler pace is suitable here.) Each player has nine seconds to check at most four markers, all of which are in close proximity.

(For those who are not familiar with scanning through a conga line for Playstation markers: Each player must identify their own marker and their own side of the tether. If their tether partner is to the left of them, then they are on the right side of the tether, and vice versa. Since their tether partner must be on a side of them, they need only check one side, and can choose the shorter one for speed. That shorter side has at most three player characters on it - for this not to be the case, both of their sides would need to have at least four player characters, for a total of nine player characters in the instance. So each player must check at most three markers other than their own, for a total of at most four markers, and they are all in an uninterrupted line that allows for smooth planned eye motion.)

From a mechanical standpoint, this gives every duty in the Savage tier its own twist on proteans: M1S has Quadruple Crossing, M2S has this, M3S has Quadruple / Octuple Lariat, and AAC Light-heavyweight M4 (Savage) (M4S) has Electrope Edge 2.

From an uptime standpoint, full uptime requires effort for anyone with significant constraints here. Yes, tanks and melee DPS are going to have to leave melee range briefly. The end of the cast is positioned late in a minute and thus moderately late in a slow GCD so that players can hit their GCD, run out, and then gap-close back in if they want to put in the effort to maintain uptime. (Which is not easy, especially since failing the timing fills the charm gauge on both the player character attempting it and their partner!) Meanwhile, those with frequent cast times must contend with the possibility of having to go almost all the way across the arena from their conga line spot, and plan instant casts and/or movement tools accordingly - there are only two spots in a conga line (the edges) that can guarantee which side of their tether they will be on, and the presence of any magical ranged DPS other than SMN in the party ensures at least three jobs with frequent cast times. And even those who are on the edges can still have to go to any of four spots around the edge of the arena, so they will want to get moving quickly (and set up healing accordingly) if they plan to slidecast their way there.

From a thematic standpoint, Honey B. Lovely is ending her performance by simultaneously thanking the player characters and making one final attempt to control them. (The debuff is called Supporting Cast because she wants to be the star. "Nobody steals the spotlight from me" and all.) Definitely a sting hidden under that sweet surface, especially to spectators in-setting who cannot see the tethers or markers!

7:57 - 8:02 Honey B. Finale (ends Beat 3)

As before, a raidwide that transforms the arena back and cleanses all hearts. Has a longer wait after the previous mechanic than the previous Finales, to allow the party to come back into healing range from the very edge of the arena. While this gives the party a moment to refocus after having to find their new spot, it is not entirely a kindness - the heart DOTs are still ticking.

As a whole, the redesigned Beat 3 is the slowest "song" of all the Beats, compared to the fast Beat 2 and moderate Beat 1. The slower pace enables the high randomness combined with large movement distances, creating a sweeping final performance before Rotten Heart.

8:10 - 8:14 Killer Sting / Stinging Slash

As before, a random one of the two tankbusters.

Consider what happens if this tankbuster rolls Killer Sting: If the first set of tankbusters put Killer Sting first, any tank player can choose in that branch of their mitigation plan to use their invulnerability both at that time and again here. If the first set of tankbusters put Killer Sting second, a WAR or DRK player can choose in that branch of their mitigation plan to use Holmgang or Living Dead both at that time and again here. Only the PLD+GNB tank pairing loses out on the opportunity to get a third productive invulnerability use out of a third Killer Sting.

(Again, this sort of thing is an inherent result of designing the tank jobs with different invulnerabilty cooldowns. Preventing that in the duty design would require delaying this tankbuster by almost a full minute, thus delaying Rotten Heart and distorting the battle overall.)

As in the original, this tankbuster placement creates a relatively open raid buff window where non-tanks can focus on their burst rotations. Players who dislike the slightly disrupted 6 minute burst window may choose to place their second potion here instead of there, but in this version that is something to think about rather than using a potion every time it is available for a burst window. (M4S has the early potion to avoid getting cut off by its transition, but that is about it. I would like to see that more often when it suits the duty design.)

This is the last random tankbuster, each of which has two possibilities, creating a total of eight possible timelines for tankbuster order and therefore tank mitigation planning. This number is on the high side relative to expected duty completions, since a player needs one to two small armour pieces and one to two accessory augments from this duty to complete a Best-in-Slot set for a role, but it is workable because the player does not need to have multiple mitigation plans to complete the duty without burdening their party even as a tank. ("Share every tankbuster with a fixed order of personal cooldowns" will work just fine for the tankbusters themselves, and the randomization does not change the timing of any mechanics.)

Branching timelines that change mechanic order have tighter limits, as the player - regardless of their role - is effectively forced to make a plan regarding their job for each timeline or else have their plan not function in some attempts. Here, it is optional skill expression even within the context of players choosing to attempt the duty, and so tanks who will be frustrated by not getting to use plans for every timeline can simply just not make them. Meanwhile, tanks who expect to go through many duty completions, or simply enjoy the planning, have the timelines there to enjoy.

8:25 - 8:30 Rotten Heart

As in the original, this raidwide transforms the arena and applies the Beeloved Venom debuffs. (It hits as hard as Call Me Honey, too. Hopefully the party has a mitigation plan by this point!) Honey B. Lovely also locks herself in the center of the arena facing North for the remainder of the battle, now focusing on getting rid of the party using her venom and Groupbees. (There is a mechanical reason for this too.)

I am adding to the mechanic though, since the original is rather easy - and boring - for a full 90 seconds of a duty, as I mentioned in my review of the original.

As in the original, there are four different debuff timer lengths, each of which is given to one player character for Beeloved Venom Alpha and another for Beeloved Venom Beta. These assignments are completely random, except that each player character gets exactly one Beeloved Venom - including KOed player characters. (Should these expire without being cleansed, they apply a heavy paralysis and a KO respectively. While they are present, they include a light DOT, comparable to a single heart during a Beat.)

Unlike the original, the order of the timers is marked for players with the First in Line through Fourth in Line convenience debuffs, to facilitate doing more than just mixing the Beeloved Venom types in a controlled fashion. (For those unfamiliar, the [number] in Line debuffs do not have any mechanical effect of their own, only serving to show players the order the boss has assigned them for the "core part" of a sequential execution mechanic. Using them here also provides a gentle introduction to this tool for new raiders, as here they have the time to realize what it does by looking at the debuff timers.) I did not use these in Beat 3 since it does not have a clear "core part" to read the sequence from; the towers and large AOEs are both equally important, even though the large AOEs are the only ones directly dictated by the debuffs.

Also unlike the original, all player characters receive a Contact Poison debuff lasting for the rest of the battle, which makes any contact at all between two player characters cause a Big Burst explosion without removing the debuffs. (From the original, Big Burst deals moderate magical raid damage that applies a seven second vulnerability to further magical damage, including Call Me Honey. There, it is caused by mixing Beeloved Venom types instead.) The two player characters who made contact receive five seconds of Poison Immune, allowing them to finish casts and move away without further explosions. Contact between the two types of Beeloved Venom is still necessary to cleanse them, and also to create the new Beeloved Antivenom debuff, which lasts for 30 seconds after its creation.

(Since the heal check here cannot be increased by heart stacks, Big Burst hits a little harder than it does in the original. Not a lot, the main danger is still timing the strong magic vulnerability, but about 10% more damage. While this is roughly equivalent to three heart stacks at 3% damage up each, Call Me Honey is not getting buffed at all and it is the bigger hit. The resulting overall heal check of Rotten Heart should be roughly equivalent to the original when the boss has one heart stack, but with a more even damage profile.)

(And yes, from a technical translation perspective, Contact Poison is correctly named. Poisons are passive, relying on the action of those affected, and Contact Poison does not do anything until the players act to make it explode.)

In turn, Beeloved Antivenom is needed to soak the Groupbee towers, which return from the redesigned Normal Mode. The towers have the same possible positions as in the redesigned Normal, and still spawn in pairs, but the pairs are no longer constrained to be across from each other. The only constraint on the random position selection is that each position will be used exactly once. (And this is why the boss is locked in place. I do not want tanks to find themselves moving the boss for momentarily leaving melee range and unable to move it back because of the Contact Poison debuffs around, especially since both tanks can have the same timer length and the melee DPS are also stuck respecting the debuffs!) A reminder diagram of the tower positions:

https://i.imgur.com/mX3zNlB.png

Without Beeloved Antivenom, player characters cannot neutralize the sting and so cannot affect the towers. As a visual cue to this, the towers do not light up from player characters without Beeloved Antivenom standing in them. (Yes, I am using the visual cue from TOP's towers requiring Looper, it is a good cue and does not need to be restricted to Ultimates.) If a Groupbee tower goes off without a player character with Beeloved Antivenom inside, it explodes for high raid damage (the same as Call Me Honey) and inflicts 30 seconds of the Stung debuff on all party members. Stung is a stacking, heavy DOT - a single stack hits about twice as hard as having three hearts during a Beat. Healers might be able to salvage one missed tower, but considering the ongoing damage, the party may not have the HP to give them the opportunity.

When a Groupbee tower is taken with Beeloved Antivenom, it does no damage at all, just removes Beeloved Antivenom from any player characters standing in it. (This is a second Savage floor, not an Ultimate; I am not making the healers do Program Loop spot healing on top of an AOE heal check. Spot healing the player characters with the longest Beeloved Venom timers on top of the AOE heal check is enough.) This makes logical sense for the creation and use of an antivenom.

As in the original, players must space out the Big Burst explosions around not only each other, but four Call Me Honey casts. These occur at 8:45 - 8:50, 9:01 - 9:06, 9:17 - 9:22, and 9:33 - 9:38. (This is not the same pace as the original, which has less time to the first raidwide but more time afterwards. The changed timing is deliberate, to allow players to check their [number] in Line and mentally prepare themselves.) Each [number] in Line's Beeloved Venom debuffs expire one second after the start of the matching Call Me Honey cast, so if players forget their timing they will get caught out by the magic vulnerability.

A pair of Groupbee towers spawns at the start of each cast and resolves 15 seconds later, at 9:00, 9:16, 9:32, and 9:48. As such, at most one pair of towers is on the field at any time, so players do not have to distinguish between towers about to resolve and towers freshly spawned. (Again, this is not TOP Program Loop.) These towers give players an extra five seconds compared to the heart towers, to help them deal with the random positions and move around everyone else.

Now that I have described what the mechanic does directly, let us go over what it suggests to the party as a solution, and what that solution asks of players.

Most obviously, player characters need to remain spread out somewhere for an extended period. To avoid accidental collisions when taking towers, those locations should be out of all the towers and discourage meeting near the towers to create Beeloved Antivenom. For the sake of uptime, strategy designers will want to put the tanks and melee DPS in melee range, and try to give melee DPS access to the positional walls.

At that point, it is easiest to put everyone well within melee range inside the ring of towers, so that no one tries to run through the towers in a panic or makes contact with someone taking a tower by accident. And to give players fixed locations with clear reference points to avoid touching, it is convenient to use the clock spots from earlier on the inner layer of the boss's targeting ring. (Traditional clock spots even put the melee DPS on the positional walls.) Players should already remember the clock spots they agreed upon from the earlier and easier proteans, even if their strategy for spreading Venomous Pop and assigning heart towers does not use the clock spots.

The party can then use the very center of the arena to cleanse Beeloved Venom and create Beeloved Antivenom. So far, this is much the same strategy that is popular in North American Party Finder for the original; the only difference is that the cleanse timing must be more precise to prevent Beeloved Antivenom from wearing off early, rather than precise timing simply being helpful for the healers.

But now, rather than simply returning to the boss's targeting ring in their clock spots, each pair of players must agree on who will take which tower in their pair. Since both the tower spots and the player character pairs are random, the whole party must agree on and remember a priority system for splitting up the towers. (Or have voice chat, but for strategies that are doable in Party Finder / Raid Finder.) Players are free to devise whatever priority systems they would like, including ones that aim to preserve specific jobs' uptime, but the simplest will be along the lines of "Venom Alpha looks at the ring of towers from North going clockwise and takes the first tower they see, while Venom Beta looks at the ring of towers from North going counterclockwise and takes the first tower they see".

(Some players will try to go without a priority system and say "yolo" or "use eyes", and that will go poorly for them. It was unavoidable that players would try, given a long time to see the towers, but the center of the arena is not safe to adjust through. Not only is there the ring of those waiting, but the next pair needs to create their Beeloved Antivenom while the current pair is getting to their towers.)

Optimization is thus a puzzle on both a party level - devising uptime-preserving priority systems the party can consistently execute - and a personal level - planning for every possible [number] in Line around the uptime and healing considerations of one's job. The Fourth in Line players can even pre-locate their towers by paying attention to all the previous tower spawns. Have fun!

On top of that, two pairs of player characters will often be outside the hitbox at once: one going to their towers, and the other returning to their clock spots from their towers. The party must therefore agree on and execute a system of "lanes" or other priorities to prevent collisions, which also encourages those who can do full damage at range to use that fact and create more room. (Some players will also try "yolo" / "use eyes" here, to slightly better effect, but getting stuck in a loop of reacting to each other is still deadly.)

This teaches players to stay still far more strongly than the original does. In the original, players are free to move their characters within their clock spot wedges so long as they only go to the center at their assigned time, and can even move further once their neighbors no longer have Beeloved Venom. (Players mostly do not use that freedom of movement because there is no need to.) In this version, players who are waiting their turn have their characters confined on all sides: in front is the central meeting spot, to each side is another clock spot, and behind are two lanes for dealing with towers! The message is clear without using any AOEs or puddles - stay the heck still, and especially do not touch outside of creating Beeloved Antivenom.

Staying still and not touching are themselves foundational to handling debuff passes and tether passes, which are recurring mechanical elements in higher Savage floors and Ultimates. (M4S has both in Mustard Bomb!) Before a player can pass a debuff or tether with the expected control, they must learn to stay still, both to not pass when they do not want to and to provide a clear target for the other player in the pass to work with. (As both players and designers, we work with the netcode we have got.)

Thematically, this does not use any ground AOEs at all, nor the puddles from Alarm Pheromones, but it sure feels related to the latter through the tight space. And it is related - the towers that make the space so tight are from Groupbees suicide bombing the party! Rotten Heart has become a twisted upgrade to Alarm Pheromones, where Honey B. Lovely has dropped both the pretense of caring for her Groupbees and the performance of the Beats to KO the party at any cost. (The towers can even be seen as a lingering hint of her performing skill from the Beats, while feeling different due to the longer timer and different context.) Players can enjoy the theme when it is not their turn to move.

And players do still have time when it is not their turn to move, to stand still and deal damage. But the longer period of movement, along with needing to find their tower (and for some priority systems, notice their timer partner's position), makes the time spent standing still a breather to a high-focus mechanic that would be too harsh for a second floor otherwise. (It is scary and precise, but players have ample time if they avoid panic and maintain focus. It would be a disaster if the whole party had to move at once.) This version is actually 10 seconds longer than the original, but it does not feel like it, between adding more than 10 seconds of movement and requiring more focus during one's movement.

Yes, this makes the duty as a whole harder - not just in making a mechanic harder, but with mechanics late in a duty being more time-consuming to progress due to the time taken to get there each attempt. But that does not make it wrong to do, especially when the concept of the duty suggests it. The original duty as a whole is too easy for anyone with a good reaction time, and it should be terrifying when a charmer and performer finally drops all their acts. All the rules they were keeping to for the sake of their acts no longer bind them, and all the energy they were spending on pretenses they now have free to use.

I also made sure that there are no "boss-assigned roles" that have to do very different things, keeping down the number of practice attempts a player needs to learn Rotten Heart - they only need to practice the nuances of the one thing they will do every single time they see it, with at most a different timer and different tower priority. (Putting a "boss-assigned roles" mechanic with individually difficult "roles" this late in a duty would be too harsh for a second floor, and more so the more "roles" there are.)

Yes, players will attempt to skip Rotten Heart. Players attempt to skip every mechanic they perceive as difficult late in a duty - unless the damage output requirements are so high as to prevent them from doing so - it is inherent to the incentives created by requiring mechanical consistency. "Skip Soar or disband" is still a meme from Heavensward. That is not a reason to make all final mechanics easy.

10:00 - 10:10 Sheer Heart Attack

As in the original, this is the effective enrage, bringing out and immediately filling the whole party's charm gauges upon completion. Thematic enrages are fun and should be kept.

I have added five extra seconds before this cast starts and therefore ends, to reward players who think about how to front-load one last partial raid buff window rather than autopiloting their rotation loop. (Not every duty should do that, but the mechanical timing suggested it here.) This also allows the Fourth in Line players to return to their clock spots to take full advantage of the last moments of the attempt - they have the same amount of time between their towers resolving and the enrage cast as there is between the original last raidwide and enrage cast.

10:14 - 10:19 Honey B. Finale

As in the original, this is the raidwide that actually KOs the party.

Since the party cannot attack through Devoted Drone, she takes her time to give one last flourish to the audience before the party KO - unless DOTs KO her first, pets KO her while their "owners" are unable to act entirely, or the party decides to go out on their own terms using Contact Poison. (The nine second tense wait guarantees exactly three DOT ticks after the charm gauges fill, though the party can choose to end it after five seconds.)

Should budget and voice acting constraints allow, there is room here to add a special line or sound effect for the enrage, and even a brief cut-off line for the party choosing to interrupt her. Not necessary, of course, but it would spark even more memes.

----------------

Not only does this redesign of Honey B. Lovely keep her lore and concept, I have even kept the highly random nature of the mechanics. In fact, Beat 3 and Rotten Heart are both significantly more random than their original versions, and Beat 2 forces the whole party to deal with the randomness rather than allowing the 0-heart group to ignore it with the "braindead" strategy. (And knowing where the Groupbees will be in Alarm Pheromones 2 comes from paying attention in Alarm Pheromones 1, not from "oh there is a hidden pattern that the first set always has safespots clockwise of each cardinal.") But the duty no longer pressures reaction time, whether to the boss itself or to fellow players, making it consistently doable by players who choose to put in the effort to learn it - even though the redesign is harder!

It is not an accident that I have heavily used forms of randomness that are common in Ultimate, it is by design, because those forms both force the party to deal with high randomness and ensure that they can learn to manage it every time. Randomness is fine, and can enhance a mechanic, even some fast-paced mechanics that require players to learn to efficiently process the information they have been given. Handing a player a failure that they cannot do anything about - they cannot change the boss's random choices or their own reaction time - is not just unpleasant but demotivating.

It is also not an accident that many parts of the design reward players for digging deeper into their jobs and the duty. (Tanks can always use their invulnerabilities if they make a branching mitigation plan for the tankbusters, because of the hidden pattern. There are multiple selection order mechanics for everyone with uptime constraints to map out all possibilities. And so on.) Pushing for greater mastery of a duty and one's job is a fun thing to do once players can consistently do the mechanics at all, it is a form of skill expression, and it should be rewarded rather than frustrated.

I have cut the occurences of the memory mechanic down, but not removed it, since memory mechanics are fine as a light seasoning with care not to let them get boring or crowd out other forms of difficulty. I want to demonstrate what an appropriate amount looks like in a duty timeline, with the understanding that not every duty (or even most duties) should use them. Even though in this case that means just cutting it from the Beats, that makes a big difference - it is four uses of the mechanic total rather than six, and two of those are the opening ensuring that players have seen all possibilities. (Indeed, if the Normal could be relied upon to teach the Splinter follow-up patterns, the opening pair of occurrences could potentially be cut down to a single one, forcing Drop of Venom since it is not a partner stack in Normal. (Much like how Alphascape V2.0 (Savage) (O10S) opens with the one animation tell AOE that is not demonstrated in the Normal, and expects the party to know all the patterns thereafter.) But there is not enough room in the Normal to reliably teach both that and Playstation markers, so I prioritized the concept that is interesting to casual players in the Normal.)

Unfortunately, I cannot demonstrate with a single duty design the appropriate balance between testing and rewarding tanks' skill at positioning a boss, and allowing for duty concepts and mechanics that indicate that the boss should position itself. First, that is a matter to be weighed over an entire Savage tier or even an entire major patch, not a single duty. Second, Honey B. Lovely's concept as a performer indicates that she largely should be positioning herself on her stage for her audience. (For an example where adding more positioning skill tests would fit a duty concept, albeit with reworked mechanics, M4S could have had the boss's hitbox gradually get smaller as she feeds her own body into her electrope.) Third, Tempting Twist and Honey Beeline both require precise dodging with floor references when combined with the memory mechanic, so to leave that in, I had to keep her positioning herself for those as well.

While no single duty design can demonstrate that duties need not and should not all feel the same mechanically, Honey B. Lovely's concept is great to demonstrate leaning into a duty concept as a way of creating mechanical variety. Not only does a charm-based enemy suggest different mechanics than a more head-on battle, but she has two "modes" within herself: the charming songstress and the venomous queen bee. I have deliberately created mechanical contrast between her modes - the Beats as performances on a wide-open stage, while almost everything else has movement restrictions from her venom and her Groupbees - to make the duty better express her concept. The contrast and the chosen mechanical styles create a push and pull between the boss and the party for control of the battle.

She has an excellent concept, really, it has just been overshadowed by the duty design issues. I hope this redesign example has helped demonstrate how to avoid those issues and make the game more fun for everyone. If Dawntrail had been designed like this, not only would players not have complained en masse about Honey B. Lovely, but I would be in a FRU static rather than having to sit out of its launch entirely.

You have promised to improve job design for Patch 8.0, and players have been complaining about job design for a long time. (Indeed, I myself am a systems-oriented player, hence my prior writing about job design.) Strong job design can and would prevent much of the boredom players experience in the Endwalker and Dawntrail duties. But none of that will matter if players are driven from the game by duties they cannot improve at, or have too many annoyances from duty design in the course of playing their jobs. You must get both designs right, individually and interlinked.

Give us challenges we can all rise to meet and aspire to master, not barriers we cannot do anything about.

I must humbly confess that I do not speak Japanese; for your convenience I have attempted to translate the above letter via Google Translate. Please pardon any mistakes so generated. In case of any confusion, please refer to the English text if at all possible, as I cannot verify the accuracy of the machine translation.

正直に申し上げますが、私は日本語が話せません。皆様の便宜を図るため、上記の手紙を Google 翻訳で翻訳してみました。間違いがありましたらご容赦ください。機械翻訳の正確さは確認できませんので、ご不明な点がありましたら、できるだけ英語のテキストを参照してください。

--------------------------------

吉田PとファイナルファンタジーXIV開発チームの皆様へ

まず、フィードバックをもっと早くお伝えできなかったことをお詫びします。グラフィックアップデートで健康と安全上の問題が発生したため、Dawntrail早期アクセス開始から3週間以上経ってから、安全にゲームを見ることができるポストプロセス方法を見つけるまで、Dawntrailのコンテンツをまったくプレイできませんでした。その結果、パッチ7.00のメインシナリオクエスト（MSQ）をパッチ7.05とAACライトヘビーウェイト（サベージ）のリリース前日まで完了できなかったため、まずサベージティアを完了し、パッチ7.0のコンテンツデザインをまとめて徹底的にレビューしたほうがよいと判断しました。

特にエンドウォーカーのコンテンツデザインに関するフィードバックに応えて、プレイヤーが楽しめるコンテンツを提供するために一生懸命取り組んでいることがわかります。私は、エンドウォーカーの任務設計に戻ることを主張しているわけではありませんし、ゲーム全体を簡単にすることを主張しているわけでもありません。実際、私はパッチ 7.0 で利用可能なすべての任務を完了しましたが、そのどれもが 3 回かそれ以下で私を飽きさせました!

パッチ 7.11 で登場する Futures Rewritten (Ultimate) (FRU) などのより難しい任務に挑戦して完了させたいと思っています。実際、私はエンドウォーカー中に The Omega Protocol (Ultimate) (TOP) を完了したので、パッチ 7.2 のリリース前にパッチで FRU を完了することを目指すのが論理的な次のステップです。

しかし、私はそれを実行できず、FRU をまったく完了できない可能性があります。そして、それはパッチ 7.0 の任務設計のせいでもあります。

ご存知のとおり、私は障害のあるプレイヤーです。医学や生体力学については触れませんが、特に翻訳では読者にとって難しいかもしれませんが、私の筋肉は「普通の」人間の体よりも反応が遅いです。うまくプレイできるだけの速さで入力するには、両腕を肩まで使って入力しなければなりませんが、それでも反応時間が少し遅くなります。もちろん、TOP はできるので、それほど遅くなるわけではありませんが、進行中 (ハイエンドの任務のメカニクスを練習中) は、最も速いメカニクスをもっと勉強して、効率的に処理する方法を学んでいます。「リーダーに従う」やコールアウトに頼るのを避け、代わりにすべてを自分で解決することを好むほどです。なぜなら、他の誰かがメカニクスを処理するのに反応することで余分な遅延が発生すると、他の全員がセーフスポットに到着する間に私のキャラクターが KO される可能性があるからです。私がすべての作業を費やしたにもかかわらず、単純な老化や高遅延によって動きが遅くなった人のように動いています。

Duty Finder のプレイヤーの大半は、私と一緒にプレイしても気付かないでしょう。 (私はデューティファインダーで「リーダー」として、パーティに仕組みを教えることが多かったです!) レイダーの大半は私と一緒にプレイしても、私が言わない限り気付かないでしょう。ただし、自分の体にとって入力パターンが悪いためにレイドを避けているジョブに切り替えるように求められたり、怪我を避けるために設定したレイド時間制限を超えるように求められたりすると、レイドグループに伝えることもあります。

なぜ時間制限があるのでしょうか? 他の 7 人のプレイヤーが私の入力を十分速く行うことに頼っているということは、私が自分自身に高い基準を課していることを意味します。私の体は健康な状態でそれを長く維持することはできません。覚えておいてください、私はすでにプレイするために両腕をフルに使っています。ゲームをプレイし始めたとき、徐々にスタミナを鍛える必要がありました。レイドのベースラインを超えると、頻繁にストレッチしたり、食事を増やしたり、特別な栄養バランスをとったり、身体に回復時間を与えるためにレイドのタイミングをより戦略的に考えたりする必要があります...アスリートが集中的にトレーニングするときにそうしなければならないのと同じです。(私も、ベースラインを超える時間を数週間以上維持することはできません。) 健常者と自分との違いは確かにわかります。なぜなら、プレイできるように舞台裏ですべての努力をしているのは私自身だからです。

(地球上で私に会っても、ほとんどの人は、高い身体能力や筋力を必要とする作業が発生しない限り、私に気づきません。オフィスが従業員全員に重いものを持ち上げたり、非生産的な残業をしたりすることを要求していない限り、私はさまざまな仕事をすることができます。医師に相談すれば、フィットネスのためにできるスポーツがいくつかあります。車の運転もできます。無意識のうちに補償するスキルがそれ自体で十分に優れていたため、障害があることを知らずに路上試験や緊急事態のトレーニングをすべてパスしました。(たとえば、ほとんどのドライバーは、車が死角にあることを最初に知るために肩越しに確認するのであって、周囲の交通を継続的に追跡して確認するのではないことを知ったのは、何年も経ってからでした。) 障害のせいで私よりもパフォーマンスが悪い人はたくさんいます。)

Dawntrail 以前、私のレイド時間は週 9 時間が基準でした。Endwalker の最後の方まで、週 12 時間を超えることはありませんでした。Dawntrail の早期アクセスの前に TOP を完了するために、Party Finder で進めていたにもかかわらず、週 14 時間までレイド時間を延長しました。当時は、そのためにやらなければならないことは信じられないほど集中的だと思っていました。インスタンスの外で完了ビデオ映像や打撃ダミーを使用して練習した時間は除き、毎日 2 時間という厳格なスケジュールで、休みはありませんでした。(誤解のないように言っておきますが、Party Finder ではそう簡単にはいきません。各パーティーでの時間を毎日集計する必要がありました)。各セッションの直前、セッション中の 1 時間経過時、直後だけでなく、1 日を通してさらに数回ストレッチしました。ガツガツと食べながらも、タンパク質、脂肪、微量栄養素をすべて食事に取り入れるよう注意深く計画しなければなりませんでした。手の負担を減らすために、パーティー チャット以外ではほとんどタイピングしませんでした。地球のスケジュールではこれらすべての対策が講じられていたにもかかわらず、私が体をここまで酷使しようとしたのは、Dawntrail のリリースとそれに伴う MSQ で 1 週間 (必要なら 2 週間) 自然に回復できるとわかっていたからです。

AAC ライトヘビー級 (Savage) の進行に週 9 時間費やすのは、あらゆる面で体に負担がかかりました。ストレッチや食事、睡眠だけではありません。Dawntrail は、私が自分のケアのためにあらゆることをしているにもかかわらず、ゲームをプレイして初めて手が痛くなったときでした (痛みを感じたらすぐにやめることはできませんでした。痛みは後でやってくるので、やめるのに丸 1 か月かかりました)。Dawntrail は、レイドだけでなく通常の任務に時間制限を適用することを検討しなければならなかった最初の拡張です。Dawntrail は、運動で徐々に増加するのではなく、私のスタミナを効果的に低下させ、それに合わせて時間制限を下げることを真剣に検討するようになった最初の拡張です。

パッチ 7.0 の任務設計のこれらの結果を考慮すると、任務全体を個人的に詳細に分析しなければ、FRU を完了できるかどうかはわかりません。もちろん、リリース後かなり経って、メカニクスのランダム性から生じる多くの可能性を適切に理解できる大量の完了映像が利用可能になるまでは、分析できません。同様に、たとえ FRU を物理的に実行できるとわかっていたとしても、さらに詳細な分析を行わなければ、少なくとも週 9 時間の FRU 進行を継続できるかどうかはわかりません。(私は、TOP や Dragonsong's Reprise (Ultimate) (DSR) のリリース前に喜んで取り組みました。Endwalker の開始時には週 9 時間の基準値があったからです。)

そして、週 9 時間未満では、リリース日からパッチの残りの期間を通じて効率的に時間を使用する静的レイド グループ (多くの場合、「静的」と略されます) であっても、リリース パッチで Ultimate をクリアすることは統計的に非常に困難になります。 (なぜそうなるのかを検証したスプレッドシートはありますが、レイドの一貫性の計算は重要ではありません。) Ultimate が要求する一貫性の高い基準を満たすメカニクスを適切に学ぶには、十分な時間がありません。

これを知っているので、良心に従って FRU を完了する意図で静的にコミットすることはできません。また、複数の友人が FRU の静的に興味があるか尋ねてきました。そうすることはグループに対して不誠実であり、私自身の健康にも害を及ぼします。Live Letter 83 で FRU に挑戦するときは健康に気を付けるべきだと皆さんが言う前から、私はそのことを知っていました。また、ローンチ パッチでパーティー ファインダーを使用して Ultimate を進めるには、DSR で経験したように、1 週間あたりさらに多くの時間が必要です。パーティーの明示された目的に対して準備が整っていないグループで無駄に過ごした時間を補うためだけでなく、任務を迅速に完了するために 1 週​​間に 30 時間以上を費やしている強力なプレイヤーに追いつくためです。彼らに遅れをとると、パーティーは良くなるどころか悪くなり始め、そこから悪循環になります。

そのため、非常に悲しいことですが、FRU から遠ざかることを余儀なくされることが分かりました。少なくともパッチ 7.2 の後半 (AAC Cruiserweight (Savage) の後にプレイヤーの関心が戻るまで)、そしておそらく永久に。私は難しい任務に興味があり、熱心に取り組んでいます。私は SCH の完全な最大アイテム レベル ベスト イン スロット セットを所有しています。これは、単に最大ダメージ出力のためではなく、FRU の進行のために設計されており、必要に応じてセットを調整するための本も持っています。(毎週 Savage ティアを繰り返しているので、必要に応じて他のジョブの武器を入手できます。) Eden Savage への懐かしささえ感じています。しかし、パッチ 7.0 の任務に飽き飽きしているのと同じくらい、私の体は現在の任務の設計方向の物理的な要求に耐えられません。

これまでの Dawntrail の全体的な設計方針、つまり義務設計とジョブ設計との相互作用の両方の結果として、現在の義務は精神的には非常に簡単で、肉体的には非常に困難です。 これは、運動の形として設計されていないゲームにとっては非常に悪い状況です。そして、そのようなゲームは、特定の形式と運動の強度を求めるニッチなプレイヤーに限定されることを受け入れています。幅広く永続的な魅力を求める MMO にとって、これは適切ではありません。この手紙の残りの部分では、この問題の両側面を修正することを目指して、より詳細なフィードバックを提供します。

私自身だけでなく、Strayborough 人形と Honey B. Lovely の義務について不満を述べた多くのプレイヤーのためにも。これらの義務は、そのメカニズムが非常に反応性が高く、若くて健常なプレイヤーでさえイライラさせるため、広く不満を集めました。 AAC ライトヘビー級 (サベージ) を 10 時間未満でクリアしたことに失望しているワールド プログレッション レイダー、および、ほとんど苦労しないタスクが興味深い難易度を圧迫していることに気付いた一般的な熟練レイダー向け。加齢によって反応速度が遅くならざるを得なくなったプレイヤー向け。(このゲームはかつて「格闘ゲーム プレイヤーの引退ホーム」と考えられていました。このジャンルは、最高レベルのプレイヤーが素早い反応に依存しているため、トップ プレイヤーは年齢を重ねてジャンルから完全に脱落します!) Dawntrail を始めて 1 か月か 2 か月で、まったくの退屈さから登録を解除したプレイヤー向け。戦闘デザインは、拡張のローンチ メジャー パッチで無料ログイン キャンペーンを実行するのが適切だと判断するほど登録数が急激に減少する唯一の問題ではありませんが、間違いなくリストに含まれています。 (確かに、World of Warcraftは独自の拡張パックであるThe War Withinをリリースしましたが、それはパッチ7.05のほぼ1か月後のことでした。Dawntrailはグラフィック、ストーリー、エンドゲームを最初に公開したので、全体的なパフォーマンスが良かったら、もっと多くのプレイヤーが残っていたでしょう。) 「これは難しすぎる」と言わなくなったプレイヤーは、適応したからではなく、身体的または作業記憶の限界に達した場合は適応できないため、他のプレイヤーに怒鳴られたり、イライラしてやめたりしたからです。 (あなたがそのことを言及したインタビューは、Dawntrail 早期アクセスの開始から 2 か月以上経っていたことに注意してください。タイム カードでサブスクライブしているプレイヤーでさえ、その時点ではサブスクライブを解除する機会があったため、ここに投稿することはできなくなりました。) 本当に、すべての人に。

(この段階で FRU のメカニカル デザインを変更することを期待しているわけではありません。すでに確定しているはずです。また、Ultimate のデザインは非常に精密で相互に関連しているため、パッチ 7.05 のリリース時にフィードバックを提供したとしても、FRU に間に合うように完全に対応することはおそらくできなかったでしょう。(Ultimate の任務デザイナーには大きな敬意を抱いています。彼らは私たちを限界まで追い込む任務を引き受けますが、それを超えることはありません!) しかし、パッチ 7.3 以降、あるいは AAC Cruiserweight (Savage) などのパッチ 7.2 の任務でもそれを活用できることを願っています。)

徹底のため、この手紙の残りの部分を 3 つの主要なセクションに分けました。まず、現在の任務設計が不十分な点に特に重点を置きながら、設計の原則と全体的なスタイルについて説明します。次に、パッチ 7.0 の繰り返し可能なインスタンス化された任務を私の視点から分析し、良い点と悪い点の両方を取り上げ、私がメカニクスをどのように体験しているかについて多くの例を示します。退屈だと感じる読者の方は、2 番目のセクションを飛ばして 3 番目のセクションに進むことをお勧めします。これは、Honey B. Lovely のノーマルとサベージの両方の任務の完全な再設計であり、Dawntrail コンセプトに適用したいデザイン スタイルを示しています。

ゲームの歴史全体にわたる任務の仕組み、および敵の背景にある伝承との関係について広範囲に説明しますので、他のプレイヤーの皆さんは注意してください。この手紙の残りの部分は、MSQ 以外の任務を含む、パッチ 7.05 までの完全なネタバレです。

As with previous parts of this letter, for your convenience I have attempted to provide a Japanese version using Google Translate, since I unfortunately do not speak Japanese myself. Please pardon any mistakes so generated. In case of any confusion, please refer to the English text if at all possible, as I cannot verify the accuracy of the machine translation.

この手紙のこれまでの部分と同様に、残念ながら私自身は日本語が話せないため、皆様の便宜を図るため、Google 翻訳を使用して日本語版を提供するよう努めました。間違いがあった場合はご容赦ください。機械翻訳の正確さは確認できないため、混乱が生じた場合は、可能な限り英語のテキストを参照してください。

--------------------------------

念のため言っておきますが、私はゲームを全体的に簡単にしてほしいとか、Endwalker のデューティ デザインに戻してほしいと言っているわけではありません。実際、私が望むデューティ デザイン スタイルを原則ではなく拡張版で説明しなければならないとしたら、最も近い既存の拡張版は Stormblood でしょう。

(新規プレイヤーへの注意: ストームブラッドの任務設計スタイルを、その任務がドーントレイルのジョブやエンドウォーカーのジョブのほとんどとどのようにプレイするかで判断しないでください。任務設計スタイルは、任務を形作り、プレイヤーが任務で自分のジョブをプレイすることに対する要求を作り出す必要があります。その要求が組み合わさって、プレイヤーが経験することになるからです。ストームブラッドのジョブは、MSQ ダンジョンからアルティメットまで、あらゆるタイプの任務で許容できる能力に到達するために、全体的に大幅に要求が厳しくなりました。それに加えて、拡張パワー クリープにはいくつかの形式があります。レイド ティアごとに利用できる食べ物やポーションの質が向上する、同期ダウンによる効果的な装備が向上する、パッチ 6.0 で適用されたステータス スクイッシュの寛大さ、元の上限レベルを下回る特大の効力を持つ新しい拡張ジョブ、パッチ 6.0 で実装された「2 分間メタ」によるレイド バフの同期の容易化などです。)

ストームブラッドの任務設計スタイルは、あらゆるレベルのプレイでプレイヤーに挑戦することを目指しています。これはアルティメットを導入した拡張版であるだけでなく、アルティメットのメカニクスタイプであるトリオをアライアンスレイドに導入した拡張版でもあります (オーボンヌ修道院のハイセラフ、アルティマ)。オリジナルの雷神シドをもたらした拡張版で、その後 3 回弱体化されました (そのうち 2 回は、デューティファインダーに極端なレベルのパーティー調整を要求したためです)。悪名高い MSQ ソロインスタンスでサドゥと戦う拡張版。タンクがローカルレゾナンス / リモートレゾナンスのボスバフを確認し、それに応じてオメガ M とオメガ F を移動させる拡張版... 通常モード。もちろん、プレイフィールド、グランドクロスオメガ、フォーセイクン、ハローワールドなど、印象的で象徴的なサベージメカニクスもありますが、チャレンジをハイエンドのデューティに限定するものではありません。

オメガのモットーを忘れないでください: 「戦い、勝ち、進化する。」任務デザイナー (およびジョブデザイナー) として、プレイヤーが自分自身に適用するように促す必要があります。

退屈を避け、成長を促すには、挑戦が必要です。したがって、プレイヤーに挑戦できますし、挑戦しなければなりません。つまり、失敗に対するペナルティを設ける必要があります。ペナルティこそが成功と失敗の違いを生み出すからです。回避可能なヒット、脆弱性アップ スタックとダメージ ダウン、KO、さらには通常の任務でのパーティ KO を配布します。

ただし、任務デザインの当面の目標は、プレイヤーを諦めさせることで「勝つ」ことではありません。プレイヤーに同じ任務を無期限に完了させようとするよう強制することさえ目標ではありません。なぜなら、プレイヤーは最終的に忍耐を失い、諦めてしまうからです。そして、ほとんどのプレイヤーと任務の組み合わせでは、「最終的に」はそれほど多くの試行ではありません。目標は、プレイヤーが少なくとも通常の任務を完了し、その過程で向上することであるため、失敗はやる気をなくさせるのではなく、やる気を起こさせる必要があります。

--------

すると、次のような疑問が湧いてきます。やる気を起こさせる失敗とやる気を起こさせない失敗の違いは何でしょうか?

私たちは、このことについてある程度直感的に理解していますが、より詳しい情報を提供できる既存の科学的研究もあります。今年 1 月に私が書いた、魔法の遠隔 DPS の役割のバランス調整に関する記事から引用します:

Quote:

---

心理学について少し話がそれます。特に、ブレムの動機付け強度理論について説明する必要があります。この理論は、ビデオゲームでテストされ、実証されているだけでなく、全体的に非常に広範囲に再現されているため、リヒター、ジェンドラ、ライトの要約論文では、読者をより限定された要約論文に誘導する必要がありました。機械翻訳にもかかわらず簡潔で明瞭にするため、バランスの問題に関連する理論のポイントのみを説明し、例を挙げて説明します。

プレイヤーが、ダンジョンをクリアするなど、難易度が固定された目標を検討しているとします。その目標を達成するために、パーティー メンバーになるために他の 3 人を見つけてゲームを教えるなど、ばかげたことをしなければならない場合、プレイヤーはほぼ確実にその目標を拒否し、他のことに時間を費やすでしょう。明らかに不可能なことをしなければならない場合も同様です。これは、プレイヤーが価値がある、または可能だと考える最大限の努力があることを示しています。これを、その目標の文脈では「潜在的な動機」と呼びます。

ただし、プレイヤーが潜在的な動機の範囲内でダンジョンをクリアできたとしても、最大限の努力を払うわけではありません。インスタンスで 1 時間を費やすつもりかもしれませんが、パーティーが 20 分でダンジョンをクリアした場合、同じインスタンスで残りの 40 分間、ターゲットのないボタンを押し続けることはありません。 (同じ瞬間に環境を探索するかもしれませんが、それは新しい目標です。) むしろ、目標を達成するために必要な最小限の努力を払い、それ以上は払いません。

この理論は、エネルギー節約の観点からは直感的に理解できます。砂漠に移住して新しい水田を建設することを主張する稲作農家は、おそらく米を生産できず飢えるでしょう。収穫物を理由もなく、身体的なトレーニングさえせずに何度も運ぶことを主張する稲作農家も、良い立場に立つことはできません。この理論を裏付ける多くの実験は、それが非常に正しいことを示しています。このようにエネルギーを節約することは、人間の脳の進化した命令であり、人間向けのゲームはそれに従う必要があります。

---

あからさまに明白なことを言う危険を冒して言うと、 モチベーションはすべてプレイヤーの視点に関するものです。 彼らは、任務デザイナーや他のプレイヤーの情報、関心、能力ではなく、自分自身の情報、関心、能力を使って作業しています。

(一般的に、この手紙の中では、何かを省略して誤解されるよりは、すでに知っていることを述べるほうがいいと思います。ある知識がすでに共有されていると誤解するのは非常に簡単で、言語の壁を乗り越えても状況は改善されません。)

--------

プレイヤーの視点から見ると、任務を完了するために必要な労力は決まっていますが、最初は不明です。(MSQ を一定のポイントまで完了したり、Savage ティアを完了したりするなど、長期的な目標についても同様です。) 任務に入る前に、プレイヤーは、ゲームでの以前の経験や利用可能な任務の難易度の理解などの要素に基づいて、任務を完了するために必要な労力を見積もっています。その最初の見積りで、必要な労力が潜在的なモチベーションを上回る可能性が高すぎる場合、プレイヤーは事前に「価値がない」と判断し、任務に挑戦することさえしません。 (これは多くの場合良いことです。レイドをしたくないプレイヤーが、ハイエンドの任務のために進歩パーティに参加するのを阻止するものです。しかし、MSQ に苦労していると自覚しているプレイヤーが、通常のレイドやアライアンス レイドなどのオプションの通常任務を追求するのを阻止する可能性もあります。たとえ、そこでは十分なパフォーマンスを発揮できるとしてもです。) 任務に参加すると、任務自体からのフィードバック、自分が費やした努力、経験したメカニカルな成功と失敗に基づいて、必要な労力の見積もりを継続的に改善します... そして、その改善によって潜在的なモチベーションを超える可能性が高くなると、彼らは「価値がない」と宣言し、欲求不満で任務を辞めます。

したがって、この手紙の最初の主要な設計原則は次のとおりです。

プレイヤーに何かに失敗したと伝えるたびに、プレイヤーは次回はよりうまくやり、妥当な努力で成功に導く方法を正しく認識する必要があります。

(例外が 2 つあります。1 つは、どんなに注意深く教えたとしても学ぶことを徹底的に拒否する少数のプレイヤーに対処する場合です。もう 1 つは、前提となる基礎スキルにまったく準備ができていない状態でプレイヤーが参加できるハイエンドの任務の場合です。この場合、準備ができていないプレイヤーに「これは現時点では時間の有効な使い方ではありません」と伝えて、自分自身とパーティー メンバーを惨めにしないようにする必要があります。この点で、TOP は見事に設計されています。最初の 2 つのフェーズでは、後のフェーズに必要な多くの機械的スキルを慎重にテストするため、後のフェーズに取り組む準備ができていないプレイヤーは、通常、パーティー全体がそれぞれ数十時間を任務に費やして特別なレベルに到達した後ではなく、早い段階でフラストレーションを感じて失敗します。驚きです。)

この文言をもう少し詳しく説明して、サブ原則を紹介させてください。

「認識」: 成功への道筋があっても、プレイヤーが少なくとも最初のステップを理解していなければ役に立ちません。

たとえば、カジュアル プレイヤーでも、Alexander - The Burden of the Father (A4N) の Judgement Nisi による KO を回避することは可能です。2 人のヒーラーが Nisi デバフの 30 秒間、お互いや他のすべてのプレイヤー キャラクターとの接触を避け、ヒーラーが定期的に自分自身を回復するようにキャラクターを配置するだけで済みます。(Nisi デバフは各ヒーラーに 1 つのタイプから始まり、接触によって広がり、複数のタイプがあると即座に KO になります。) しかし、カジュアル プレイヤーは、ヒーラーの役割であっても、特にそれ以外の場合は、デバフ バーを読み取ることを期待できません。これが A4N の Nisi デバフの唯一の兆候です。したがって、ジャッジメント ニシによってキャラクターが KO されたプレイヤーは、KO の直接的な原因さえ認識せず、自分のキャラクターが「謎の爆発」を起こしたとしか認識しないことがよくあります。そして、彼らは自分の混乱状態について正しいのです!

したがって、技術的には任務によって成功する方法とそれを発見するための情報の両方が与えられているにもかかわらず、制御できないと思われるペナルティに苛立ちを感じます。以前の任務で、カジュアル プレイヤーが自分のキャラクターがデバフされたときに確実に気づくように教えられていたら、彼らは与えられた情報を認識し、それを手がかりとして受け取るでしょう。同様に、ニシのデバフに、カジュアル プレイヤーが確実に気づく追加のインジケーター (The Epic of Alexander (Ultimate) (TEA) の頭上のマーカーなど) があった場合も、彼らは与えられた情報を認識し、それを手がかりとして受け取るでしょう。

パッチ 7.0 の任務では、このサブ原則 (および次の原則) の違反は、圧倒的なビジュアル、プレイヤーにインターフェースとの戦いを強いること、または時には積極的に役に立たない非ダイジェティックな手がかりによって、プレイヤーがメカニクスを理解するために正しい場所を見ることを妨げるメカニクスの形で定期的に現れます。それが起こると、プレイヤーは 「常にあらゆる場所を見るようにする」という考え方に入り、楽しい種類のストレスではありません。 (特に、人間の目のレイアウトでは細部に中心視力を使用する必要があるため、高忠実度では物理的に不可能です。さらに、プレイヤーは任務を学習せずにその状態からリラックスできるかどうかを知る方法がありません。) ハイエンドの任務では、プレイヤーにどの情報がメカニクス的に関連するかを理解するのを求めるのが適切な場合もありますが、通常の任務で実行可能にするには試行回数が多すぎます。

「次回よりうまくやるための方法」: プレイヤーは失敗から何かを学ばなければなりませんが、必ずしもメカニズムの完全な解決策である必要はありません。 プレイヤーの進化は、改善するために取る多くの小さなステップの累積的な効果です。

プレイヤーが「よし、この攻撃を受けたので、攻撃の出どころに注意する必要がある」と正しく考えれば、それは許容されます。プレイヤーが「その攻撃のタイミングを誤って判断したので、ダメージを与えるために留まるのではなく、より早く動く必要があった」と正しく考えれば、それは低頻度では許容されます。これは、タイミングを発達させるプレイヤーにとって必要なフィードバックですが、常に早く慎重に動くのではなく、そのスキルを大幅に向上させるには、多くの任務で長期間練習する必要があります。(これが、攻撃アニメーションとの視覚的な不一致にもかかわらず、多くのプレイヤーがキャストバーの最後にあるスナップショットを好む理由です。キャストバーは、タイミングの判断に使用するための公平で一貫した手がかりを提供します。) プレイヤーが「それが現れる前にそれを計画するために、任務のタイムラインを学ぶ必要がある」と正しく考えれば、それはハイエンドの任務でのみ許容されます。通常の任務では練習と暗記が多すぎるだけでなく、任務のタイムラインに関する事前の知識が必要になると、通常の任務のストーリー体験が損なわれることがよくあります。プレイヤーが「このメカニズムの参照シートが必要だ」と正しく考えた場合、それが任務全体に関する 1 つの小さな参照シートである場合のみ許容されます。2 つ目のデバイスやモニターに頼るのではなく、自分で書いたり描いたりできます。(これが、リドルアナ灯台が許容される理由です。計算障害のあるプレイヤーは、「数学のボス」の前のインスタンスのメモや他のプレイヤーから与えられたメモに基づいて、自分で 1 つの小さな参照シートを作成できます。)

プレイヤーが「どこに行くべきかがわかったらすぐに移動したが、速さが足りなかった」と正しく考えた場合、次の試みでどこへより速く移動すべきかを知るための実行可能な手順を認識し、それを実行する意思がある限りにおいてのみ許容されます。 (これはプレイヤーに要求できる速度を制限するだけでなく、そのような要求を主にハイエンドの任務に限定します。) 「ああ、このキャストには常に対応するパターンの別の攻撃が続くので、フォローアップを見るのを待つ必要はありません」は、プレイヤーに攻撃間のつながりに気付かせることができれば、実行可能です。 「このキャストが終了したら、デバフ バーを確認する必要があります」は実行可能ですが、ハイエンドの任務でのみ確実に認識されます。これは、カジュアル プレイヤー全体が現在、デバフ バーを確実に確認することをまったく学んでいないためです。 「このメカニクスのために計画した移動シーケンスを練習する必要があります」は、すでにそのようなシーケンスがあり、単一のメカニクスに外部の学習時間を費やす意思があるプレイヤーに対してのみ実行可能です。これは、Savage 未満ではまったく当てにできない特性です。Savage と Ultimate でも、プレイヤーが代わりに諦めるまでの厳格な時間制限があります。

「もっと早く反応しないといけない」というのは、どんな状況でも実行不可能なので受け入れられません。 Ultimate のレイダーの大多数でさえ、「一般的な高速処理ドリルの時間をスケジュールに組み入れ、数週間から数年にわたってそれを続け、最大で 10% から 20% の改善が得られたら、任務の進行に戻る」という選択肢は認識しません。直接そうするように指示すると、彼らは躊躇するでしょう。

プレイヤーの地球環境に対する提案された変更も、ほとんどの場合、実行可能ではありません。光の速度とさまざまな地域のインターネット インフラストラクチャによって、避けられない遅延が発生しますが、だからといって「サーバーのそばに引っ越す」ことがプレイヤーにとって実行可能になるわけではありません。(実際、プレイヤーはさまざまな地球環境のために元の地域から移動したかもしれませんが、さまざまなサーバー関連の機能 (住宅など) を失ったり、その地域のデータ センターで友達とプレイしたりする能力を失ったりせずにキャラクターを移動することはできません。) 予算が限られているプレイヤーにとっては、「VPN を入手する」ことさえ実行可能ではないかもしれません。ましてや、「インターネット サービス プロバイダーを変更する」や「有線インターネット接続が可能な住宅を探す」など、実行可能ではありません。同様に、「宣言された最小仕様を超える高性能のコンピューター/コンソールを入手する」、「PC でプレイしてアクセシビリティのサードパーティ ツールを使用する」、「多くの図表をめくる別のモニターを入手する」などは、予算上の理由で実行可能ではありません。

「それが彼らを成功に導く」: プレイヤーは、行き止まりに陥ったり、ますます小さな差で失敗し続ける無限後退に陥ったりするのではなく、最終的には成功する必要があります。そうしないと、潜在的なモチベーションが枯渇し、彼らは諦めてしまいます。 これは非常に単純で、プレイヤーが「ただより速く反応する」という見通しにやる気をなくすもう 1 つの理由です。彼らは、反応時間のトレーニングは最終的には行き止まりになることを直感的に理解していることが多く、十分に改善できるかどうかわからないまま、それに多額の投資をすることを嫌がります。

(パズルの仕組みはここで興味深いケースです。パズルを解こうとするプレイヤーの視点から見ると、潜在的な解決策のいくつかの合理的な方向性は行き止まりになる可能性があり、それがパズルの難易度に必要となる場合もあります。しかし、適切に設計されたパズルの仕組みは、間違った潜在的な解決策すべてについて、たとえ仕組みのランダムなバリエーションだけであっても、プレイヤーに明確なフィードバックを与え、計画が不完全に実行されたのではなく間違っていることを伝えます。このフィードバックは、プレイヤーが行き止まりに陥っても行き詰まるのを防ぎます。なぜなら、最終的にその方向の潜在的な解決策を除外するように指示するからです。「最終的に」については、次のサブ原則を参照してください。)

「彼らが合理的だと思う努力の範囲内で」: プレイヤーが任務に費やす累積的な努力、つまりあなたが彼らに与えるすべての機械的失敗と彼らが努力して得たすべての機械的成功の両方は、彼らがそれに費やそうとする努力を超えてはなりません。 プレイヤーは、少なくともあなたのゲームに関しては、そして彼らの人生全体に関しては、オメガのような容赦ない進化への意欲を持っていません。彼らはいつでも立ち去ることができます。あなたは彼らが喜んで与えてくれる時間と労力で働かなければなりません。

言うまでもなく、プレイヤーの「妥当な努力」の定義は状況によって大きく異なります。メカニズムの事前知識がなければ、インスタンス タイマーが切れる前に通常の任務を完了できないと、ほぼ誰でもかなりイライラするでしょう。一方、究極に挑むレイダーは、1 つをそれほど早く完了できたとしたら当然怒るでしょう。

「正しく認識する」: プレイヤーが前進する方法が存在すると誤って信じている場合、前進する方法であると信じていることを試し、失敗し続け、諦めます。前進する方法が存在すると正しく信じているが、それが何であるかが間違っている場合、プレイヤーは機能するルートではなく、認識しているルートを試し、費やした累積的な労力を追加します。したがって、 意図した解決策がプレイヤーの既存のメカニズムの知識から遠いほど、より広範囲に答えを導く必要があります。任務を簡単にしたいほど、プレイヤーをメカニズムの解決に導くためのガイドを広範囲に行う必要があります。 これが、通常の任務にパズルのメカニズムを持たせることができない理由です。Duty Finder のマッチメイキングでは、パーティー全体で忍耐強く待つことはおろか、パズルを解けるプレイヤーがパーティー内に 1 人でもいることさえ保証できません。そして、最も手の込んだパズルのメカニズムはすべて Ultimate にあります。

しかし、その影響はパズルのメカニズムに限定されません。Duty Finder のプレイヤーは実際にはかなり幅広いメカニズムを学習できますが、通常の任務を通じての指導は小さな段階的なステップで行う必要があります。そこでは、プレイヤーはクエストやルーレットを完了しようとしており、複雑な概念を一度に学習するために任務に挑戦する以外に何かを学習する興味はありません。(Shadowbringers と Endwalker が Stormblood よりも優れていた重要な点は、ガイドと指導です。マーカー、NPC の解説、時には NPC のデモンストレーションを使用して、小さな部分でプレイヤーに教えます。)

「プレイヤーに何かに失敗したと伝える」: プレイヤーが何らかの意味で失敗してもペナルティを感じていない場合、改善するためのステップを認識しているかどうかは問題ではありません。いずれにしても、改善する理由はありません。

これは多くの場合、職務のメカニクス自体が合格/不合格であるため、ジョブ設計の問題ですが、ジョブがメカニクスの成功度を複数段階に分ける職務設計にも関連しています。カジュアル プレイヤーが通常の職務のみに従事し、エンレイジ キャストを実際に見たことがない場合は、装備をアップグレードせずにダメージ出力を高めるために実行できるステップを理解していなくても問題ありません。これには、キャスト時間が頻繁なジョブのスライドキャストや、近接範囲から離れる必要があるメカニクスのスナップショットの直後にボスにギャップ クローズバックするなどのメカニクス解決テクニックが含まれます。このようなことを達成できないことは、最適なパフォーマンスに比べて「失敗」であっても、プレイヤーを苛立たせません。より困難な任務を引き受けるためであれ、単に内なる楽しみのためであれ、彼らが仕事でより高いスキルを追求することを選択した場合、彼らはダウンタイムや低い効力出力をペナルティと認識するようになる可能性があり、その時点で原則の残りの部分が再び適用されます。

これは、それに基づいて行動できるプレイヤーにフィードバックを与えることに関して、通常の任務設計にも関連しています。

たとえば、Alexander - The Fist of the Son (A5N) には「タンクスワップ」メカニズムが含まれています。これは、最も敵対的なプレイヤーキャラクターをターゲットにした一連の攻撃で、同じプレイヤーキャラクターがそれらをすべて受けると特に危険であり、一連の攻撃を分割するために、フルパーティーの 2 人のタンクが敵対的な人を交換することを奨励または要求します。A5N では、関連する一連の攻撃は非常に遅く、フィードバックの層があります。ボスの Ratfinx Twinkledinks は、4 つの個別の Gobjab 攻撃を使用する前に劇的に成長し、それぞれが Headache のスタックを与えます。頭痛のデバフの説明には、「軽度の頭部外傷を負っています。与えるダメージが減ります。外傷が増加すると脳震盪になります。」と書かれています。これでプレイヤーが気づかない場合、脳震盪のデバフにより、プレイヤーキャラクターは特別なアニメーションで行動できなくなります。これはプレイヤーにとって非常に明白です。その後、ボスはゴブストレート タンクバスターを使用します。これは、脳震盪のあるプレイヤーキャラクターが受けると大きな真のダメージを与え、KO をもたらします。最初のゴブジャブからゴブストレートの開始までの間の任意の時点でタンクを交換すると KO を回避でき、最初のゴブジャブから 4 番目のゴブジャブまでの間の任意の時点でタンクを交換すると、脳震盪も回避できます。

最大の問題は、このフィードバックのすべてが、現在最も敵対的なキャラクターを持つプレイヤー、つまり「メイン タンク」に提供されることです。最も敵対的なキャラクターになるために行動する必要があり、したがってフィードバックを受け取る必要があるプレイヤーは、もう一方のタンク、つまり「オフ タンク」です。 (現在では、メインタンクはオフタンクをシルクできますが、オフタンクがメインタンクの敵意の半分以上をすでに持っていない限り、オフタンクを最高の敵意に交換することはできません。また、シルクはパッチ 4.0 ストームブラッドで追加されたため、A5N が最新だったときはタンクにそのオプションはありませんでした。簡単に敵意を高くするにはダメージ出力を犠牲にしなければならないジョブ設計の環境では、オフタンクが絶対に行動する必要がありました。) そして、オフタンクは、自分の副タンクのデバフ バーを見るだけでなく、デバフの効果を読まない限り、これをまったく認識しません。これは、大多数のカジュアル プレイヤーが行いません。彼らは、自分の副タンクが「不可解に爆発した」ことだけを認識し、おそらく軽減または回復の失敗に起因すると考えるでしょう。

この任務設計を考えると、カジュアル プレイヤーのオフタンクが、副タンクのデバフを読むという非現実的な期待を抱かずに、次回よりうまくやれるようになるには、何が起こる必要がありますか?まず、メインタンク（またはパーティー全体のデバフを監視している他のパーティーメンバー）は、オフタンクが攻撃を分割する行動をとることが可能であることを理解する必要があります。次に、これを理解した人は、オフタンクにメカニクスに失敗したことを伝えなければなりません。これには、社交的な意欲とある程度のタイピング速度の両方が必要です。3番目に、オフタンクは他のパーティーメンバーを責めるのではなく、これを受け入れる必要があります。4番目に、パーティーメンバーの誰かが、タンクの交換という比較的複雑な概念を分解して、オフタンクがいつ最も敵対的なスポットを獲得するか、それをどのように行うか、そして以前のメインタンクは将来のタンクの交換でいつか最も敵対的なスポットを取り戻さなければならないことを理解できるようにする必要があります。

カジュアル プレイヤーはこれらすべてのステップを大きなハードルと見なしているため、A5N は、Duty Finder の人口にタンク スワップを教えることはできませんでした。(ハイエンドの任務を追求するプレイヤーはそれを達成します。ハイエンドの任務を取り巻く期待により、より熱心で、学習にもっと努力を払うことを期待し、他の人からのフィードバックを受け入れられないか受け入れないプレイヤーをキックする意思のあるプレイヤー人口が生まれるからです。) もっと広い意味では、カジュアル プレイヤーが他の人にフィードバックを与えたり受け取ったりすることは当てにできないため、任務デザイナーであるあなたは、それに基づいて行動できるプレイヤーにフィードバックを与える必要があります。

(実際、Heavensward は Stormblood よりも私が望むデザイン スタイルから遠いですが、それは挑戦に対する姿勢のためではなく、プレイヤーに教えるのが下手だったためです。Alexander - The Burden of the Son (Savage) (A8S) は、私が Ultimate を追求する決心をさせた任務です。なぜなら、その難しさと集中力の強さを心から楽しんだからです。)

カジュアル プレイヤーにタンク スワップを教えたいですか? 教えられますが、そのようには教えられません。

まず、タンクの交換が必要な攻撃シーケンスのマーカーを作成します。これにより、プレイヤーは、敵が危険な攻撃を終える前に、将来の任務でそのコンセプトを適用する手がかりを得ることができます。(これは、パッチ 5.0 Shadowbringers で導入された「中断可能なキャスト」マーカーのようにキャストバーに表示することも、Shadowbringers と Endwalker で導入されたさまざまなタンクバスター タイプのマーカーのようにボスの現在のターゲットに表示することもできます。どちらも、「ターゲットのターゲット」HUD 要素上またはマーカーをボスの新しいターゲットに移動することで、交換が成功した場合に追加のフィードバックを提供する機会を提供します。) 次に、指導とデモンストレーションを行うことができる NPC の共同タンクと一緒に、ソロ インスタンス化された任務を作成します。たとえば、現在タンクをプレイしているかどうかに関係なく、Dawntrail のエンドゲームに参加するすべてのプレイヤーにそれを学習させたい場合、ソロ インスタンスは、Wuk Lamat と Gulool Ja Ja との戦闘訓練の Echo ビジョンとして MSQ に表示することができます。 (キャラクターとしての Wuk Lamat が嫌いなプレイヤーでも、新しいメカニカルな知識を得ることができ、既存の MSQ のテーマにも合うでしょう。) その任務では、プレイヤー キャラクターを最初にメイン タンクにし、ボスが非常にゆっくりと派手なタンク交換メカニクスを開始して、プレイヤーに初めて新しいタンク交換マーカーを表示します。(A5N は「ゆっくりと派手」な部分を正しく行いました。) 一連の攻撃中に、NPC の共同タンクに危険についてコメントさせ、プレイヤーに正しいタンク交換を実演させ、一連の攻撃の最後にボスを動かさない位置に移動します。次に、おなじみのメカニクスの短い「スペーサー」期間の後、ボスに NPC をターゲットにして同じタンク交換メカニクスを再び開始させ、プレイヤーに 2 度目のタンク交換マーカーを表示します。プレイヤーがタンク交換の実演を見たので、NPC はプレイヤー キャラクターがボスの注意を戻す必要があるとコメントでき、任務では一部のソロ インスタンスのように、非ダイジェティックなプロンプトを表示することもできます。プレイヤーがメカニクス中に NPC から最も高い敵意のスポットを正しく奪い、一連の攻撃の最後までそれを維持した場合 (Provoke とタンク スタンスを使用)、NPC から感謝のフィードバックを与えて続行します。そうでない場合は、NPC を KO し、任​​務 (または任務の一部) を最初からやり直します。

(その基礎の上に構築された概念、たとえば Shirk やダブル タンク スワップを使用するタイミングなどは、直後に説明できますが、プレイヤーが基本的なフォームを固めた後で説明するのが最適です。)

そうすると、将来の任務で共同タンクにマーカーが再び表示されたときに、タンク スワップを試みます。(両方の側面を見たことがあるため、別のタンクが自分から交代しても混乱しません。) 最初の数回はすぐには対応しない可能性がありますが、最初の 2 つの Dawntrail トライアルでは、恥ずかしい思いをすることなく、プレイヤーが任務サポートを練習できる良い機会だったでしょう。一部の概念については、カジュアル プレイヤーに取り上げてもらうために、プレイヤーの手を実際に導く必要があります。

「必須」: なぜ「すべき」ではないのでしょうか? 「すべき」では、ゲームを設計してプレイヤーの体験を向上させる方法を真剣に検討することなく、ゲーム設計ではなくプレイヤーに責任を押し付けようとする誘惑が生まれるからです。

カジュアル プレイヤーが、通常のレイドを本格的なレイドだと誤解して避ける場合、それを修正するのはコミュニティの責任ではありません。そうではないと伝えるのはあなたの責任です。(デューティ サポートを通じて MSQ を完了したプレイヤーは、他のプレイヤーとほとんど話さないかもしれません!) 色覚異常のプレイヤーが Alphascape V3.0 (Savage) (O11S) または TOP の 2 種類の拳の違いを判別するのに苦労する場合、それは彼らのせいではありません。それは、色の手がかりに頼らずに形を十分に区別できなかったためです。 (TOP のその他の要素は、色覚異常のプレイヤーにとって完全に区別がつきます... Asphodelos: The Fourth Circle (Savage) (P4S) Act 2 のテザーとは対照的に、プレイヤー キャラクターの頭上に表示される色以外にはまったく区別できる特徴がありません。同様に、Emerald Weapon (Extreme) のオーブは、色によってのみ区別されます。) プレイヤーが Seat of Sacrifice Normal で Active Time Maneuver (クイック タイム イベントまたは QTE とも呼ばれます) を完了できない場合、または完了しようとして怪我をした場合、それは学習の失敗ではありません。ゲームが実際に必要とする、またはプレイヤーに必要と思わせる入力のペースを考慮した設計ミスです。

はい、その特定の QTE には、バーの最初の 75% が急速に消耗してプレイヤーが常に失敗の瀬戸際にあると思わせる「フェイクアウト」があることは承知しています。プレイヤーが失敗しそうに見えても、1 分あたり 200 アクション (APM) で完了できるというメトロノームのデモを見たことがあります。このゲームの QTE では特にクリックする必要がないことはわかっています。メニューを開いたり、チャット ウィンドウに入力を送信したり、オペレーティング システムがゲームからフォーカスを外したりしないボタンであれば、どれでもかまいません。それでも、プレイヤーに求めるのは無理な要求です。実際のゲームプレイは高速ジョブでもそれほど速くないため、設定したい興味深い課題に対応できるプレイヤーを不必要に排除してしまうからです。

非常に高速なジョブを取り上げ、計算してみましょう。プレイヤーが完璧な MCH ハイパーチャージ ウィンドウを実行でき、ヒート ブラスト / ブレイジング ショットを 1.5 秒ごとにヒットし、ヒート ブラストの合間にガウス ラウンド / リコシェ / ダブル チェック / チェックメイトを均等にヒットするとします。これは 1.5 秒ごとに 2 回の入力、つまり 0.75 秒ごとに 1 回の入力です。プレイヤーがハイパーチャージの短い期間だけでなく、継続的にこのペースの入力を維持できると仮定します。0.75 秒ごとに 1 回の入力は、60 秒ごとに 80 回の入力なので、80 APM になります。プレイヤーがアニメーション ロックのあるアクション (主にジョブとロール アクション) に片手だけを使用し、もう一方の手は移動とターゲットに完全に自由に使用でき、両手を完全にずらしたままもう一方の手で同じ APM を実行できる場合でも、それでも 160 APM にしかなりません。プレイヤーは、ゲームで最も速いジョブの 1 つを快適にプレイできるにもかかわらず、その QTE を物理的に成功させることができない場合があります (または、繰り返し試みることで怪我をするリスクが高くなります)。

また、1 人のプレイヤーがそれに失敗した場合のペナルティは、MSQ 任務で そのプレイヤーをゲームの残りの部分から締め出す即時パーティー KO であるため、

ちなみに、この計算は私の APM とほぼ一致しています。このゲームでの QTE に対処するには、その間にどのキーボード キーを押すかを制御することをあきらめ、代わりに、各指を 1 つの軸に沿ってずらしながら「転がし」、各手を別の軸に沿って「転がし」、各指の動きごとに複数のキーを押すという複雑な方法を考案します。(この方法はほとんどのプレイヤーには思いつかず、プレイヤーの運動協調性に非常に要求されるため、APM が限られているプレイヤーは、思いついても実行できない可能性が高いことに注意してください。) それでも、Seat of Sacrifice の通常 QTE の持続入力要件をほとんどクリアできませんでした。Ultimate をクリアできる人が MSQ を完了できないほどブロックされていたのは不合理です。 これは明らかに設計上のエラーです。

対照的に、アレキサンダー - 創造主の心 (A11N) の QTE は、個人の KO を回避するために数秒に 1 回の入力が必要であり、その機能も初心者のゲームプレイに必要であるため、まったく問題ありません。同様に、アルファスケープ V3.0 (O11N) のオメガ ジャマーの使用は、失敗すると通常はパーティーが KO になるものの、まったく問題ありません。オメガ ジャマーには 1.5 秒のクールダウンがあり、パーティーはシールドを形成するために必要な共同使用回数にかなりの余裕があるためです。これは、非常に遅いジョブをカジュアルにプレイするよりも入力速度が速くなることはなく、入力のリズムも良くなる必要はなく、最も苦労しているプレイヤーは、パーティー メンバーがより多くの入力を提供することで、任務を乗り切ることができます。これらは、要求がはるかに少なく、コア ゲームプレイに本質的に必要なタスクに匹敵するため、すでに必要な十分な理由があります。

このゲームでは QTE はメカニカルに興味深い用途には使用されていないため、アクセシビリティ設定で QTE を削除したり、A11N のようにボタン 1 回押しにしたりすることができます。あるいは、オメガ ジャマーのような寛容なリズム チェックで全員を置き換えることもできます。この方法でも、ジョブ アクションとは無関係にプレイヤー キャラクターが戦っていることをうまく表現できます。技術はあります。しかし、毎回そのようなことを考えなければ、設計エラーが漏れてしまいます。

すべての拡張には、Dawntrail でさえ、さまざまな優れたメカニクスがあります。しかし、ほとんどが優れたメカニクスでは十分ではありません。1 つの悪いメカニクスだけで任務を台無しにできます。1 つの悪い任務だけでルーレットを台無しにできます。プレイヤーが実行できないメカニクスが 1 つあるだけで、続行できなくなります。

毎回、苦痛や過度の怪我のリスクなしに任務をプレイできるという最低限の基準を満たす必要があります。これは単に、タスクの厳しさです。

--------

最初の主要な原則は、プレイヤーに必要な労力の見積もりがどのように変化するかを考えることです。プレイヤーの潜在的な動機がどのように変化するかについてはどうでしょうか?

では、そもそも何がプレイヤーを任務に駆り立てるのでしょうか? 任務に費やすと予想される労力ではなく、任務から得られると予想されるすべてのもの、つまりストーリーの進行、自分の仕事をこなすというパワーファンタジー、後の任務へのアクセスとパワー、他のゲームシステムやエリアへのアクセス、コスメティック、軽い娯楽、個人的な達成感、社交、コンプリート主義などです。ギアは、コスメティック、後の任務へのアクセスとパワー、および/または達成の形式のいずれかです。ギルも同様ですが、任務以外のゲームプレイにも使用できます。ゲーム内アイテムの意味は、私たちが彼らのために、または彼らと一緒に行う旅から生まれます。それがなければ、彼らは単なるピクセルです。

そして、プレイヤーの観点からは、任務を完了することで得られる総報酬は固定されていますが、任務を完了するために必要な労力と同様に、最初は不明です。 (エキスパートルーレットを何回やれば欲しいトームストーン装備をすべて入手できるかはわかっていても、最初はその任務や装備からどれだけの楽しみが得られるかはわかりません。) そのため、彼らは任務を完了することで得られるものについての見積もりを継続的に精査し、潜在的なモチベーションを変えていきます。潜在的なモチベーションが任務を完了するために必要な労力の見積もりを下回ると、彼らは「価値がない」と宣言して任務を放棄します。原因によっては、フラストレーション、退屈、無関心などが考えられます。

これが、この手紙の 2 番目の主要な設計原則です。

ゲームを難しくしようとするときは、ゲーム全体を通して楽しくすることを忘れないようにしてください。退屈を避けるためにプレイヤーに挑戦する必要がありますが、プレイヤーに挑戦できるのは、その挑戦に取り組むことで得られるものに応じてのみです。

いくつかの帰結を検討してみましょう。

プレイヤーは、現在ゲームをどれだけ楽しんでいるかに基づいて、後でゲームをどれだけ楽しむかという見積もりを常に精緻化しています。

たとえ成功が任務の完了という形を取らなくても、その改善に関連する何かに実際に成功したという肯定的なフィードバックを定期的に与えなければ、プレイヤーが戦闘での改善を長期間続けることは期待できません。したがって、特にカジュアル プレイヤーにとって、「戦う、勝つ、進化する」は実際には正しい順序です。改善は失敗から学ぶことで得られますが、成功も必要です。 Ultimate レイダーでさえ、取り組んでいるメカニズムを正しく実行せずに何時間も進歩に費やすことにはイライラします。

同様に、プレイヤーの見積もりには、当然、任務の設計外の要素、および戦闘外の要素が含まれます。 プレイヤーがストーリーに興味を持っている場合、Wuk Lamat の書き方が下手だと感じると、MSQ で彼らに挑戦する機会が少なくなります。MSQ では、苦労の甲斐あって Wuk Lamat がもっと見られると当然期待します。プレイヤーが ジョブから何か特定のものを得たい場合、その希望に合うジョブにプレイヤーを誘導しないと、義務から得られるものが少なくなり、義務に努力を注ぐことへの寛容さが低下します。プレイヤーが光の戦士を道徳的に立派なヒーローとしてプレイしたい場合（ほとんどのプレイヤーがそうします）、エンドレスを人間と見なしている場合、エンドレスによる大量虐殺を強制すると、たとえゲームを続行したとしても、将来のすべてのストーリーラインと関連する義務から永久に遠ざかる可能性があります。Dawntrail は、すでに「戦闘が得意」ではないプレイヤーに粘り強くプレイする理由を与えるという点で、かなり失敗しています。

(個人的には、古いクラフト システムのためにゲームをプレイし始め、MSQ の任務での最初の苦労を乗り越えて、クラフトと収集のための新しいゾーンに到達しました。実際、将来の MSQ 要件に備えるために最初はハイエンドの任務に挑戦し始めましたが、その後、通常の任務のこれまでのどの設計スタイルよりも、レイドの計画重視の設計スタイルとジョブの習熟を促す設計スタイルの方が好みであることに気付きました。そして、そのおかげで、ゲーム外での身体的な制限への対処方法が改善され始め、レイドへの非常に強い動機付けになりました。プレイヤーの粘り強さに関しては、私は非常に例外的です...それでも、今日プレイし始めた私のような人は、エンド ゲームにはまったく到達できません。新しいクラフトと収集のアクションは、ローテーションの深さを回復するのには良いですが、パッチ 7.0 の MSQ 任務の落胆するような側面を私のような人が乗り越えるにはまったく十分ではありません。)

もちろん、プレイヤーが将来の任務を楽しめるかどうかは、任務を完了するかどうかにかかわらず、現在行っている任務によって決まります。(そして、プレイヤーはゲームに健全な関心を持っている場合、ほぼ常に、同様の難易度の将来の任務を完了することに何らかの関心を持っています。プレイヤーがその関心を欠く唯一の無関係な理由は、魅力のために単一の究極を追求することです。) プレイヤーをあまりに厳しく追い詰めると、たとえ完了したとしても、任務を楽しむことはできなくなります。また、あるカテゴリの任務を楽しめないままでいると、そのカテゴリをあきらめる十分な理由があります。 これは、パーティーが完了まで「運んで」くれる場合や、何度もやりくりしてパニックになり、将来の任務を心配して楽しみが押しのけられる場合など、ソロの場合に発生することがあります。ほとんどのプレイヤーは、カジュアルなストーリー重視のプレイであっても、パーティーに役立っていると感じたいと考えています。

(The Copied Factory が Patch 5.10 で初めてリリースされたとき、私は特定のメカニクスに苦戦し、反応が速すぎると感じました。一度経験するのは問題ありません。しかし、レイド用に特定のギア ピースを入手するために最初の週に何度も任務を完了しても、メカニクスの反応に問題がある場合は別です。その後も状況が改善されなかったため、Savage の最終フロアと同じように、そのようなメカニクスごとに詳細な計画を立てました。キャラクターだけでなく、手、目、カメラも事前に配置しました... その結果、キャラクターはセーフスポットから出るのではなく、セーフスポット内で KO されるだけでした。その後、TEA をすでに完了していた 1 人を含む数人のレイダーの友人に計画の改良を手伝ってもらいましたが、誰も改善策を提案できませんでした。)

(その時点で私は挑戦に立ち向かうために全力を尽くしましたが、反応時間の問題という最初の印象が正しかったことがわかりました。そのため、The Copied Factory に入るのをやめただけでなく、The Puppets' Bunker (Patch 5.10 でリリース) がリリースされたときにも、The Copied Factory に入るのをやめました。 5.30) 私は、アライアンス ルーレットで個人的に楽しめる任務を獲得できる可能性をさらに下げないように、アンロックを拒否し、装備の理由で抗議する人には、私の Eden's Verse Savage 装備を紹介しました。私が楽しむだろうと思って抗議する人には、「Ultimate をクリアしたらまた聞いてください」と返答しました。これはわざと馬鹿げた発言ではなく、論理的な結論です。より良い計画と HUD レイアウトによって実際の応答時間を改善するあらゆる迅速な方法をすでに使い果たしてしまったので、残された唯一の選択肢は、メカニズムを効率的に処理する練習を何百時間から何千時間も行って改善を期待することだけでした...そして、すでに Savage に慣れていたので、改善を示すことができる唯一の高難易度は Ultimate でした。)

(たとえ私のキャラクターがずっと KO されていたとしても、他の 23 人のプレイヤーが Alliance Raid を乗り切ってくれることは十分わかっていました。私はすでに DPS をプレイしていました。とにかく私は惨めで、当然ながら、私を惨めにさせている同じ任務で、反応時間の延長ドリルに挑戦する気はありませんでした。言うまでもなく、通常の任務で個人的な改善が見られないことに苛立たしいと感じているプレイヤーの大多数は、Ultimate でそれを修正しようとさえせず、ゲームをやめてしまいます。)

また、ゲームの繰り返しベースのゲームプレイ ループにも注目する価値があります。プレイヤーが任務の詳細をすべて知っていることで退屈になったり、イライラしたりする場合、その任務を繰り返したくはないでしょう。 特に、パッチ 7.0 の任務に基づくと、次のようになります。

• 敵が行っているメカニクスのランダムなバリエーションの最初の表示を遅らせても、任務が面白くなったり、楽しくなったりすることはありません。 プレイヤーは、最初のキューが表示されたときにそれを学習し、そのメカニクスを解決することで生じる短いストレスにもかかわらず、退屈してそれまで待つことしか学べません。その結果、短いストレスの瞬間が中断される退屈な体験になり、楽しい瞬間はまったくありません。ストレスの多い難しい任務でさえ退屈になる可能性があります。これが、以前の拡張版の任務よりも Dawntrail の任務で眠りに落ちる可能性が高くなった理由です。
• シーケンス内で反応を強要する動きのシーケンスは楽しくなく、面白​​くもありません。 プレイヤーは計画を立てる機会を奪われ、代わりに繰り返し反応する以外に手段がないと学びます。その結果、反応のストレスの最中でも退屈してしまいます。(プレイヤーが反応することで何らかの楽しいアドレナリンラッシュを得ることを期待しているのであれば、私には理解できません。また、プレイヤーのさまざまなフィードバックから、そう思っているのは私だけではないことがわかります。私のような人にとってはせいぜい迷惑なだけですし、プレイヤー人口は古いターンベースの ファイナルファンタジー ゲームによって形成されており、このゲーム自体は歴史的にリアルタイムゲームとしては反応時間に関してかなり寛容です。反応のスリルを味わえる人には、他にもたくさんの選択肢があります。) 代わりに、プレイヤーに動きのシーケンスを計画するために必要な情報を与え、それから、そのメカニズムの役割と意図された効果に適した速度でそれを実行するように求めます。
• 必要な視覚的指標を見にくくすることは、楽しさを損ないます。 (同じコンソール/コンピューター設定を使用している健常者と反応の速いプレイヤーの間でも、視覚処理のスタイルには大きな違いがあることに注意してください。さらに、メカニックをはっきりと表示するスクリーンショットは、限られた時間でプレイヤーが見ているものを正確に表しているわけではありません。人間の目は中心視野の外側では詳細を捉えることができず、さまざまなアクション効果によって画面が一時的にぼやけるためです。) 視覚障害、光過敏症、迅速な処理への依存、または単に予想外の視覚処理スタイルを持つプレイヤーにとっては、これは積極的な不利益です。そして、他のすべての人にとって、どちらでも問題ではありません。任務の伝承に適したビジュアルを用意するのはもちろんですが、視覚の明瞭さを優先する方法で行ってください。
  (「必要」という言葉は重要です。爆発で見えにくくなる A8S メガビームは、実際には計画テストなので、とても楽しいです。このビームが最初に登場すると、ランダムなプレイヤー キャラクターに餌付けされていることがパーティにわかるので、ボスが餌に誰を選んでも、前のメカニズムに対する解決策をアレンジするだけで安全地帯を作ることができます。さらに、すべての任務には、メカニズムとは関係のない視覚的要素が他の理由で含まれており、それらはアーティストや伝承ライターが望むほど微妙な場合があります。)
  実際、視覚的な明瞭性を確保することで、より楽しく、設計しやすいメカニズムが生まれます。なぜでしょうか。見えにくい視覚的要素は、一部のプレイヤーを他のプレイヤーよりも大幅に遅らせますが、明瞭な視覚的要素はその遅延を排除するため、任務に挑戦するプレイヤーの間で比較的均一な識別時間が得られます。 (隠されたビジュアルの多くの形式は、メカニクスのランダムな変化に基づいて、プレイヤーをどれだけ遅らせるかにも変化します。) その後、その判断時間を考慮して、決定を下したりキャラクターを動かしたりするなど、メカニクスの他のタスクに時間を残す計画を立てることができます。 これにより、プレイヤーがそれらのタスクに費やす時間をより細かく制御できるため、達成可能なレベルを維持しながら、チャレンジ レベルをより正確に調整できます。
• 任務にランダム性を追加しても、自動的に任務がより楽しくなったり、退屈でなくなったりすることはありません。 役に立つためには、ランダム性は興味深いものに貢献する必要があり、ランダムな変化によって作成されたすべての可能な任務タイムラインが、すべてのジョブに対して同様のスキル フロア (完了するために必要な能力) を持つようにする必要があります。 (Alphascape V1.0 (Savage) (O9S) や Abyssos: The Eighth Circle (Savage) (P8S) のような、メカニクスの順序に関するタイムラインの分岐は、一般的には問題ありません。(すべての可能なタイムラインを計画する努力をするプレイヤーは、任務に戻るのをやめる前に、それらの計画をすべて使用できると合理的に期待するという点で、限界があります。) 提示されるメカニクスに大きなばらつきがあると、スキル フロアに予期しない違いが生じやすいため、避けるのが最善です。) 間違った場所でのランダム性は、プレイヤーが強力なジョブ アクションの使用を計画する機会を奪い、任務の楽しさを損なうことさえあります。

プレイヤーに初回体験を有意義なものにすることと、退屈な繰り返し体験を避けることを両立させるために、スキル表現を奨励し、報酬を与えて、プレイヤーが自分のペースで任務にさらなる挑戦を加えるようにします。スキル表現は歓迎されるべきであり、恐れられるべきではありません。 (スキル表現は、それを行うプレイヤー以外のプレイヤーには明確である必要はないため、仲間のプレイヤーを不快にさせる必要はありません。) これはジョブ設計と密接に関連しており、各ジョブのスキル表現を定義するのはジョブに依存しています。魔法の遠隔DPSロールのバランス調整に関する以前の記事から引用します:

Quote:

---

[難易度を提供するために任務のみに依存することによる] 5番目の問題は、プレイヤーにとって適切な難易度であっても、特定の任務の寿命に関するものです。レイドで経験したことがあると思いますが、任務の仕組みを学んでいくと、どんな任務でもだんだんと簡単になります。任務にランダム性が組み込まれているとしても、任務を確実に完了するには、やはりバリエーションを学ばなければなりません。実際、任務に慣れると、どんな仕事と装備の組み合わせでも、その任務を完了するためのスキルの最低水準が下がります。(ただし、メカニズムで異なる役割を果たす異なる役割のジョブの場合は、それほどではありません。) その結果、ストーリーからアルティメットまで、どの難易度でも、より良い装備がなくても、完了できる任務は、その後の完了時に十分な注意を払うことができません。では、退屈を防ぐために、残りの注意を払うことができるものは何でしょうか? 必然的に、任務ではありません。そうできるのは仕事だけです。

---

ただし、任務の設計は、プレイヤーに自分の仕事のスキルを表現する機会とツールの両方を与えることができ、また与えるべきです。キャストタイムが頻繁なため、必要以上に移動することは価値がありませんが、AOE がどこにヒットするかをプレイヤーが正確に把握し、不要な移動を排除できるのは、デューティ デザインのおかげです。(パッチ 7.0 のデューティでは、プレイヤーが参照ポイントとして使用できる床の形状が頻繁に欠落しています。) 近接攻撃範囲に制限があるため、敵の近くにいることは価値があります。しかし、タンクが敵を後のメカニクスのために安全な場所に再配置できるようにできるのは、デューティ デザインのおかげです。軽減ツールと回復ツールの使用が制限されているため、それらの使用を計画することは価値がありますが、計画するためのタイムラインとそのパターンを作成するのは、デューティ デザインのおかげです。 (特にダンジョンでは、一度に 2 つのパックと戦うこと (「ダブル プル」と呼ばれる) は、スキル表現の 1 つの形式です。特定のゴミパックに対してプレイヤーにそのオプションを拒否する場合は、代わりにそのゴミパックで別の形式のスキル表現を提示する必要があります。)

満足のいくメカニクスは、プレイヤーのスキル表現に積極的に報い、任務に楽しみと長続きを追加します。 メカニクスやプレイヤー戦略を満足のいくものにするのは、プレイヤーが自分のスキルや任務の知識を示す機会だけでなく、うまくやっていることを明確に確認できることです。パーティがそのように設定したため、前のメカニクスが後のメカニクスに対処する余地を残したり、後のメカニクスに対抗するために使用されたりします。回転する AOE 回避で、最後のヒットがプレイヤー キャラクターの位置にほぼ到達しますが、完全には到達しません。これは偶然ではなく、プレイヤーが回転する距離を知っているためです。パーティが戦闘タイムラインを計画したため、クールダウンを正確に連続して複数回効果的にクールダウンを使用する回復および緩和計画。 (これはハイエンドの任務に限ったことではありません。Eden's Promise: Anamorphosis (E11N) では、魔法の遠隔 DPS が、最初の Burnished Glory で早めに Addle を使用し、その後 2 回クールダウンのタイミングで Addle を使用することで、3 回連続で Burnished Glory レイドワイド Addle を実行できます。クールダウンをドリフトすると、最後の Addle は間に合わなくなります。) また、パーティが設定後に解決するのを見るシーケンスもあります。派手なシーケンスは、パーティに作業の結果を見せ、次のメカニズムの前に一息つく良い方法です。既存の Dawntrail 任務にはこのようなメカニズムがひどく欠けており、その寿命には何の恩恵もありません。

--------

プレイヤーはこれまで、Dawntrail における作業記憶と反応時間への要求について頻繁に言及してきたため、この 2 つの任務設計の側面には特別な注意を払う価値があります。

作業記憶の仕組みは寿司にかかる醤油のようなものです。少量であれば風味が増しますが、多すぎると風味が損なわれます。 (醤油は米ではなく魚にかけるべきです!) 適切な難易度の記憶の仕組みとはまったく別に、パッチ 7.0 には記憶の仕組みが多すぎて、他の形式の難易度を提供できる他の仕組みが排除されています。その結果、退屈で反復的な任務になっています。

適切な難易度の記憶の仕組みについては、プレイヤーの脳には作業記憶の限界があり、それを変更することはできません... そして、その限界は、仕組みを実行するためだけでなく、任務で自分の仕事をするために使用する必要があります。 (はい、ダウンタイムのメカニズムでも、ジョブには回復、バフベースの軽減、リソース構築ツールなど、考慮すべきツールがまだあります。) プレイヤーは、拡張されたワーキングメモリからではなく、スムーズに作業できるより大きな「チャンク」を学習することで、ジョブや任務のメモリ使用量が少なくなるように練習します。 たとえば、中級の RDM は、初心者が「エンチャント リポスト、エンチャント ツヴェルヒハウ、エンチャント リダブルメント、ヴェルフレア / ヴェルホーリー、スコーチ、リゾリューション」と分解しなければならないものを「基本的な近接戦闘コンボ」と考えることができます。

さらに、プレイヤーがチャット ログに記憶のタスクをオフロードすることは期待できません。戦闘中に入力するための APM があっても、手の届く範囲にキーボードやテキスト マクロ用の追加のホットバー スペースがないコントローラーを使用している可能性があります。彼らが頼りにできるのは、任務で提供された情報を脳の外に保存するためのツールが 1 つだけであり、それは周囲のメカニズムが許す場合に限られます。つまり、キャラクターをメカニズムに関連する場所 (直接的または概念的な「最初のセーフスポット」など) に移動することです。(パッチ 5.2 より前は、パーティー全体でウェイマークを移動して、表示されるランダムな変化を思い出すようにすることもできましたが、プレイヤーは戦闘中にウェイマークを移動できなくなったため、メモリ メカニズムではそのオプションはなくなりました。)

では、人間の脳のワーキング メモリの限界はどのくらいでしょうか?そうですね、古典的な「7プラスマイナス2[記憶に保持される項目]」はせいぜい上限であり、より新しい実験は、項目の数が固定されているわけではなく、より多くの項目を記憶しようとすると、それぞれの複雑さと詳細度が低下し、最終的には記憶として役に立たないほど詳細度が低すぎるという考え方を最もよく当てはめています。 記憶すべき項目がすべて、数字の個々の桁など、非常に単純で馴染みのあるものであれば、「7プラスマイナス2」はほぼ正確です。 (実際、市外局番のない電話番号は、「数字の範囲」と呼ばれる特定の作業記憶タスクのわかりやすい例です。)

メモリ メカニクスは、プレイヤーが「スプレッド」や「左」などの適度に単純で馴染みのある概念だけを覚えていればよいまで細分化できることが多いですが、ジョブの回転チャンクはそれほど単純ではありません。また、それをそれほど単純にすると、ひどく退屈になります。したがって、「7 プラスマイナス 2」の下限、つまり 5 つのアイテムを取り、プレイヤーがジョブを続行できるように、1 つのアイテムだけでなく、少なくとも 2 つ、場合によってはそれ以上のアイテムを確保する必要があります (他のメカニクスを事前に計画する余地も増える可能性があります)。したがって、メモリ メカニクスから「まだ解放されていない保存アイテム」の合計は、任務のタイムラインのどの時点でも 3 つの単純なアイテムを超えることはなく、ほとんどの場合それより少なくなります。 プレイヤーの作業記憶が良好であれば、隠れたパターンを探したり、後のメカニクスを計画したりして、ジョブに専念できますし、そうすべきです。

例:

• ボスがランダムに 1 つを選択するメカニズムの繰り返しセットは、メモリ メカニズムの「1 つのアイテム」としてカウントされます: {spread / pair / light party stack / full party stack}、{left / right / front / back}、{in / out}、{Deep Freeze / Pyretic}。
• Eden's Promise: Eternity (E12N) では、保存されたプライマルの組み合わせの地上 AOE は 2 つのアイテムとしてカウントされます。これは、初めてプレイするプレイヤーは一般にプライマルを個別に覚え、重複するセーフスポットを解読するためです。Savage バージョン (E12S) では、同じプライマルの組み合わせが 2 つのアイテムとして始まり、1 つのアイテムとして終わります。これは、プレイヤーが進行の過程でそれらを十分に目にし、重複するセーフスポットを記憶し、個々のプライマルではなくそれらを操作するためです。 (そして、最後のサベージフロアでは、プレイヤーがセーフスポットを思い出すのに苦労している場合は、任務外で5つの可能性を積極的に研究することができます。これにはそれほど時間はかかりません。) FRUは、Ultimateでの外部研究と任務の試行回数に対する期待が高いため、最初からそれらを1つのアイテムとして扱うことができます。
• Seat of Sacrifice Extremeでは、Quintuplecastは3つのアイテムルールに違反しているように見えます。ただし、パーティは、Absolute Stone III Proteansの場合はクロックスポットで、Absolute Holy IIIスタックの場合はお互い(およびボス)に近い場所でメカニズム全体を実行するように手配できます。これにより、プレイヤーはそれらの攻撃がシーケンスのどこにあるかを覚えておく必要がなくなります。残りの3つの攻撃は、プレイヤーが前の2つの攻撃から最後の攻撃を再構築することに抵抗がある場合でも、3つのアイテムルールに完全に適合します。
• Mica the Magical Muに対するLiving Memory World FATEであるMascot Murderは、3つのアイテムルールに著しく違反しています。まずボスはプレイヤーにカード（すべて1から6の番号が付けられている）を覚えておくように促し、アリーナを覆う6枚のカードを見せます。ボスはメカニクスの次のセクションの前にこれらのカードを隠すので、プレイヤーは最大6つのアイテムを覚えておくことになります。次にボスは帽子からカードを1枚引き、プレイヤーはそれをアリーナを覆うカードと一致させなければなりませんが、一致するカードにキャラクターを移動させることができるため、覚えておくべきアイテムは6つのままです。（ただし、このメカニクスの後の繰り返しでは、帽子のカードが表示された後、プレイヤーキャラクターの上に水たまりが落ちるので、回避方法に注意しないとアリーナのカードから追い出されてしまう可能性があります。）次にボスは帽子から2枚目のカードを引き、プレイヤーはそれを再びアリーナを覆うカードと一致させなければなりませんが、今度は7番目のアイテムとして場所を覚えていなければなりません。次にボスは帽子から3枚目で最後のカードを引き、プレイヤーは引いたカードをもう一度アリーナを覆うカードと一致させてから後者の記憶を捨てることができます。プレイヤーがマッチングをどれだけ速く行うかによって、一時的に保管アイテムが 8 個に増えてから 2 個に減ることがあります。(最後に、ボスはプレイヤーに、最初に引いたカード、2 番目に引いたカード、3 番目に引いたカードを答えるように求めます。その後の繰り返しではランダムな順序で、間に他の AOE が挟まれます。) しかし、彼らはずっと仕事をこなしており、数字のスパン テストをしているだけではないことを考えれば、7 個でもまったく受け入れられません。
  World FATE はカジュアルな仕事であり、作業記憶に苦労しているプレイヤーを運ぶことができるからといって、彼らを惨めにするのは適切ではありません。(私の経験では、隠れたパターンがあります (アリーナの床にある安全なカードはエッジを共有していません) が、それは問題ではありません。プレイヤーは隠れたパターンを見つける前に良い経験を積む必要があるだけでなく、注意力が要求される状況ではそれに気づくのが難しくなります。)

--------

反応時間については、1 月にこの件について書きましたが、私が書いた内容を詳しく説明する必要があります:

Quote:

---

2 つ目の問題 [難易度を提供するために任務のみに頼ること] は、アナバシオス サベージでもすでに経験していますが、より微妙な問題です。任務のメカニズムに要求できるのは限られています。プレイヤーは、反応が速く、正確に動き、集中できる時間など、限界があります。ゲームプレイの基本を習得した後は、これらの属性を「自然な」練習で大幅に向上させることはできません。また、これらの属性について意図的かつ思慮深く練習するほど熱心であるプレイヤーでさえ、脳を含む身体の限界に達するまでには限界があります。プレイヤーとしてこのような制限に遭遇するのは、特にそれが障害によるものであったり、人間の身体の短い全盛期を外れている場合、ひどい気分になります。これは、高齢の MMO プレイヤーと、一緒にゲームを楽しみたい家族の両方にとって懸念事項です。(障害を持つレイダーとして、私はこれについて一日中話すことができますが、それはこのシリーズの範囲外です。)

通常の任務とアルティメットの両方が何年もの間これらの制限と格闘してきましたが、エクストリームとサベージも徐々にそれらに直面し始めています。 (一例として、アナバシオス: 第 10 サークル (サベージ) とそのジェイド パッセージ オーブを考えてみましょう。パンデモニアック レイ アニメーション内に出現し、その後すぐに発射されるという事実は、特に後にデーモニアック ボンドと組み合わされて派手なアニメーションと移動要件が追加されると、非常に反応的になります。この種のデザインは、派手なアニメーションを削除したり、情報をより迅速に取得したりするためにサードパーティ ツールを使用することをプレイヤーに奨励し、アナバシオス サベージにはそれが満載です。) プレイヤーはすでに任務のペースが速いことに言及し始めています。

これらの物理的属性は、作成できる戦闘メカニクスの範囲を制限するだけでなく、しかし、どれだけ難しくできるか、そして、プレイヤーの年齢が短い場合でもハイエンドの任務をこなせるようにするには、どれだけ難しくできるかが問題です。仕事に難しさを持たせることを拒否した場合、同じ制限が、曲線の上限をどれだけ難しくできるかにまで及び、満足のいく方法ではなくなります。現在、スキルの向上できる面で行き詰まっているプレイヤーは、さらに高いレベルを目指し、さらに先へ進み、プレイする意欲を維持できます。身体的限界に行き詰まっていて、それを知っているプレイヤー、または他のプレイヤーの身体的限界によって最高難易度が妨げられているプレイヤーは、必然的にやる気を失います。なぜなら、ゲームにどれだけ努力を注いでも、その問題は解決できないことを知っているからです。

---

これを書いたとき、Shadowbringers と Endwalker の過程で、反応時間に対するゲームの徐々に高まるプレッシャーが続くことを想像していました。パッチ 7.0 で、より反応的なメカニズムに一気に移行するとは思っていませんでした。 (特に、エンジンと、クライアントとサーバー間の長い移動に伴う遅延は、反応的なメカニズムにはあまり適していません。) この点では、パッチ 7.0 のデューティ設計の方向性は、役に立つどころか、まったく逆でした。

これは明らかにリアルタイム ゲームであるため、プレイヤーは一定の反応時間基準を満たす必要があります。(見込みのあるプレイヤーが HUD レイアウトとキーバインドを好みに合わせて設定しても、基本的な地上 AOE に 10 秒未満で反応できない場合は、おそらく、プレイヤーが合理的に対応できない何らかの制限があるのでしょう。) ただし、ゲームのリアルタイム性では、プレイヤーは固定の基準を満たし、そこに留まることだけが求められます。プレイヤーにさらに速く反応するよう要求する必要はありません。実際、プレイヤーにさらに速く反応するよう要求することは、反応時間トレーナーとして設計されたニッチなゲームでのみ役立ちます。アクション ゲームでも、最終的にはプレイヤーの身体的限界に達するからです。 ファイナルファンタジーXIVは、元々ターン制だったファイナルファンタジーシリーズの伝統を引き継いでおり、多くの「非ゲーマー」が家族や友人とプレイすることになるため、高齢化するMMOプレイヤーベースを除けば、寛大な反応時間の基準がこのゲームに最適です。

それでは、なぜ Copperbell Mines のアップデート版のメカニクスのペースが The Royal Menagerie (Shinryu Normal) や The Ghimlyt Dark よりも遅いというのは直感的に正しいのでしょうか。結局のところ、それらの任務はすべて MSQ で要求されるものです。

• プレイヤーが MSQ を開始する時点で、タブ ターゲットやホーリー トリニティ MMO のジャンルの詳細はもちろん、一般的なビデオ ゲームに精通しているとは想定できません。ゲームプレイの過程でそれらを習得することが期待できます。Copperbell Mines では、画面に表示される膨大な情報量や、「クラス / ジョブ」、「パーティー」、「ターゲット」などの基本的な概念にまだ慣れていない可能性があります。何かをゆっくりと見せることはできますが、その時点では、プレイヤーがそれを見ることをまったく期待できません。 The Royal Menagerie では、ゲームの基本システムに慣れているだけでなく、基礎概念に精通しているため、「スタック」、「タンクバスター」、「レイド全体」、「追加」などの繰り返し使用されるメカニズムにも慣れています。反応時間はまったく変化せず、与えられた情報を整理して既存の概念に当てはめることができるため、任務で求められていることをより迅速に理解できます。
• プレイヤーは、デフォルトの HUD レイアウトとキーバインドでゲームを開始しますが、これはおそらくプレイヤーにとって理想的ではなく、まったく難しく直感に反する場合もあります (たとえば、ゲームでは、プレイヤーに標準とレガシーの両方の移動モードを試すように指示されていません)。Copperbell Mines では、インターフェイスを自分に合わせて設定する際に何を考慮すべきかがわからない可能性があり、移動キーバインドがまだ遅い場合もあります。The Royal Menagerie では、デフォルトに固執して自分で調整する場合でも、インターフェイスに慣れて確認しています。
• 人間の脳の一般的な高速処理を改善し維持する方法の 1 つは、リアルタイム ビデオ ゲームです。このゲームのように。プレイヤーが反応時間を簡単に改善できる場合、初期の MSQ やその他の通常のタスクを通じてそれを実現します。

これらすべての理由のうち、本質的に無期限に続く唯一の理由は、プレイヤーのゲーム関連の概念のライブラリを拡張することです。そのメカニカル レイヤーは「メカニカル ボキャブラリー」とも呼ばれます。Stormblood タスクの平均的なプレイヤーは、メカニカル ボキャブラリー、ゲーム全体の親しみやすさ、インターフェイスの快適さ、および反応時間において、初期の A Realm Reborn タスクの平均的なプレイヤーを楽々と上回ります。Dawntrail タスクの平均的なプレイヤーは、メカニカル ボキャブラリーにおいて Stormblood タスクの平均的なプレイヤーを楽々と上回りますが、メカニクスを扱う速度とは関係ありません。

したがって、Dawntrail の任務では、プレイヤーが Stormblood のときよりも、初めて任務に挑むときに任務が何を求めているかをより正確に推測し、既存の概念をより効果的に使用し、新しいメカニクスの概念をより速く習得することを期待できます。しかし、Dawntrail の任務のプレイヤーは、Stormblood のときよりも速くメカニクスに対応することはできません。 実際、Stormblood が既存の拡張パックの中で私が望む任務のデザイン スタイルに最も近い理由の 1 つは、練習してもまだ速すぎるメカニクスが徐々に導入される前の最後の拡張パックだからです。

一部のプレイヤーは、メカニクスが反応しすぎて一瞬で終わるように感じると不満を述べています。また、他のプレイヤーは、メカニクスの最初のキューを提示してから、プレイヤーキャラクターがメカニクスを正常に解決するために必要なアクションを完了したかどうかをスナップショットするまでの時間が、実際には 1 秒未満ではないと返答しています。後者の意見は正しいものの、次の 2 つの理由から、プレイヤーのフラストレーションに対する有効な対応ではありません。

• サーバーがメカニクスの最初のキューを送信してから、プレイヤーキャラクターがメカニクスを正常に解決できる正しい状態にあるかどうかをサーバーが確認するまでの間には、プレイヤーが意識的に認識する時間以外にも、決定/反応するステップが多数あります。これらには、キューを示すパケットをプレイヤーのゲーム クライアントに送信するサーバー、それらのパケットを処理してプレイヤーにキューを提供するクライアント システムの遅延、プレイヤーの感覚神経伝導時間、視覚的なキューの場合、プレイヤーが目を動かしてキューに焦点を合わせる時間 (すでにキューを見ている場合、または計画的または偶然に周辺視野でキューを管理するように訓練されている場合を除く)、プレイヤーの感覚知覚処理時間、プレイヤーの決定/反応を神経信号に変換する運動計画時間、プレイヤーの運動神経伝導時間と筋肉反応、プレイヤーの入力を処理するクライアント システムの遅延、プレイヤー キャラクターがクライアント側で必要なアクション (距離を移動するなど) を実行するのにかかる時間、そして最後に、完了したアクションを示すパケットをプレイヤーのゲーム クライアントがサーバーに送り返す時間が含まれます。
  その他のすべての手順を差し引くと、プレイヤーは完全に個別のメカニズムであっても 1 秒未満で決定/反応することができ、その時間を正確に報告できます。 (スタックやスプレッドなどの基本を含む、パーティに対する相対的な行動を伴うメカニクスの場合、パーティが事前に配置を決めていない限り、より多くの手順があります。)
• さらに重要なのは、[B]プレイヤーのモチベーションにとって、身体によって課せられた最小反応時間というハードな物理的壁に実際にぶつかったかどうかは重要ではありません。重要なのは、プレイヤーが現在持っている実用的な反応時間というソフトな壁にぶつかったかどうかです[B]。つまり、次の試みでもっとうまくやりたいという気持ちが、改善するための実行可能な方法につながっていないということです。そして、プレイヤーが不満を言っているのであれば、それはすでにそうなっているのです! (プレイヤーが、合理的で実行可能だと考える改善方法に気付いた場合、不満を言う代わりにそれを実行するでしょう。) モチベーションはプレイヤーの視点にかかっていることを忘れないでください。

今、何かが反応しすぎていると思ったら、おそらく実際にハードな物理的壁にぶつかっているのでしょう。しかし、そこに到達するために私がしたすべての困難を乗り越えるようプレイヤーに頼んだ場合、彼らはほぼ間違いなく従わないでしょう。代わりに彼らは「そんなことを考えるなんて馬鹿げている、ましてや実行しろ」と言って辞めるだろう。

--------

さまざまなタイプの任務でプレイヤーの反応時間を圧迫しないようにすることが何を意味するのか、例を使って検証してみましょう。

基本スプレッドは新生エオルゼアのメカニクスで、通常の任務とハイエンドの任務の両方で何度も登場しています。頻繁に登場するため、パッチ 7.10 で追加された初心者の館の新しいセクションで最初に教えるマーカーになっています。では、なぜ 11 年経っても任務ファインダーの人口はこれを習得していないのでしょうか。経験豊富なプレイヤーが通常の任務で 2 つ以上のスプレッド AOE にキャラクターが当たったり、スプレッド マーカーが積み重なってレイズ可能なすべてのプレイヤー キャラクターが KO されてパーティがスロー KO になったりするのはなぜでしょうか。

デューティ ファインダーは互いに反応することで広がりを見せ、プレイヤーに他のプレイヤーの選択に反応するよう求めると、必然的に反応時間にプレッシャーがかかります。 あなたや私が設定した反応時間の固定基準に関係なく、そのような状況で 1 人のプレイヤーがその基準を満たすと、クライアントとサーバーの遅延により、パーティー メンバーへの警告が少なくなり、反応する時間も少なくなります。プレイヤーは、「他のパーティー メンバーから逃げる」や「頻繁にキャストするプレイヤーは動かない」などのヒューリスティック ガイドラインを学習することで、ある程度まで上達しますが、その方法では、他のプレイヤーに合わせて調整するというループから抜け出すことはできません。パーティー メンバー全員の位置を継続的に追跡しても、完璧になることはありません。

(クライアントとサーバーの遅延だけが調整ループの原因ではありません。人々は対面でループを経験します。典型的な例は、廊下ですれ違うときに、両方が同じ方向に動いているときです。しかし、解決は通常、廊下の方がはるかにスムーズです。「画面では私があなたより先にいた！」ということはありません。)

対照的に、パーティー ファインダーとレイド ファインダーの人口は、ハイエンドの任務に慣れていないプレイヤーや、初めてのエクストリーム ミッションをプレイしているプレイヤーでも、完璧に分散されます。これは、パーティー ファインダー / レイド ファインダーのプレイヤーがゲームに長けているからではありません。任務ファインダーのマッチメイキングでは、ワールド クラスのレイダーでいっぱいのパーティーをまとめることができますが、その環境では分散に失敗する可能性があります。なぜでしょうか?

パーティー ファインダーとレイド ファインダーは、反応ではなく計画によって分散を行うためです。パーティーは、各プレイヤー キャラクターが行く場所について合意し、お互いが自分の割り当てられた場所に行くことを期待します。進行中にメカニズムが再び表示されるのを見て反応するかもしれませんが、お互いに反応しているわけではありません。 (そして、ワールド プログレッション グループが、ノーマル レイドがサベージ バージョンと同時にリリースされた頃に行っていたように、通常の任務でもそのような計画を立てることができます。) 実際、計画に従うのではなく、お互いに反応しようとすることは、「決して調整しない」などのフレーズで嫌われることがよくあります。

--------

もっと広い意味では、通常の任務は任務ファインダーの人口とプレイヤーの期待に合わせて設計する必要があります。

任務ファインダーがマッチメイキングできる「最悪の合理的な」パーティーを想像してください。任務について事前に何も知らない初心者で、全員が最低アイテム レベルにあり、仕事について大まかな知識しかなく、特定のメカニズムを計画する方法を知りたくても知らず、以前に会ったことがなく、ゲームに復帰した人もいれば、長い一日の仕事や介護の後で疲れて水分不足の人もいます (はい、後者は反応時間を悪化させます)。彼らは全員、クエストを完了し、任務に付随するストーリーを続けようとしており、少し焦っています。彼らはその瞬間に最善を尽くしていますが、その瞬間の先を見ていません。また、個人的な進化の壮大な計画も持っていません。

そのパーティーは通常、インスタンス タイマーが切れる前に任務を完了する必要があります。任務を完了しない場合は、次のマッチメイキングで個別にキューに戻って任務を取得できるだけの知識を得たことを正しく認識する必要があります。いずれにせよ、彼らは任務に費やす時間を楽しむべきです。彼らに挑戦し、場合によってはパーティーを数回KOさせても構いませんが、ストーリーの続行を妨げないようにしてください。その結果、次のようになります。

• Duty Finder はゲーム外で準備することはなく、Duty に挑戦する以外で準備することはほとんどありません。 Hall of the Novice アップデートが歓迎される理由の 1 つは、「このソロ インスタンスを試す」が、たとえパーティー全体が協力している場合でも、カジュアル プレイヤーが互いに実行できるほぼ限界であることです。(結局のところ、プレイヤーはビデオを見たりガイドを参照したりするための別のデバイスがない状態でコンソールを使用している可能性があります。)
• Duty Finder は常に反応で動作しています。初心者は見たことのないものを計画できず、頻繁に Duty を完了するプレイヤーでさえタイムラインを記憶している可能性は低いためです。
• Duty Finder の実際の応答時間は速くないため、プレイヤーは十分なリードタイムと長時間の情報可視性を必要とします。 プレイヤーは、必要な情報を最初に提供されたときに適切な場所を見ておらず、インターフェイス内の情報ソース間を効率的に視線を移動させる方法について積極的に考えたことがなく、すべてをまとめるのに時間がかかります。これは非常に大きな影響があり、マッチメイキングされたパーティーよりもセーフスポットに先んじることがよくあります。(パーティーチャットを呼び出したり読んだりするために、キューから完全に目をそらしている可能性もあります。) エンドゲームに落ち着いて以来、現在のタスクで KO されたパーティーメンバーを復活させる点で、健常者で経験豊富なヒーラーよりも定期的に先んじています。これは、デューティタイムラインからでさえない反応テストであるにもかかわらずです。なぜなら、私はパーティーリストを周辺視野に入れながら KO に気付くように自分自身に教えてきたのに対し、彼らはそうしなかったからです。 (時には、パーティーの位置追跡で、迫りくる KO を事前に警告し、プレイヤー キャラクターをターゲットにして、画面に現れる前に Swiftcast を使用することができます。)
• Duty Finder は特定のメカニズムを準備しません。これは、どのようなメカニズムが来るか分からないことの当然の結果です。これが、Thunder God Cid がうまく機能せず、弱体化されなければならなかった理由です。各タワーに 3 人のプレイヤー キャラクターを配置し、6 つの Shadowblade AOE すべてを重複しないように配置することは、分散のより高度なバージョンであり、したがって信頼性には、Party Finder および Raid Finder スタイルで誰がどこに行くかを計画する必要があります。
• Duty Finder に、プレイヤーが焦って任務の開始エリアから逃げ出す前に、厳密に制限された量の一般的な準備 (特定のメカニズムを事前に知らなくても役立つことの 1 つ) を実行させることができます。Stormblood では、その 1 つは「ヒーラーがすべてのパーティー メンバーに Protect をキャストするのを待つ」でした。 Heavensward と A Realm Reborn では、パーティーに CNJ / WHM がある場合は、Stoneskin を待つことも含まれていました。別のタイムラインでは、マクロまたはウェイマークを介して「クロック スポット」について簡単に合意し、それを現在の Party Finder や Raid Finder のようにさまざまなメカニズムに再利用することができました。(本当にやりたいのであれば、今でもそうすることができますが、Hall of the Novice から新しい期待をプレイヤーに教えるためにかなりの時間が必要になります。)
• Duty Finder では、ほとんどのメカニカル ミスは、その試行内で回復して任務を完了できるようにする必要があります。 計画も事前の知識もなければ、合理的な期間内に試行ごとにミスをほとんどしないという基準に達することはできません。そのため、どのようなミスが即時または遅延したパーティー KO につながる可能性があるかについて、メカニズムを慎重に検討する必要があります。ほとんどのミスは、優れた装備、回復、軽減を備えたプレイヤーにとって、個々の KO を保証するものではありません。
• Duty Finder は、特に準備不足によるペナルティをまれにしか受け入れません。 プレイヤーは、理性的である限り、時折のスタックスプレッドや同様の事故に対してパーティーメンバーやゲームに腹を立てることはありません。しかし、そのようなことの発生率が高すぎるとイライラするため、そのような事故を招くようなメカニズムはお勧めできません。
• Duty Finderで熟練したプレイヤーは、ベースラインとはまったく異なる体験をします。スキルを発揮する機会と報酬を与えて、彼らにも良い体験をしてもらうのはあなた次第です。最も簡単な任務であっても、それらの機会を無視しないでください。 ヒーラーの設計と魔法の遠隔DPSの設計に関して説明したように、カジュアルなストーリー志向のプレイヤーの前に置くのに適切なものは、究極はもちろん、極限に慣れているパーティーに挑戦することはできません。 (そして、エキスパート ルーレットで同じダンジョンを 90 回目に見たときに、たとえ「最も理性的でない」プレイヤーであっても、まったく異なる体験をすることになります。) ゲームプレイ ループでは、すべてのプレイヤーが通常の任務に就くことになるため、その異なる体験は、スキルの表現に没頭するか、退屈するかのいずれかになります。任務を単に早く完了するだけでは十分ではありません。強いプレイヤーは、すでに習得したスキルだけで眠りに落ちながら任務を完了することができます。これが、ジョブ デザインと満足のいくメカニズムとのつながりを強調した理由です。

この同じアプローチにより、カジュアル プレイヤーは、ゲーム (または仲間のプレイヤー) が時間をかけて概念を小さな部分に分解し、そのように教えない限り、複雑な概念を学習しません。彼らの努力と報酬の追跡は長期的ですが、彼らはその瞬間にしか努力を惜しみません。カジュアル プレイヤーを一歩ずつ導けば、彼らは山を登ることができます。説明なしに彼らの前に丘を置くと、彼らはそれを壁として認識するだろう。

--------

対照的に、ハイエンドの任務は、それに取り組むプレイヤーのさまざまな期待に合わせて設計する必要があります。プレイヤーは、個人的な進化の壮大な計画はまだ持っていないかもしれませんが、目の前の任務を向上させる計画は持っています。任務ファインダーは瞬間に存在しますが、パーティーファインダー、レイドファインダー、統計は未来に存在します。つまり、仕事のパフォーマンス基準、パーティーの調整、任務のタイムラインに基づく事前の計画、進行中のそれらの計画の実行の練習などです。そして、より多くの努力を払うことで、プレイヤーは任務が自分たちをどのように扱うかという点でより多くの報酬を期待します。ギアの報酬だけでなく、任務自体のゲームプレイ体験にも報酬があります。

(もちろん、これには程度があります。エクストリームのプレイヤーは、キャストバーを見てメカニクスの計画を思い出し、ためらいながらそれを実行するでしょう。これは計画の一般的な考え方としては正しいのですが、アルティメットの計画にはまったく不十分です。たとえば、TOP のウェーブキャノンの最終反復を SCH で正しくシールドすることを考えてみましょう。どちらの場合も、ウェーブキャノンのキャストバー中は、アルファオメガの先行メカニクスのためにパーティが広範囲に動き回らなければなりません。そのため、スウィフトキャストを消費せずに最初のシールドをキャストする正しいタイミングは、先行メカニクスのキューを見ながら、先行キャストバー中であり、シールドがダメージを吸収する前に期限切れにならないように慎重に計画します。(デュナミスのバフの構造上、レイズはほとんど役に立たないので、スウィフトキャストを消費してシールドしても大きなトレードオフはありませんが、そうしない方が楽しいです。) そして、最初のシールドが完全に使用されてから 6 秒後に 2 番目のシールドを配置する必要があります。その間に、SCH も移動してパーティを完全な状態に戻す必要があります。 HP が 100% なので、少しでもためらうとパーティが KO になってしまいます。また、SCH とパーティ全体が、どこでどのクールダウンを使用する予定かに関係なく、すべてのヒットに適用しなければならない他の軽減策も覚えておく必要があります。計画全体は、アルファ オメガの攻撃がそれぞれいつ着地するかを、攻撃が始まる前に知ることに基づいています。これは喜ばしいことですが、ハイエンド デューティを初めてプレイするプレイヤーはそうしません。)

• ハイエンド デューティでは、プレイヤーはデューティに挑戦する以外にも、特定のメカニズムを含め、準備することを予想し、望んでいます。 Deltascape V1.0 (Savage) (O1S) は世界初クリアに 1 回の試行で簡単すぎるというコミュニティの反応を覚えていますか? Savage にとっては簡単すぎますが、正確に言うと、世界初チームが 1 回の試行でクリアできたのは、通常バージョンへの短い露出と、Savage メカニズムを反応で実行するための一般的な準備を利用したためです。彼らの側には優れた反応時間がありました。これはワールド プログレッション チームが選ぶ物理的な特性です。しかし、任務が彼らにメカニクス固有の計画を立てることを強いていたら、パーティの KO は避けられなかったでしょう。(私が初めて O1S に入ったとき、タイムラインがなかったので、反応でメカニクスを実行することができませんでしたが、ワールド ファースト チームが実行できた理由はよくわかりました。) O1S ではメカニクス固有の計画は要求されません。そのため、エクストリームでも簡単すぎます。
• ハイエンドの任務では、メカニクスを学習する過程で、プレイヤーのメカニクスに対する実際の反応時間が大幅に向上し、計画が広範であればあるほど改善されます。ここで、プレイヤーは、任務が最初に必要な情報を与えたときに常に適切な場所を見るように計画し、情報源間を効率的に視線を動かし、今後のメカニクスの可能性のあるランダムなバリエーションにキャラクターを事前配置することができます。さまざまなランダム ソースがどのように組み合わされるかを事前に把握することができます。手を事前に配置したり、アニメーションを最初の視覚的な手がかりにまで絞り込んだりすることを積極的に考えなくても、計画を効率的かつためらうことなく実行する練習をすることができます。
  TOP は、Run:****mi* (Omega Version) のオープニングで、この優れた例を示しています。これは、The Weapon's Refrain (Ultimate) (UWU) の Ultimate Predation に似ていることから、一般に「捕食回避」として知られています。TOP を知らない読者は、Sausage Roll の研究プレイリストで、リアルタイムで捕食回避の多くの例を見ることができます。映像が TOP からのものであることを知らなくても、一般のプレイヤーの観察者は、このメカニクスが自分には適していないことがわかります。彼らはそれを Ultimate メカニクスとして認識しないかもしれませんが、これを Alphascape V4.0 (O12N) に組み込むと悲惨な失敗になることを理解しています。そして、それらは正しいです - たとえそれらの AOE 形状のすべてが Alphascape Normal の一般プレイヤーに知られており、AOE の組み合わせの概念と同様に管理可能であってもです! (オメガFは、O12NでクロスAOEの最適化ブリザードIIIと「ホットウィング」のフロントからバックラインの安全なAOEであるスーパーリミナルスチールをキャストします。オメガMも同様に、O12Nで至近距離AOEの効率的なブレードワークとドーナツAOEのビヨンドストレングスをキャストします。実際、ビヨンドストレングスのセーフスポットは、最適化ブリザードIIIとの組み合わせに対応するためにTOPでより大きくなっています。どちらも、O12NでTOPと同じ武器の兆候を持っています。(彼らはO12Nでも同時にAOEをキャストしますが、そこでは彼らの位置と向きはタンクの選択に依存し、組み合わせはより制限されます。)そして、カジュアルプレイヤーはファイナルオメガと戦うことはできませんが、彼らはラティテュディナルインプロージョン/ロンジチュディナルと同じセーフスポットを持つアルファスケープV1.0 (O9N)の砂時計AOEシーケンスを見てきました。インプロージョン。) カジュアル プレイヤーは、床にオレンジ色の電報がなくても AOE パターンを覚えているという考えさえ持っています。では、なぜプレデーション ドッジはカジュアル プレイヤーの手の届かないところにあるのでしょうか。
  このメカニズムは純粋に個別の回避であるため、調整は必要ありません。避けられないダメージはないため、回復や軽減の要件もありません。また、フェーズ全体にハード エンゲージがありますが、ダメージ出力の要件はカジュアル プレイヤーが見ているものではありません。彼らが正しく認識しているのは、すべての情報が与えられているにもかかわらず、十分な速さでどこに行けばよいかわからないということです。
  カジュアル プレイヤーは、オメガ M とオメガ F のペアが常に反対側のインターカーディナル上に出現することを知らないため、インターカーディナルを見てメカニズムを開始するときに、自分のキャラクターをカーディナル上に事前配置してすべてのインターカーディナルが画面に表示されるようにすることも、1 つのクローンを見つけてその武器を処理した後すぐに反対側のインターカーディナルを見ることも知りません。 TOP の進行のこの時点で、アルティメット レイダーはそれをすべて知っており、ブラインド進行であっても、TOP パーティー シナジーで 1 つのペアが先に出現したことで、反対のスポーン位置を予想するように準備されています。(実際、私は各クローンの腕を見て、プレデーション ドッジのペアをすべて確認します。武器の 3 つがそこに表示され、「武器を持っていない」は常にスーパーリミナル スチールのスケートであるためです。パーティー シナジーにのみデコイ クローンがあります。)
  カジュアル プレイヤーは、キューが表示されながら、各組み合わせのセーフスポットの重なりを解明する必要があります。TOP の進行のこの時点で、アルティメット レイダーはパーティー シナジーを学習したときに各クローン ペアの組み合わせを記憶しており、今では各クローン ペアがファイナル オメガの砂時計クリーブとどのように重なるかを記憶できます。 (そしてそこから、各組み合わせが他の可能な組み合わせにどのように流れていくか、すべて任務外で。これが、研究プレイリストが存在する主な理由です。)
  カジュアル プレイヤーは、ゴールド オーバーレイの有無にかかわらず、ダーク グレーのクローンのライト グレーの武器をミッド グレーの床から選択する練習をしていません。TOP の進行のこの時点で、アルティメット レイダーはパーティー シナジーのためにそれを十分に練習しており、ここでもいくつかの個別の KO (パーティー KO につながる可能性があります) を完全に想定しています。実際、この時点でアルティメット レイダーは TOP で何百ものパーティー KO を経験しており、カジュアル プレイヤーは任務完了のためにそれほど長く粘るつもりはありません。
  捕食回避の合図の表示速度を大幅に遅くして、カジュアル プレイヤーが任務中に重複を解く時間を与え、同時に 1 つのクローンを武器の移行/作成アニメーションと共に表示し、その後「フリーズ」して 2 つずつ発射することで、プレイヤーの視線をアリーナ全体に誘導すれば、そのバージョンのメカニクスは O12N に組み込むことができ、AOE をフル スピードで解決しても、カジュアル プレイヤーにとってほとんど問題にはなりませんでした。(残る唯一の問題点は、最適化されたブリザード III と超能力の組み合わせで、ウェイマークなしでオメガ M から正しい距離を計算することです。) この仮想の通常レイド バージョンと既存のアルティメット メカニクスの違いは、純粋にプレイヤーに要求される処理速度であり、アルティメット レイダーは特定のメカニクスを計画し、練習することでその速度を獲得します。基本的な反応時間はまったく変更する必要はありません。私の場合、プレデテーション ドッジは、反応時間が頭打ちになってからずっと後になってから、数時間の学習と練習で恐ろしいものから快適なものへと変わりました。スナップショットを知る前でも、2 つのセーフスポット間の移動距離が最も長い「外側」の組み合わせ (最適化されたブリザード III + 効率的なブレードワーク) を除いて、移動に余分な時間があります。
• 計画と職務スキルの基準を守る努力と引き換えに、プレイヤーは、ハイエンドの任務に妥当な努力を注ぐことで、（練習して正しく実行すれば）反応時間に十分な余裕が生まれ、毎回「クリーン」に任務を完了できる計画が生まれることを期待しています。（「クリーン」とは、「KO、ダメージダウン、またはメカニクスの失敗によるその他のペナルティがない」という意味です。失敗したメカニクスからの回復には、必然的に失敗に対する反応が伴います。）ハイエンドの任務は通常の任務よりも速く進むことができ、そうあるべきであり、一般的にはそうしていますが、それはプレイヤーに計画を立て、適切な流暢さで実行させるためだけです。必要な情報をすべて事前に十分に与えてから、特定の時間に行動すること、または一時停止せずに一連のアクションを実行することを要求するのは、はい。その情報の一部に、他のプレイヤーのキャラクターに与えられたデバフやマーカーを含めるのは、はい。次のメカニクスまたはメカニクスの一部に関する情報を受け取って処理しながら行動することを要求するのは、はい。限られた時間だけ必要な情報を表示することはできますが、最も厳しい制限は Ultimate でのみ実行可能であり、プレイヤーに処理の順序と手がかりの記憶を強制します。その結果、プレイヤーがまだ学習中は、特定のメカニズムを効率的に処理するための計画を練っていないため、メカニズムが反応的であると感じられることがあります。しかし、その感覚は、そのメカニズムに対するプレイヤーの実際の反応時間が向上するにつれて、適度な練習で消えなければなりません。プレイヤーがメカニズムを学習したら、反応時間が実質的に要因にならないだけの十分な時間が残っているはずです。習熟はターンベースで、このリアルタイム形式で古い Final Fantasy ゲームの感覚を取り戻す必要があります。計画段階ですでに戦いの半分は勝利しています。
• したがって、プレイヤーが他のプレイヤーの選択に反応することを要求するメカニズム、またはそのような反応を排除するすべての計画が任務にふさわしくないほど精巧である場合、ハイエンドの任務では受け入れられません。 (Eden's Promise: Anamorphosis (Savage) (E11S) Prismatic Deception のように、他のプレイヤーが渡す必要がある情報に応答するには、2 人のプレイヤーの反応時間と追加のクライアントとサーバーの往復に必要な長いリード タイムと、タイピング、マクロ スペース、またはゲーム外の通信に頼らずにプレイヤーがその情報を渡す方法が必要です。E11S はそれを正しく行いました。しかし、そのようなメカニズムは決して高速ではありません。) はい、これは、一部のメカニズムが通常の任務では許可されているが、ハイエンドの任務では許可されていないことを意味します。プレイヤーはハイエンドの任務での個々の試みにより多くの労力とリソースを費やしており、強制的な反応のためにパーティー KO または個人的なペナルティを受けるのは迷惑です。 (この点では TOP はまったく問題ありません。Run: ****mi* (Sigma Version) と Run: ****mi* (Omega Version) で提示される「オープンな選択肢」に対して、プレイヤーが過度の計画や実行の労力をかけずに優先順位システムを考案できるからです。パーティー ファインダーのプレイヤーが代わりにサードパーティ ツールに頼ったのは、デューティ デザイナーのせいではありません。(技術的には、Run: ****mi* (Delta Version) にも「オープンな選択肢」がありますが、これは Patch の Remote Regression および Local Regression テザーを使用してさらに簡単に割り当てられます。) 手動でサインを配置することで、パーティーの優先順位システム管理の負荷を軽減する方法も知っていますが、個人的にサインを配置する APM を持っていないだけです。)
• ハイエンドの任務では、プレイヤーは任務が失敗に対する厳しいペナルティを伴う致命的なダンスを提示することを期待しています。これは難しさの一部であるだけでなく、これらの任務に挑戦するスリルの一部でもあります。ハイエンドの任務で、少数のミスを許容する全体的な調整ではなく、特定のメカニクスに失敗した場合に寛容さを組み込む傾向がある場合、そのメカニクスはそもそもハイエンドの任務に適していない可能性が高いです。寛容さを求めることは、メカニクスが適切な計画に従わないことの兆候です。
• ハイエンドの任務では、プレイヤーにベースラインの経験を得るためにある程度のスキルが必要ですが、プレイヤーは苦労に見合った満足のいくメカニクスと、ベースラインを超えてスキルを発揮する十分な機会を積極的に期待しています。すべてのメカニクスが正しく実行するためにそれ自体が満足できるものである必要はありませんが、AAC ライトヘビーウェイト サベージ全体にそのようなメカニクスが 1 つもないのは残念です。

ハイエンドの任務では、プレイヤーはパズルを解き、全体像を把握できる程度まで仲間のプレイヤーから複雑な概念を学びます。(プレイヤーがバハムートの封印の巻 - ターン 5 でツインタニアのダイブボムを傾けることを学んだ後、アリーナの床はほぼ完全に平らになっているかもしれませんが、彼らの本当の 3 次元は常に時間です。) 最高のプレイヤーは、任務のタイムライン全体とパーティー全体を 1 つとして扱い、8 人のプレイヤーによって実行される単一の分岐計画を作成します。そして、その見返りとして、計画レイヤーと実行レイヤーの両方が、任務のすべての動きにスムーズに対抗する喜びを提供してくれることを期待します。プレイヤーのパーティー全体が、古い ファイナル ファンタジー ゲームでキャラクターのパーティーを制御する単一の「プランナー」プレイヤーの役割を担い、計画を実行する際の個々の表面的な役割も担います。

したがって、ハイエンドの任務では、努力する価値のある計画の問題を提供してください。スプレッドシート、ストライキング ダミーでの時間、単一の任務での時間に対して報酬を与えてください。反応的でなければ簡単なメカニズムでは、決してそれはできません。

--------

欠点はあれど、Heavensward、Stormblood、Shadowbringers はこれらの原則に十分近かったので、私は戦う意欲が湧きました。怯える新米から Ultimate 対応まで、着々と成長しました (Shadowbringers には私にぴったりのジョブがほぼあったので、間違いなく助けになりました)。また、アラガン トームストーンの週上限に達してからずっと後も、ジョブをレベルアップさせることなく、ただ楽しむためにルーレットをしました。Sastasha でも気にしませんでした。ジョブに取り組むことで、そこで楽しむことができたからです。

Endwalker が私のジョブを壊し、それは私にとって大きな打撃でした。この手紙はデューティ デザインに関するものですが、私は常にシステム指向のプレイヤーであり、ジョブは戦闘のコア システムです。私は、身体的制限に対処する個人的な向上が私にとって重要だったし、今も重要だったため、レイドを続けるためにロール スワップを行いました。アラガン トームストーンの週上限に達し、ジョブをレベルアップするまでは快適にルーレットをしましたが、それ以上はしませんでした。

ドーントレイルは基本的に私のジョブを変えませんでした。しかし、アラガントームストーン装備を使用できるジョブがあるにもかかわらず、週の上限に達するように自分自身に積極的に思い出させる必要があったのはこれが初めてです。何年も続けてきた習慣が2週間も経たないうちに壊れました。ドーントレイルの任務は肉体的に非常に厳しいため、私の脳は過度の怪我のリスクを認識し、私を遠ざけてしまうからです。（もちろん、パッチ7.0の任務設計の方向性によってFRUから外されることを余儀なくされたという事実は役に立ちません。安全に究極に取り組めない場合、究極レベルのスキルはあまり役に立ちません。）そして怪我のリスクはさておき、パッチ7.0の任務は退屈です！

ドーントレイルが私を動かしたのは、RuneScapeをもっとプレイすることです。ある意味、あのゲームはこれよりもアクセスしにくいです。戦闘システムは、古い手動の Embrace SCH よりもはるかに高い APM 上限があるため、完璧にプレイすることはできませんが、少なくとも戦闘システムを自分のやり方でソロでプレイすることができ、ゲームは私に報酬を与えてくれます。彼らは私のために何かを持っていて、それは私に合った任務がないよりはましです。

彼らの 最新のボスラッシュダンジョン は、とても楽しいだけでなく、攻撃テレグラフと任務サポートトライアルでの制限付き KO システムをコピーして、プレイヤーのメカニクスを教えています... KO の許しをオフにするハードモードが大人気です! 繰り返し登場するメカニクスのコンセプトの「チュートリアル ボス」 には、指導的な NPC だけでなく、ランダムな順序で個々のメカニクスを選択できる機能があり、プレイヤーは練習に集中したいものを選ぶことができます... また、すべてのメカニクスをアクティブにして強化する独自のハード モードもあります。ボスがセットを完了すると、メカニクスはランダム化されて繰り返されるため、ボスが 2 回続けて同じメカニクスを選択し、最初のプランのクールダウンが利用できない場合に備えて、プレイヤーは各メカニクスによる KO を回避するために少なくとも 2 つのプランを用意しておく必要があります。 また、ハード モードは、個々の完了に対して低エクストリームから中サベージの難易度までスムーズに調整され、最大 200 キルを連続して達成するとより良い戦利品が獲得できる「キルストリーク」システムが提供され、そのたびにスケーリングが強化されます。これにより、Ultimate レベルの一貫性がうまく報われます。彼らの 最新のギャザリング ボス でさえ、スキルの表現を奨励し、報いるために、遅いが労力の少ない戦略から速いが労力の多い戦略まで、意図的にさまざまな戦略を許可しています。

（ちなみに、RuneScapeはあなたの任務設計の歴史の最良の部分をコピーしているだけではありません。彼らはまた、昨年完全な教育クエストラインを備えたペットクラスの独自のバージョンをリリースするなど、真剣にツールチップ、戦闘フィードバック、および全体的な新規プレイヤーエクスペリエンスを改善してきました。その結果、意図せずして彼らのプレイヤーコミュニティはあなたの「ジョブゲージ」という用語を知ることになりました。なぜならプレイヤーが彼らのためにジョブゲージを作成し、それが十分に人気を博し、他の戦闘形式をサポートし始めたからです。彼らのプレイヤーは文字通りあなたの任務設計の歴史の最良の部分を彼らにもたらすことができます。インターフェース。)

適切に設計すれば、ゲーム内でレイドほど素晴らしいものはありません。RuneScape は、戦闘のチャレンジとして真剣に検討されることはないでしょう。しかし、パッチ 7.0 の任務設計の方向性は、長期的な楽しみと適切なチャレンジの両方を見失っています。

私たちには、どうすることもできない障壁ではなく、私たち全員が挑戦して克服したいと願うチャレンジを与えてください。

As with previous parts of this letter, for your convenience I have attempted to provide a Japanese version using Google Translate, since I unfortunately do not speak Japanese myself. Please pardon any mistakes so generated. In case of any confusion, please refer to the English text if at all possible, as I cannot verify the accuracy of the machine translation.

この手紙のこれまでの部分と同様に、残念ながら私自身は日本語が話せないため、皆様の便宜を図るため、Google 翻訳を使用して日本語版を提供するよう努めました。間違いがあった場合はご容赦ください。機械翻訳の正確さは確認できないため、混乱が生じた場合は、可能な限り英語のテキストを参照してください。

--------------------------------

次に、既存の Dawntrail の任務とその仕組みについて、任務設計の観点から、良い点と悪い点の両方について見ていきます。念のため言っておきますが、このセクションは必然的に退屈な内容なので、興味のないプレイヤーの読者は、Honey B. Lovely の任務の再設計にスキップしてください。(ナビゲーション リンクはこのスレッドの先頭にあります。)

MSQ ソロ インスタンス任務、ハント、トレジャー ダンジョン、通常の FATE、およびワールド FATE など、特定の任務は、正確な仕組みと調整を繰り返してレビューするには非常に時間がかかります。テスト サーバーにアクセスせずに徹底的にレビューするには膨大な時間がかかるため、これらを省略して、利用可能な 21 の繰り返し可能なグループ任務に焦点を当てました。省略した任務のほとんどは、含まれている任務とほぼ同じ設計スタイルに従っているため、これと原則の説明は、それらの問題によく当てはまるはずです。

----------------

長さのため、このセクションを 4 つのサブセクションに分割し、許容可能な難易度と予想される難易度の順に並べました。

まず、パッチ 7.00 MSQ に登場する任務から、登場順に紹介します。これらの任務の完了はプレイヤーがゲームを続ける上で機能的に必須であるため、復帰した非常に弱いプレイヤーにも励みになる体験を提供するために、これらの任務は最高水準に保たれなければなりません。彼らは、戦闘やその他のすべてのコンテンツへのアクセスがこれらの任務の完了に依存していることを知っており、その能力を心配すると、その知識が重荷になります。より強いプレイヤーがこれらの任務を完了し、その後も楽しめるようにするには、これらの水準を犠牲にしてはいけません。

----------------

Ihuykatumu (レベル 91 ダンジョン)

https://i.imgur.com/SiqXaNm.png

最初のボス前のモブ

3 つのパック (川船上)

これらは強制的なシングル プル (一度に 1 つのパック) であり、基本的な地上 AOE を少し行うだけです。プレイヤーにシングル プルを行わせる場合は、強いプレイヤーが飽きないように、個別に興味深いものにしてください。各モブが死ぬたびに前のパックの残骸を食べに来る魚を追加したり、Bakool Ja Ja の「ギフト」からのパターン AOE などを追加したりします。ダウンタイムのない Holminster Switch の強制シングル プル セクションほどリソースを消費する必要はありません。

拡張パックの最初のダンジョンのオープニングなので、プレイヤーが久しぶりにグループ デューティをこなす可能性は十分にあるため、穏やかなオープニングが妥当ですが、それでも面白いかもしれません。これはそうではなく、MSQ のためではなくレベルを上げるためにダンジョンをプレイする場合、ストーリー シーケンスがさらに不快なものになります。

最初のボス - Prime Punutiy

https://i.imgur.com/LIrdYH7.png

Hydrowave

狭い円錐状の AOE で、簡単に回避でき、特別なものではありません。問題ありません。

Resurface

キャスト バー後の AOE スパムは、その速度で視覚的に圧倒的です。AOE スパム中にドーナツ AOE をキャストする Flytrap アドが現れ、プレイヤーは自分のキャラクターの足元だけでなく他の部分にも目を向けざるを得なくなります。プレイヤーが自分のキャラクターの足元だけを見ることができれば、視覚的には高い方ですが許容範囲です。なぜなら、プレイヤーはキャラクターの足元に焦点を合わせることで、気を散らすものを排除し、関連する手がかりを見る時間を短縮できるからです。

私はこれを行うことができますが、楽しいことではありません。特に、ハエトリグサについて知らない最初のときは、できる限りあらゆる場所を常に見るようにするだけです。ハエトリグサについて知っている場合でも、ハエトリグサは通常、キャラクターの足元からかなり離れているため、AOE スパムが終了するまで、アリーナ全体を見上げたり、キャラクターの足元を見下ろしたりと、視覚的に行ったり来たりする必要があります。

代わりに、ボスがキャストバー中にハエトリグサを引き上げ、アリーナに入れてプレイヤーの敵意リストに入れたとします。 AOE シーケンスはその後まったく同じように展開されますが、初めてプレイするプレイヤーは、ドーナツ テレグラフが表示されて自分のキャラクターの周囲全体をカバーしてしまう前に、フライトラップを見て「これは何か効果があるかもしれない」と考える可能性が十分にあります。今後の試みでは、より強いプレイヤーは AOE スパム中にフライトラップの既知の場所まで移動するオプションがあり、それが自分の仕事とどのように相互作用するかがより興味深いものになります。

プヌティの歌 (追加)

これは良いメカニズムで、拡散による軽いダメージが追加によるタンクの回復の増加に反映されます。パーティーが追加のほとんどの位置を、ターゲット可能になったときにおびき寄せられるようにすることで、スキルを表現できる機会を提供できますが、ここでは 2 つの拡散が非常に大きいため、あまり効果がありません。

Shore Shaker

簡単だけど楽しい。自分のキャラクターの前の電信が消えたら、そこに踏み込むだけです。プレイヤーは電信の連続表示を見て、そこを見る必要があるとわかっているので、タイミングは問題ありません。

全体的な印象 (ボス)

このボスは、Resurface が問題になることを除けば、かなり簡単です。拡張パックの最初のグループ ボスとして、ビジュアル デザインの問題から始まると、すぐにプレイヤーの懸念が生じます。

第 1 ボスと第 2 ボスの間のモブ

第 1 パック - コウモリとウィドウメイカー

特に面白いものはありません。ただし、ダブル プルが可能なので問題ありません。

第 2 パック - ドブリンおよびトレント

同感です。

このパックが死ぬと、ラダーをターゲットして操作する必要がなくなるのはありがたいです。このダンジョンの後のラダーでも同じです。

第 3 パック - ミミクロット

「彼らは私たちの姿をコピーした」というアイデアはクールですが、どうやらこれらの敵をスタン耐性にする以外には何も行われていないようです。また、パーティーに WHM がいなければ、プレイヤーはそれを見ることはないでしょう。役割の「クラシック」ジョブに由来する基本的な AOE をそれぞれに与えるといった単純なことでも (その認識の一部はジョブの変更によって損なわれていますが)、コンセプトは機械的に売り込まれます。タンク (MRD/WAR) には正面の円錐、ヒーラー (AST 以外のヒーラー) には至近距離の AOE、DPS (LNC/DRG) には直線 AOE。

(ヒーラーにはそのような AOE がありますが、他のヒーラーにはそのような AOE はなく、キャストも Holy には見えません。さらに、没入感を高めるために最初に Duty Support を使用し、次に他のプレイヤーの体験を妨げないように Duty をレビューしたにもかかわらず、自分の完了ではなくビデオからそれを見つけなければなりませんでした。)

このパックではダブル プルが可能なので、退屈ではなく、単に機会を逃しただけです。

第 4 パック - Doblyn と Widowmakers

まだ面白いものはありません。

第2ボス - ドロウジー

https://i.imgur.com/KbPsEVV.png

種をまく

このキャストでプレイヤーが種が落ちた場所を見ることを期待しているのであれば、それは無理な期待です。種は小さすぎて、表示される時間が短すぎるため、プレイヤーが4つのオブジェクトを追跡できません。プレイヤーは後のドロウジーダンス、テレグラフ、ウォーロップで種を見るのに十分な時間があるので、これは主に「警告、何かが起こった」キャストになります。

ドロウジーダンス

テレグラフで種を育てます。まあ、当然です。

ウォーロップ (追加によるキャスト)

ここでは対面マーカーが役に立ちます。これにより、プレイヤーは各根の根元を見るだけで済むため、これを回避できます。各ルートを上に向かってたどって向きを探すのは大変です。

ミミクロット

ミミクロットは以前のゴミから戻ってくるので、デューティ サポートでこれを完了したときのように、単に AOE の餌食になるのではなく、毎回メカニクスを実行する絶好の機会です (プレイヤーはゴミに基づいて適切な推測を行うことができます)。ボス戦で AOE の餌食になることは時々問題ありません。回復要件が変更され、全員が単一ターゲットと AOE のローテーションを切り替える機会が提供されますが、これはそれ以上のことをする機会を逃しています...

他のプレイヤーの完了のビデオを見ると、ミミクロットにはここでメカニクスがあります。デューティ サポートはなぜメカニクスを完全に排除できるのでしょうか?

くしゃみ

大きな円錐、それをかわすための長いキャスト、問題ありません。

全体的な印象 (ボス)

このボスの名前と眠り始めるという事実は、こっそり通り抜けることを示唆しており、自分のキャラクターに基づいてそのままにしておきたいプレイヤーと、倒して XP が欲しいプレイヤーの間には必然的に緊張が生じます。その緊張を取り除くには、おそらくこれは他の主張者のチームの 1 つが戦闘を強制するポイントであるべきでしたか?

メカニズム的には、これは問題ありません。

2 番目と 3 番目のボスの間のモブ

4 つのパック

ここではメカニズム的に興味深いものはありませんが、2 回引っ張ることができます。

テーマ的には、死んだ動物はボスへのクールなヒントですが、それはメカニズムではありません。

3番目のボス - アポリオン

https://i.imgur.com/kDPMpD9.png

レイザーゼファー

基本的な AOE ライン、妥当なキャスト時間。

出現する獣

テーマ的にクールな「警告、何かが起こった」と、ボスがパワーアップしたことをほのめかす赤いテキストが気に入っています。

群がるイナゴ

ボスがシーケンスの事前の通知なしに飛び回るのは好きではありません。視覚的な追跡の観点からだけですが、このような状況では適度に許容されます。ジャンプはそれほど速くなく、次の攻撃ではボスの正確な位置を知る必要はありません。

Blades of Famine

強化されたライン AOE で、4 つあるため、ジャンプ後にボスから離れることはプレイヤーにとって迷惑ではなく、むしろ有益です。

ジャンプを含めて 2 回繰り返されます。

Levinsickle

このメカニズムはかなり気に入っています。放射状の円錐の各セットは、そのシーケンス全体にわたってアリーナをカバーし、2 つあるため、プレイヤーは空のスペースから重なり合ったスペースに足を踏み入れたくなります。すべての情報が明確に示され、適切な時間があるため、プレイヤーは考えるよう促されます。

(キャストバーなし) 出現する獣

別のミニフェーズで、今回はタンクバスターの後にドーナツを追加します。

Windwhistle

ライン AOE の雪片が脈動している間、結果として生じる Whirlwind に目を光らせておく必要があるのは好きではありません。実行可能だと思いますが、動きが明確に予測できないため、視覚的にストレスを感じることは間違いありません。

全体的な印象 (ボス)

Windwhistle にもかかわらず、これは実際にはかなりしっかりしたボスです。他の捕食者を食べ、その攻撃を引き受けるボスというコンセプトは良いもので、Tower of Zot の最終ボスのような圧倒的な性質なしに、「プレイヤーがすでに見たことがあるが、強化された / 連続したメカニズム」をある程度実現できます。

全体的な印象 (ダンジョン)

退屈なオープニングは素晴らしいとは言えません... また、Resurface がダンジョン全体に迫っています。最初から「常にあらゆる場所を見る」モードになりましたが、それは楽しくありません。

--------

ワーコール ゾルモア (レベル 93 ダンジョン)

https://i.imgur.com/Ec8sDtk.png

最初のボス前のモブ

4 つのパック

ここではメカニクス的に興味深いものはありませんが、2 回プルできます。

最初のボス - リョウコール テルテ

https://i.imgur.com/t3L2W4N.png

フラッフル アップ

ボスが一部の追加を凍結することで変更する、追加からの AOE シーケンス。最初の 2 つは、さまざまなタイプの追加のパターンを示しています。カーディナルの小さな Rorrloh Teh は、アリーナの前方の 4 分の 1 を攻撃します (各攻撃のセーフスポットは互いに向かい合っています)。インターカーディナルの大きな Quorrloh Teh は、周囲の至近距離を攻撃します (各攻撃のセーフスポットは互いに隣り合っています)。最初の 2 つは問題ありません。

氷のテザーの視覚的な明瞭さは良好で、プレイヤーはボスから追加攻撃に目を向けることができます。

Snowscoop

アリーナの中央を通る 2 セットのライン AOE の間をうまく横切ります。

Sparkling Sprinkling

これらのスプレッドは、前のメカニズムの実行中に表示され、プレイヤーは近くにいる他のプレイヤーに気付いて回避しようとします。Duty Finder はスプレッドの場所を計画しないため、プレイヤーがスプレッドを重複させることはありますが、これは MSQ にとって十分に簡単なメカニズムの重複です。

Fluffle Up (繰り返し)

これで 2 つのパターンが組み合わされ、ボスは 4 つの氷のテザーを出します。ボスは両方のパターンを反転させていますが、それぞれの反転は以前に個別に表示されていましたが、初めてのプレイヤーはそれを知らず、個々のパターンが常に反転していることさえ知らないかもしれません。ボスが各追加を個別に操作して、2 回ではなく 3 回のヒットのシーケンスを作成しているか、パターンの 1 つだけを反転している可能性があります。

したがって、テザーが出てくると、以前は機能していた「最初のセーフスポットに立って/注意し、テザーが現れたら、テザーされた追加がヒットするエリアに移動する」という方法は機能しなくなります。初めてのプレイヤーは、どの追加が最初に爆発するかを思い出す必要がありますが、それ自体が 4 つのものが一度に表示された空間記憶メカニズムとして簡単ではなく、それらをテザーと一致させる必要があります。与えられた時間内でそれを分解するのはあまり合理的ではありません。

プレイヤーが戦闘後にこのメカニズムについて考える時間が持てば、それは問題ない、あるいは簡単すぎるものになります。「ボスは両方のパターンを反転させるので、オレンジ色がすべてある場所に立ってください」というのは明らかな推測であり、正しいものです。しかし、これは MSQ ダンジョンであり、プレイヤーの最初の試みは重要なストーリー体験であり、多くのカジュアル プレイヤーはメカニズムに隠されたパターンという概念をまったく持っていません。

これは修正が難しいメカニズムです。繰り返しプレイするプレイヤーを退屈させるほど戦闘を遅くしたくないからです。ダンジョンよりもトライアルや通常レイドの方が適しているかもしれません。これらのゲームでは、待機中の呪文マーカーでプレイヤーをリフレッシュするなどのことができる長いボスが許され、さらに別のコンセプトのための時間も残されています。

全体的な印象 (ボス)

これは、追加を凍結することで AOE シーケンスのタイミングを操作することに焦点を当てた、優れたボス コンセプトです。初めてのプレイヤーにとっては、後半のフラッフルアップのキャスト中だけ、手に負えなくなります。

最初のボスと 2 番目のボスの間のモブ

4 つのパック

ここでは機械的に興味深いものはありませんが、2 回引くことができます。

2番目のボス - Kahderyor

https://i.imgur.com/a8N44lx.png

クリスタルクラッシュ

ここでのスタックタワーは、パーティーが直後のウィンドショットで成功するように準備します。このボスは回避不可能なダメージがかなり大きいため、特に便利です。

ウィンドショット (プレイヤードーナツ + クリスタルセーフ)

パーティーがタワーがあった場所に留まり、クリスタルが出現した場合、大丈夫です! この風のメカニズムが何をするのかを知る良い入門です。ただし、風のAOEの色は、より鮮明な緑にする必要があります。

アーゼンショット (拡散 + クリスタルから離れる)

これも良い入門です。プレイヤーは、異なる要素には異なる対応が必要であることを合理的に推測するはずです。最後の対応は「タワーがあった場所に留まる」でした。ただし、土の AOE の色は、以前のメカニズムよりもさらに鮮明な黄色にしたり、後のメカニズムを考慮して別の要素に置き換えたりすることもできます。

クリスタル ストーム (ライン クリスタル)

クリスタルを残す 3 つの小さなライン AOE。

これが、エレメント AOE に鮮明な色を使用できる理由です。プレイヤーが 3 つのラインすべてからの AOE を一度に確認し、ドーナツとスプレッドの両方でパーティの動きに反応する必要がある場合、高コントラストはビジュアルをすばやく処理するのに役立ちます。(各要素を 1 回しか見たことがないため、初めてのプレイヤーはクリスタルの場所とエレメントのセーフスポットのつながりに気付かない可能性があります。カジュアル プレイヤーは、入る前に考えることがたくさんあります!)

ウィンド ショット (プレイヤー ドーナツ + クリスタル セーフ)

どのラインにも足を踏み入れると地上 AOE を回避できるため、プレイヤーはすべてのラインの位置を覚えたり思い出したりする必要はありません。ドーナツを積み重ねたり広げたりするのに十分なスペースがあります。

アースン ショット (拡散 + クリスタルから離れて)

今ではスペースがほとんどなく、パーティーは拡散のために互いに反応し、アース AOE は薄茶色の上に黄色です。これは、完全な色覚異常ではないとしても、赤から緑の範囲で色覚が低下している人にとっては悲惨なことだと思います。

真面目な話、アース メカニズムは火や雷に置き換えるべきだったのではないでしょうか。火はすでに地上 AOE に関連付けられており、雷はすでに拡散に関連付けられており、埋め込まれたクリスタルは紫色です。

シード クリスタル

クリスタルの追加から抜け出します。簡単です。

シャープニング サイト

視線 + AOE ミニ フェーズに入り、最初に スタラグマイト サークル (至近距離 + ボスから目をそらす) をキャストし、次に サイクロン リング (ドーナツ + ボスから目をそらす) をキャストします。

視線と AOE の組み合わせは非常によくできており、プレイヤーに自然に成功する組み合わせを示してから、地上 AOE をかわすときに視線を尊重するように求めます。敵のアイコンではなく収縮する視線リングを使用することは、このコンテキストでプレイヤーに視線を認識させる優れた方法です。特にボスの頭が画面外にある場合があります。

全体的な印象 (ボス)

エレメント エフェクトは、クリスタル ラインから来ているときに視覚的な明瞭さを維持するために、地上でもう少し目立つようにする必要があります。それ以外は、しっかりしたボスです。

2番目と3番目のボスの間のMob

4つのパック

ここでは機械的にはあまり面白くありませんが、ダブルプルが可能です。優秀なタンクはクラウドトラップをグループに組み込むことができ、またそうすべきです。その逆も同様です。

3番目のボス - グルファール

https://i.imgur.com/WI7qdQj.png

ストーンワーク (火、拡大する正方形のAOE + スプレッド)

最初から組み合わせのメカニズムですが、初めてのプレイヤーには大きなものを期待し、このメカニズムが石板から来ていることを理解するのに十分な時間を与えています。

ストーンワーク (水、ノックバック)

これは組み合わせではありませんが、水のノックバックはプレート自体ではなくプレートのある側から来ることをプレイヤーに示しており、これは後で重要になります。

スレッジハンマー (マルチヒットラインスタック)

レイドヒーリングの観点からは、これはそれほど大きな打撃を与えませんが、そうするとカジュアルプレイヤーには大きすぎます。大きなマーカーと威嚇的な名前が付いているこのスローキャストは、カジュアルヒーラーが反応するよりもパーティーを癒す準備をするように誘い、パーティーの残りのメンバーが軽減するように誘うのに十分であり、どちらも良いスキルです。

アーケイン ストンプ (ダメージアップ オーブ)

ここでインターセプトするオーブの数が多いため、より速いセクションが作成され、プレイヤーは移動ツールを効率的に使用してアリーナを自分たちで分割するように促されます。巨人の遅くて強力な攻撃を犠牲にすることなく、ボスを「攻撃的なダミー」にしないのに適しています。(繰り返し実行すると、プレイヤーはボスがオーブをインターセプトするのはアリーナの中央に到達したときだけで、ヒットボックスに到達したときではないことに気付く可能性があり、その事実を利用して動きを小さくします。)

注目すべきことに、オーブによるバフはプレイヤーキャラクターでは 1 分間しか持続しませんが、ボスでは無期限に持続するため、カジュアルプレイヤーの視点から見ると恐ろしく見えます。また、これはレベルアップ ダンジョンなので、暗黙的にアイテム レベルが厳密に同期されています。

Enduring Glory (レイド全体)

これもレイド ヒーリングの観点からはそれほど大きな影響はありませんが、ダメージ アップ後のレイド全体は概念上恐ろしいものです。

ボスのダメージ アップ スタックをすべて消去しますが、これはおそらく必要ありません。現状の Sledgehammer は、ヒーラーが Anabaseios トーム ギアを持ってきた場合、Liturgy of the Bell / Seraph / Macrocosmos / Panhaima で完全に打ち消すことができます。これだけのリード タイムがあり、ダメージ アップ スタック専用のメカニズム全体があることを考えると、カジュアルなヒーラーは複数の AOE ヒーリング ツールを見つけるように求められます。これにより、スタックの半分または 3 分の 2 を消去できます。複数の AOE ヒーリング ツールが見つからなくても、パーティは KO されるのではなく、HP が低いままになる可能性が高いでしょう。

(そして、もしパーティーがダメージアップスタックの残りの一部をパーティー KO にしてしまった場合、ダメージアップの知識があれば、もう一度挑戦することができます。アーケイン ストンプの明白なわかりやすさにより、(分別のある) プレイヤーはオーブに気づき、敵のバフを見なくてもオーブがボスをパワーアップすると推測するため、ヒーラーを責めることはまずありません。)

ストーンワーク (風、竜巻 + ダブル ノックバック)

プレイヤーにノックバック耐性の使用を考えさせる組み合わせです。プレイヤーは距離がわからないため、センター プレートからの最初のノックバックはコーナーに運ばれる可能性が高く、竜巻が始まる前にプレイヤーが竜巻の位置取りをうまく決めることができます。 2 つ目は、プレイヤー キャラクターがアリーナを出る直前に竜巻に突っ込む可能性があるため、すぐには楽しくありません。しかし、任務設計の観点からは、プレイヤーに問題を提示しながらも、それを解決するためのツールが手元にあることを確認するため、これは素晴らしいことです。

Stonework (火 + 水)

これは恐ろしいように見えますが、これにより、プレイヤーは組み合わせとメカニズムの相互作用について考えるようになります。水のノックバックは、正方形の火の AOE の最初のヒットと一致するようにタイミングが調整されているため、プレイヤーはキャラクターを安全な場所に押し流し、スプレッドに移動するだけで済みます。以前に組み合わせのピースを確認したため、プレイヤーはタイミングを推測し、ノックバックの直後にキャラクターを移動する準備ができます。

Sledgehammer (マルチヒット ライン スタック)

前と同じですが、今回はレイド ワイドの直後に発生します。キャストが長いため、回復するのは依然として非常に合理的です。

全体的な印象 (ボス)

素晴らしい! ダンジョンのボスとしては、これに近い。視覚的に明瞭で、声は「歴史の重み」というアイデアを売り込んでおり、メカニクスは、初心者でも自分のミスを理解できるほど穏やかである一方、強いプレイヤーには頻繁に組み合わせや最適化の機会がある。このスペクタクルは、ストーリーを伝え、プレイヤーにヒントを与えるのに役立ち、これはすべてのコンセプトで可能というわけではないが、うまく機能すれば素晴らしい。

すべてのダンジョンのボスがこのように遅い必要はないが、巨人の場合は、より遅く、より強力な攻撃でうまくいく。

全体的な印象 (ダンジョン)

問題なく全体的に楽しく、少なくとも繰り返しクリアすると問題は改善される。メカニクスはテーマを非常にうまく実行している。

--------

ワーコール・ラー・ドール（ヴァリガルマンダ、レベル93通常トライアル）

https://i.imgur.com/JZXYC62.png

（キャストバーなし）アニメーション テル

このボスには、3つの要素を使用した基本攻撃のアニメーション テルが3つあります。風の渦巻くドーナツ（プラットフォームの安全地帯の中央）、火の踏みつけ（近接攻撃範囲を離れる）、稲妻のくちばしの円錐（アリーナの安全地帯の前隅）です。これらのアニメーション テルは、最初に赤いテキストで教えられることになっており、その後は他のヒントなしでプレイヤーが認識するはずです。

それでは、これら3つのアニメーション テルと、それらを教えるためのテキストについて考えてみましょう。

• 風：「ヴァリガルマンダの周りには激しい風が集まっています。」これは、プレイヤーにボスのモデルを見て渦巻く風を見るように指示し、アリーナの前面にドーナツがあることを示唆しています...そこは風がある場所ではなく、風はプラットフォームの周りを回っており、ヴァリガルマンダは後に渦巻きに入ります。プレイヤーがここから学べる最も正確な教訓は、ポップアップテキストが間違っている可能性があるということです。
• 火:「ヴァリガルマンダは力を集めます。」これは、どこを見るべきかについてはまったく何も述べていません。(ボスの体が赤く輝き始め、脚がプラットフォームに近いため、どこを見るべきかはボスの脚です。)
• 雷:「ヴァリガルマンダの大きく開いたくちばしの中に、目がくらむようなエネルギーが蓄積されます。」これは、どこを見るべきかについて正しい説明を示していますが、このアニメーション中、デフォルトのカメラの傾きまたはそれよりわずかに高い傾きで、ヴァリガルマンダはくちばしを画面から上げており、ほとんどのプレイヤーはカメラの傾きを調整することはありません。確かに、キャラクターのカメラ方向または非常に横から見ることによって地面の視認性を犠牲にすることなく、ヴァリガルマンダのくちばしを快適に画面に表示する方法はありません。その結果、プレイヤーは、このサインを見ながら他の地面の AOE をスムーズに回避するためにインターフェイスと格闘することになります。さらに悪いことに、このアニメーション中、ボスも赤く光ります!

https://i.imgur.com/bS0ncqM.png

これらの教育テキストはどれも役に立たないか間違っています。サインの 1 つは、画面に確実に表示されていません。そして、これは、任務中に他のことが起こっている間、プレイヤーがアニメーションだけで認識することを意図しているメカニクス セット用ですか?!

これに対して私がしたことは、すぐに「常にあらゆる場所を見るようにする」という考え方を始めて、その後の MSQ 任務でもその考え方を維持することでした。また、当時は私にとって謎だったアニメーションの手がかりを見つけるよりも、彼らの追跡の方が得意だと気付いた後、この任務のアニメーション メカニズムについて任務サポート NPC に従うこともしました。これは楽しいことではありませんが、ガイドを見つけて、通常モードのトライアルのタイムラインを見つけて記憶する以外では、私にとって最善の選択肢でした。

他の手がかりがなくてもアニメーションの手がかりを持つことができます。Aurum Vale の古いバージョンではそれが実現されていて、問題ありませんでした。注意すれば、大きなボス モデルにアニメーションの手がかりを持つことができます。Alphascape V2.0 (O10N、Midgardsormr) ではそれが実現されていて、問題ありませんでした。ただし、プレイヤーが実際に手がかりを学習し、常にあらゆる場所を見ようとしなくても、一貫して手がかりを見ることができるようにする必要があります。「このキャストバー攻撃の直後にこのアニメーションの手がかりメカニズムに当たった。どういうわけか、両方を同時に見なければならない」では、実行できません。

(このボスがアニメーション テル攻撃を使用するときには、ある種のリズムがあります。私はエクストリーム バージョンを実行した後でこのトライアルに戻る方がはるかに快適でした。しかし、エクストリーム バージョンを完了していないプレイヤーにとっては、これは役に立ちません。さらに、これは高レベルのトライアルであり、デューティ サポートがあるため、トライアル ルーレットをキューに入れたプレイヤーでも、自分の初回以降は非常にまれにしか見られません。初回の経験は重要であるだけでなく、デフォルトであり、プレイヤーはそのリズムについて何も知りません。)

ボスのコンセプトに大きな変更を加えずに、これを大幅に改善する方法の 1 つ:

• 風: 指導テキストを「ヴァリガルマンダは周囲に激しい風を巻き起こします。」や「ヴァリガルマンダはプラットフォームの周囲に激しい風を巻き起こします。」などに変更します。プレイヤーは、アリーナの隅に集まる風という、妥当なリードタイムで表示される特徴を見るように指示される必要があります。風の効果には、コントラストのためにもう少し緑色を使用することもできます。
• 火: 指導テキストを「ヴァリガルマンダは脚に火を流します。」に変更します。
• 稲妻: 指導テキストは不完全ではありますが正確です。アニメーションの特徴を、ボスがプラットフォームに嘴を下ろすところから始まるように変更します。これにより、プレイヤーはデフォルトのカメラの傾きでボスを確実に見ることができます。アニメーションの輝きを、ヴァリガルマンダの頭からのみ紫色の稲妻の光になるように変更します。これらの変更により、色と動きも火の攻撃とは劇的に区別されます。
• すべて: これらの変更により、プレイヤーは、ヴァリガルマンダの片方の脚の休止位置のほぼ周辺に「目を休める」ことができます。これは、すべてのアニメーション テルがそれらのスポットのすぐ近くにあり、教育テキストでそのことが示されているためです。この「休止ポイント」の概念は、アニメーションを見るタイミングについて別のヒントを与えることなく、プレイヤーにアニメーション テルを見せるために不可欠です。(Aurum Vale の「休止ポイント」はボス自体であり、それらはかなり小さく、見るキャスト バーもほとんどありませんでした。O10N は、画面の大部分に表示される非常にドラマチックなアニメーション テルに加えて、毎回プレイヤーにヒントを与えます。)「この任務では、敵のキャスト バー / 自分のキャラクターの足 / パーティー リスト / ... を見ていないときは、このスポットを見てください」は実行可能であり、前のメカニクスの後でプレイヤーが「休止ポイント」に目を戻す時間があるときはいつでも、アニメーション テル メカニクスを使用できます。

これらのアニメーション テルに関する 2 つ目の問題は、このボスは実際には繰り返しが始まる前にプレイヤーにすべての異なる要素を示すようにプログラムされていないことです。風、火、風、(その他の攻撃) をロールし、最終的に任務の途中で雷攻撃を教えることになります。これは任務がすでに増加している場合には適切ではなく、「ああ、神様、どこもかしこも見て」という考え方につながります。

O10N のように、各アニメーション テルをプレイヤーに複数回表示する必要はありません。また、固定された順序で表示する必要もありませんが、メカニズムの導入期間で一貫してメカニズムを導入する必要があります。高いランダム性が必要な場合は、最初の 1 つは完全にランダムにし、2 つ目は使用されていない 2 つの要素のうち 2 つ目を完全にランダムにし、3 つ目は残りの要素にします。

オープニング フェーズ

アニメーション テルを除けば、これは十分にシンプルです。実際のところ、それはほとんどアニメーションで説明されており、プレイヤーがそれらを学ぶ方法について適切に指示されていない場合、ほとんど役に立ちません。

Skyruin

レイド全体にフェーズ遷移の警告がわかりやすく表示され、アリーナの前面にランダム要素がはっきりと表示されます。

氷の段階

(キャストバーなし) 雪崩

ここで、赤いテキストはプレイヤーにカメラを回して雪崩を確認するように指示する便利なもので、実際に毎回表示されるため、プレイヤーはカメラを常に回す必要はありません。これはそのままでも問題ありませんが、私はプレイヤーがテキストだけで回避を学べるエクストリーム バージョンの方が好きです。

凍りつく塵

警告マーカーが移動期間の開始までカウントダウンし、凍結デバフがその期間の終了までカウントダウンする、優れたマークのディープ フリーズです。

(キャストバーなし、スノーフレーク AOE) アーケイン スフィア

赤いテキストの縮小する黄色の下線は、AOE が発動するタイミングを軽く示しています。私はこれで問題ありませんが、赤緑色覚異常のプレイヤーは苦労するかもしれません。

しかし、これらのセーフスポットは比較的小さく、このフロアにはプレイヤーがセーフスポットの端がどこにあるかを把握するためのジオメトリがありません。短いテレグラフに基づいて反応的に自分の位置を修正しようとするのは楽しいことではありません。このようなメカニズムをサポートするには、プレイヤーが雪片を視覚的にトレースできるフロアジオメトリをいくつか提供します。これは、Eden's Verse: Reflugence (E8N) とその Savage バージョン (E8S) で実証されているように、雪/氷のフロアでは完全に可能です。

これは、これらが雪崩や Freezing Dust と組み合わされる可能性がある火フェーズの後ではさらに悪化します。

Thunder Phase

Thunderous Breath

私はこれが気に入っています。これは稲妻のくちばしの論理的な強化であり、テレグラフはプレイヤーに、オレンジから抜け出す以外に利用できるツールがあるはずだとすぐに知らせます。これにより、プレイヤーは床の浮遊セクションを試し、後のメカニクスの使用方法を確実に理解できるようになります。

羽根の雨

ここでは、エクストリームとは異なり、プレイヤーは羽根がどのようなパターンで出現するかを知らないため、キャスト自体がレイド全体にマルチヒットし、後で避けるべき羽根を配置することになります。羽根は床に対してもっとコントラストがあってもよいですが、羽根が大きいことは役立ちます。

枯れボルト

このコンセプトは気に入っています。浮遊が常に望ましいわけではないことをプレイヤーに示し、浮遊セクションを離れると消えるテレグラフも完備しています。赤いテキストは、プレイヤーが稲妻に対抗する方法を考えるよう促し、「地面に落ちる」ことは論理的に理にかなっています。

ただし、実行には2つの問題があります。まず、正しく実行したとしても、別のプレイヤーがうまくいかず、近くの安全なエリアにボルトが当たることで、プレイヤーは攻撃を受ける可能性があります。 2 番目に、先ほどの羽根は、このキャストが終了すると同時に至近距離の AOE を持ち、プレイヤーにそのことを知らせるための雷撃は、プレイヤーがキャラクターの足元で「浮遊しない」を探しているときに発生します。これにより、特に羽根はどれも浮遊しないので、羽根のロジックはプレイヤー キャラクターの場合と異なるため、プレイヤーは意図されているものを見ることができなくなります。

(キャスト バーなし、ライン AOE) アーケイン サークル

これらは氷フェーズの雪片と同じサークルであり、うまく機能します。赤いテキストはプレイヤーに注意を払うように指示し、ボスからのテザーはプレイヤーに場所を伝え、浮遊している / 浮遊していない床セクションは、少なくともセーフ スポットの 1 つの端を見つけるのに十分なジオメトリを提供します。最適化のためにプレイヤーが望む端ではないかもしれませんが、プレイヤーは短いテレグラフに反応する必要はありません。

これらは、Thunderous Breath と組み合わせると、ファイア フェーズ後も引き続き有効です。

フェーズの追加

Ruinfall

スタック タンクバスター タワーは、3 つのマーカーを組み合わせているにもかかわらず非常に明確であり、タンクが直後のノックバックで Arm's Length を使用するように促すクールなコンセプトです

(キャストバーなし) ライン スタック + コーン AOE の繰り返し

デューティ ファインダーは、メカニクス固有のプランを作成しません。デューティ サポートでさえ、NPC が地上 AOE を回避する場所についてプレイヤーに明確なプランを提供しません。円錐 AOE を回避するのは面白くなく、スタックを取るためにパーティー全体をまとめようとするのは大失敗です。

プレイヤー キャラクターがデューティ サポートでスタックを持っている場合、NPC と同じように回避しないことで、ここで 1 回だけ許された KO を無駄にしてしまうというコイントスがあります。また、メカニクスの学習にすでにスタックを使用している場合は、実際には間違いではない何かのために最初に戻されます。

ファイア フェーズ

Tulidisaster

適切に大きく、長く、恐ろしいキャストで、複数のヒットが前のフェーズのさまざまな要素を通過してから、永続的な DOT でファイア フェーズを開始します。

永続的/長時間持続する DOT は、別の種類の回復チェックとして興味深いもので、カジュアル プレイヤーにとって十分に優しい適切な導入です。

噴火

これらの水たまりは、かなり不透明で赤くする必要があります。水たまりがいっぱいになると十分に見えますが、パーティがいっぱいになると見えなくなる可能性があります (特に最初の回避後)。また、1 つだけでは、赤くなった床に対してきれいに見えません。

全体的な印象

ハイデリンのために、アニメーション テルを作成する場合は、プレイヤーが実際にテルを学習し、いつテルを探すべきかを認識できるようにしてください。分岐フェーズの順序や永続的な DOT などの楽しいアイデアは、あらゆる場所を探そうとする絶え間ないストレスによって台無しになっています。

--------

スカイディープ セノーテ (レベル 95 ダンジョン)

https://i.imgur.com/MXRKnOY.png

最初のボス前のモブ

4 つのパック

ここではメカニクス的に興味深いものはありませんが、2 回プルできます。

最初のボス - フェザー レイ

https://i.imgur.com/aXBVeOk.png

厄介な尻尾

キャスト、各プレイヤー キャラクターの周りの迷惑なデバフ効果、および「あなたはパーティーの迷惑になるように呪われています...」というテキストは、ボスが自分のキャラクターにどのようなメカニクスを適用しているかに注意する必要があるという適切な警告を提供します。

Worrisome Wave

狭い前方円錐 AOE で、数秒後に各プレイヤー キャラクターがこれを繰り返します。頭上のマーカーは、何かが起こりそうなことを示す優れた手がかりを提供します。Troublesome Tail を見たプレイヤーは、何が起こるかをよく推測できます。そこからは、パーティ メンバーの動きに本質的に反応しているにもかかわらず、十分な時間があります。

Hydro Ring

領域を水で覆ったままにするドーナツですが、実際には次のメカニズムの準備です。

Blowing Bubbles

「Feather Ray が大量の泡を発生させます...」で示される弾幕シーケンスを開始します。泡のセットが 4 つあり、各プレイヤー キャラクターに 1 つずつ向けられているため、プレイヤーは少なくとも 1 回はキャラクターを動かす必要があります。

これは私のタイプのメカニズムではありませんが、この形式では、パーティ メンバーがどこにいたか、したがって泡がどこに向けられているかを確認するためのリード タイムがあるため、許容できます。

ローリング カレント

アリーナ上に静止した大きな泡の配列を作成します。この泡はノックバックによって押し出され、配列が見えるだけの時間があります。最初は、プレイヤーはノックバックが自分のキャラクターに当たるのか泡に当たるのかわかりませんが、ノックバックが解決するのを見るとわかります。これが繰り返されると、プレイヤーは成功するための知識を得ます。

ハイドロ リング

アリーナを再び効果的に縮小し、弾幕戦をさらに開始します。

厄介な尻尾

ここで本当の問題が始まります。

厄介な泡

ボスはここでパーティーに直接泡を放出しませんが、各パーティー メンバーは「フェザー レイが大量の泡を生成...」というテキストが表示されたらすぐに、他のすべてのパーティー メンバーに泡を放出します。つまり、プレイヤーは 3 つの異なる方向から自分のキャラクターに向かってくる泡を処理し、自分のキャラクターが放出する泡をフィルタリングし、近くの空きスペースを探し、次の泡が新しい場所に向けられるためこれを繰り返す必要があります。すべて反応的に行われます。その結果、視覚的に圧倒的な高速移動泡の入力による反応シーケンスが生まれます。

ああ、これに失敗すると、個々の KO が非常に早く発生します。プレイヤーが最初のローリング カレントを成功させたとしても、2 つの泡でタンク以外の敵をフル HP から KO するのに十分です。プレイヤーは、物理的な反応時間に関係なく、「フェザー レイは大量の泡を生成します...」というテキストが消える前に、簡単に自分のキャラクターを KO することができます。 (プレイヤーがトラブル バブルのキャストバーを見るのはこれが初めてであることを覚えておいてください。プレイヤーは弾幕とプレイヤー キャラクターによる繰り返しのメカニクスを予想する理由がありますが、繰り返しのメカニクスが始まる前に、異なるキャストバーの異なるパターンと頭上の警告を予想する理由もあります。)

さらに悪いことに、これらのバブルが非常に速く移動するという事実は、プレイヤーのクライアント サーバー間の ping が中程度であっても、サーバーがバブルの位置を含むパケットをプレイヤーのクライアントに送信し、プレイヤーのクライアントがプレイヤー キャラクターの位置をサーバーに送り返すまでの往復で、各バブルがかなりの距離を移動することを意味します。(高 ping プレイヤーが現在の居住地と一致しない地域のデータ センターに接続する前に、一部の地域は非常に広い地理的領域と古いインターネット インフラストラクチャにまたがっていることに注意してください。) その結果、プレイヤーが画面で見る位置は、サーバー上の戦闘の実際の状態とは大幅に異なります。これにより、視覚的に明らかなよりも回避時間が短くなるだけでなく、矛盾の形態により、プレイヤーは、サーバーが自分のキャラクターが攻撃を受けたと表示しているのに (ダメージと脆弱性アップに基づいて)、自分のキャラクターが同じバブルを回避している位置を示しているのを見ることがよくあります。これは、プレイヤーが「私はそこから抜け出せた!」と叫ぶのにうってつけのレシピです。

もちろん、同じクライアント サーバー遅延は、原則としてすべてのメカニズムに適用されます。これは、レイダーがハイエンドの任務で移動オブジェクトのメカニズムを研究するときに回避することを学ぶものであり、カジュアル プレイヤーでさえ、キャストバーの終了を判断するときに ping を直感的に考慮することを練習で学びます。しかし、これらのバブルの速度とターゲット設定により、すでに反応性の高いメカニズムで不公平に見える視覚状態が生まれ、カジュアル プレイヤーにすぐにわかり、メカニズムと壊れた没入感の両方に苛立ちを感じます。

泡がプレイヤーキャラクターの発射位置に対して固定された非常に単純なパターン、たとえば「常にキャラクターの背後」であれば、このコンテキストでは許容されます。「キャラクターのすべての中間基数」のようなより複雑な固定パターンは、パーティーがパターンに沿って計画できる Criterion では許容されます。現状では、失敗は視覚的な過負荷や反応時間から生じる可能性が高いため、これは楽しくないストレスの山です。

全体的な印象 (ボス)

プレイヤーが弾幕シューティング ゲームを望んでいる場合、このゲームではなく弾幕シューティング ゲームをプレイすることになります。仲間のプレイヤーが泡を発射する場所への反応は面倒で、楽しいものではありません!

第 1 ボスと第 2 ボスの間のモブ

2 つのパック

ここでは機械的に興味深いものはありませんが、ダブル プルが可能です。

スカイディープ採石場からスカイディープ船へ

これが、時々強制的にシングルプルを行う理由です。スピーキングストーンは強制チェーンプル（すでにプルされたモブが死ぬ前に次のパックを追加）になり、クールダウンとローテーション管理のテストになります。

2番目のボス - 銃器

https://i.imgur.com/zAIw0zJ.png

ミラーマニューバ

キャストは短いですが、プレイヤーが次のメカニズムを複雑にするミラーを見る（または少なくとも探すように促す）良い機会です。

サンダーライトバースト

これは楽しいメカニズムです。危険なAOEはアリーナの側面と同じ半径で巨大であり、プレイヤーはボスの狙いを追ってミラーを通り抜けてオーブに到達し、AOEがどこから爆発するかを把握する必要があります。しかし、長くて明確なキャストにより、プレイヤーがこれを行うのは非常に合理的であり、最初に失敗した場合に何が起こったかを理解できます。後の繰り返しにより、ミラーバウンスが増えますが、妥当なままです。

古代の砲兵

次のメカニズムのために AOE が拡大する場所を設定します。

出現する砲兵

もう 1 つの楽しいメカニズム。最初からクロス AOE を組み合わせ、最初は 2 つの AOE (4 つのセーフスポット) で導入し、その後 3 つの AOE (1 つのセーフスポット) を使用します。3 つの AOE の反復にはその後のスプレッドもあり、タイミングは穏やかな組み合わせに非常に適しており、スプレッドはクロス AOE 拡大の終わり近くに表示され、クロス AOE の後に解決されるため、プレイヤーは自然にキャラクターを前のセーフスポットの端に移動してから外に出ることができます。

最初は単純な側面がありますが、大きなアリーナに適合し、プレイヤーがキャラクターの移動に多くの時間を残すことができ、それでも楽しいです。

全体的な印象 (ボス)

良い! ダンジョンのボスにはこれに近い。プレイヤーはメカニクスに失敗するかもしれないし、複数失敗するかもしれないが、なぜ攻撃を受けたのかを理解するための情報はすべて持っている。そして、シンプルであるにもかかわらず、アリーナが広いため、動きの最適化に取り組みたいプレイヤーはそうするだろう。

2番目と3番目のボスの間のモブ

スカイディープ パッカー

バリアの理由や、ダンジョンのアレクサンドリア セクションを明らかにするための急な曲がり角も理解できるが、なぜこれが 1 回のプルなのだろうか? バリア ガーディアンか何かを入れればいいのに。

2 つのパック

ここではメカニクス的にはあまり面白くないが、2 回プルできる。石工の HP が高いのは、弱いプレイヤーの体験に影響を与えずにスキルを表現できる良い機会であり、強いプレイヤーは意図的に石工をターゲットにして、単一ターゲットのクールダウンとフォールオフ AOE を使用できる。

3番目のボス - マウルスカル

https://i.imgur.com/gHEuY7P.png

ストーンカーバー

キャストバー、「マウルスカルが拳に力を引き寄せる！」という赤いテキスト、そして各AOEがロックインするたびに光るパルスのオーディオキューの間にある、非常にわかりやすい「パンチ」AOEシーケンス。床にテレグラフが表示されないため、これは技術的には古い形式のメモリメカニズムですが、初めてのプレイヤーでも2回目のヒットのために1つの{ステップスルー/ステイ}を覚えるだけでよいため、まったく問題ありません。

この教えは、これがコンビネーションで再び登場するときの基礎を築くために重要です。

スカルクラッシュ

プレイヤーはノックバックサークルをよく知っているはずですが、比較的小さなアリーナでフロントセンターの配置を見るのはこれが初めてかもしれません。これにより、プレイヤーはアリーナの後ろの隅に向かってノックバックに乗ることを狙う機会が得られます。

フォローアップのスプレッドは問題なく、プレイヤーが広がる方向が限られており、後のメカニズムの前に練習する機会があります。

Maulwork

より高速でドラマチックな AOE シーケンス。速いように感じても問題ありません。

Deep Thunder

大きくて恐ろしいマルチヒット スタック タワー。ヒール チェックというよりは見ごたえがありますが、通常の任務では難しいヒール チェックはできません。

Ringing Blows

Stonecarver と Skullcrush を組み合わせたもので、ノックバックから抜け出すときにプレイヤーが素早く移動する必要があります。これは、カジュアル フレンドリーな方法で興奮を生み出す優れた方法です。これは、高速で派手な動きのシーケンスですが、プレイヤーはすでにすべてのピースを見ているためです。

巨大な衝撃

アリーナの後ろの角からノックバックし、その後にスタックまたはスプレッドが続きます。何回か発生します。

これに関する私の唯一の問題は、スプレッド バリアントでは、プレイヤーのキャラクターの視界がボスによってブロックされる可能性があることです。(カメラをノックバック サークルに向けると、プレイヤーはノックバックを希望の角度に向けることができますが、ノックバック後はカメラがアリーナの前面と側面に向いてしまいます。ボスが非常に大きく、アリーナの上に迫っているため、トップダウン ビューでも同じ問題が発生します。) プレイヤーはインターフェイスではなくボスと戦う必要があります。

代わりにノックバックがアリーナの前方の角からであれば、プレイヤーの視界がボスによってブロックされる可能性は大幅に低くなります。ボーナスとして、これにより、プレイヤーはノックバック後にジョブの最適化、近接範囲への移動 (またはスプレッド中に遠距離ジョブをプレイしている場合はその場に留まる) について考えるより自然な機会が得られます。 (現状では、プレイヤーは Arm's Length を使用しないときに後で移動することで近接 DPS またはタンクを最適化しようとすることができますが、狙いを定める時間が短くなるため直感的ではありません。)

全体的な印象 (ボス)

インターフェースと戦う瞬間を除いて、全体的にほぼ良好です。ベテラン レイダーは、Eden's Gate: Sepulture (Savage) (E4S) のメカニズムの多くを認識できるというボーナスを得られますが、もちろん象徴的な E4S のメカニズムは、Duty Finder やダンジョン ボスの短い形式には適していません。

全体的な印象 (ダンジョン)

ここでの唯一の大きな問題は最初のボスですが、Trouble Bubbles はダンジョン全体を本当に台無しにします。残りはもう少し磨きをかける必要がありますが、それでもかなり楽しいです。

----------------

1 つの投稿に 20 枚の画像制限があるため、このサブセクションを 2 つの投稿に分割する必要があります。続きは下にあります。

1 つの投稿に 20 枚の画像制限があるため、これは上記と同じサブセクションの続きで、パッチ 7.0 MSQ 内の任務に関するものです。

----------------

ヴァンガード (レベル 97 ダンジョン)

https://i.imgur.com/ZV58QV6.png

最初のボス前のモブ

2 つのパック

ここではメカニカルに興味深いものはありませんが、ダブル プルできます。

ヴァンガード セントリー

これらはダブル プルできるだけでなく、そのオープニング AOE によってメカニカルな興味が増し、タンクが最初のパックに近づく方法が変化します。(モブは遠くからタンクに気づきますが、通常はテレグラフ ダッシュによりパーティの後ろに回ります。) このようなものを見るのは楽しいです。

最初のボス - ヴァンガード コマンダー R8

https://i.imgur.com/VKkmBq2.png

強化された機動性

すぐに空間推論メカニクスが使用可能になり、「ウィングブレードに注意！」というテキストで、テレグラフされた AOE の向こう側に側面があることをプレイヤーに警告します。素晴らしい仕事です。

バイク ダッシュを実行するたびに、短い強制ダウンタイムが発生します。強制ダウンタイムがあってもまったく問題ありません。タイミングは非常に明確なので、強いプレイヤーはダウンタイムに合わせて計画を立てることができます。

ディスパッチ

バイク アドを召喚して独自の基本メカニクスを実行します。最初のセットはプレイヤーがゴミ箱で見たダッシュを実行し、2 番目は最初のテレグラフから拡大する大きな円形 AOE を実行します。どちらのタイプも同じキャストバーを持っていますが、前者は地面に留まり、後者は空中に飛び上がるため、初期段階では視覚的に区別できます。

後期バージョンでは、同時ダッシュ（互いに垂直）または連続した円形 AOE のいずれかで、2 セットのアドがあります。大きな危険ゾーン、フェアなタイミング、特に円形シーケンスがコーナーからコーナーではなくコーナーからセンターであるためです。

エレクトロサージ

同時ダッシュと同時に発生するスプレッドで、各プレイヤーにダッシュ用の異なるセーフスポットを見つけるように指示します。スプレッド マーカーが見える時間はあまりありませんが、スプレッド マーカーが表示される前にプレイヤーがダッシュ セーフスポットをうまく把握しているはずなので問題ありません。

全体的な印象（ボス）

非常にフェアで空間指向のボス。 (私にとっては遅い方ですが、空間推論に苦労している人にとってはおそらく恩恵でしょう。これは MSQ ダンジョンです。) すべてのボスが空間推論にこれほど重点を置くべきではありませんが、これはその重点を置いた堅実なボスです。これは間違いなく私が望むデザイン スタイルにふさわしいものです。

第 1 ボスと第 2 ボスの間のモブ

2 つのパック

ダブル プルのオプションがあり、2 番目のパックには優れたタンクがグループに持ち込む遠距離セントリーがあります。適度に大きなスプレッド ショット AOE がその存在に注目を集めます。

さらに 2 つのパック

これもダブル プルできます。ヴァンガード エアロスタットのペアは The Twinning を非常に彷彿とさせ、同じゴミ モデルとドーナツ AOE をペアで共有しているため、それを覚えているプレイヤーにハイテク テーマをアピールするのに役立ちます。 The Twinning と同様に、近接攻撃とキャストアップタイム戦略が完全に異なるため、これはトラッシュで時々使用する楽しい AOE タイプです。

タレット エレクトロビームは、以前のダッシュとは異なり、通り抜ける経路がないため、あまり良くありません。プレイヤーはパーティーの先頭に立つか、待つか、AOE を受けるかのいずれかを行う必要がありますが、最後の 2 つのオプションは楽しくも興味深いものでもありません。順番に発射するか、間にスペースがあれば、楽しいものになります。

第2ボス - プロテクター

https://i.imgur.com/2N7Jlyi.png

サーチ アンド デストロイ

ラウンドごとにすべてのプレイヤー キャラクターの下に AOE を囲んで並べ、Another Mount Rokkon を彷彿とさせる視覚的に密度の高い弾幕効果を作成します。これは、プレイヤーが AOE の配置にすばやく反応し、何か面白いことをするのではなく常に移動することを強いられるため、迷惑な種類のストレスです。(Fulminous Fence との後の組み合わせは役に立ちません。Fulminous Fence は優れたメカニズムですが、根本的な要件は依然として繰り返しの反応であるためです。)

このボスは、キャストの名前とボスがタレットを使用するというアイデアの両方に適合し、テーマに沿ったプレイヤーを誘う AOE をはるかに興味深い方法で持つことができます。このボスは既に加速爆弾を使用しているため、加速爆弾をすべてのプレイヤーキャラクターに（現在警告が出ている状態で）配置することから始め、加速爆弾の有効期限が切れる直前に、すべてのプレイヤーキャラクターの下に 1 つの低速円形 AOE をドロップすることができます。これにより、反応して逃げるのではなく、動かずに AOE に対処してから行動するという計画に固執するテストが作成されます。その後、繰り返しの加速爆弾をめぐるさまざまな単純なおとりを計画する段階に進むことができます。

Fulminous Fence

通過すると麻痺を引き起こす稲妻フェンスを作成します。これは、後の AOE でプレイヤーの通常のパターンから動きを変えるための優れた方法ですが、必要に応じて頻繁に使用することはできません。

Battery Circuit

回転する円錐と定期的な円形 AOE は、稲妻フェンスの存在によって興味深いものになります。これがこのボスで移動を重視する唯一のメカニズムである場合、移動ツールを効率的に使用する場所として計画すると気分が良くなります。しかし現状では、頻繁にキャストするジョブの移動ツールはおそらく Search and Destroy で使用されているだけであり、Search and Destroy の煩わしさが増しています。

モーション センサー

これは Acceleration Bomb の優れた警告で、キャスト名と赤いテキスト「モーション センサー爆弾があなたに取り付けられました」の間にあります。プレイヤーは、そのコンセプトを定期的に確認するために非 MSQ の任務を一度も行ったことがなくても、テキストを読まなくてもデバフを理解することができます。

ブラスト キャノン (タレット追加によるキャスト)

アリーナをカバーする単純な一連のライン AOE ですが、Acceleration Bomb によって複雑になります。非常に公平で、優れたメカニズムです。

ヘビー ブラスト キャノン

この標準ライン スタックはスペーサー メカニズムとして機能し、レイド全体に直接つながるため、ヒーラーは少し作業する必要があります。それほど難しくはありませんが、良いことです。

全体的な印象 (ボス)

なぜパーティー全体の下に AOE をスパムするのでしょうか? 代わりに、Fulminous Fences と Acceleration Bombs の周りで通常とは異なる動きを使用することに重点を置いた場合、これは素晴らしい、本当に楽しいボスになった可能性があります。

2 番目と 3 番目のボスの間の Mob

2 つのパック

ここでは機械的に興味深いものはありませんが、ダブルプルできます。

さらに 2 つのパック

これもダブルプルでき、このペアの最初のパックは、ラインとドーナツ AOE の組み合わせで開始することで、機械的に興味深いものになっています。

3番目のボス - スネークスキナーのザンダー

https://i.imgur.com/m0KEM0H.png

ソウルベイン セイバー

ボスの正面に直線状のAOEが出現し、その後爆発して半部屋の裂け目となり、キャスト中にボスの正面にあったエリア全体を攻撃します。

これを一度見せ、プレイヤーが見たものについて考えることができるタンクバスターを行った後、ボスはアリーナの真ん中で回転しながらこれを繰り返します。これは、現在のキャストの最初のAOEが前のキャストの2番目のAOEとほぼ同時に発動するため、それ自体とのスムーズな組み合わせです。これは、プレイヤーがすべてのピースを最初に見る機会があったため、拡散を追加するために増加した後でも公平な組み合わせです。

このメカニクスは、ダンジョンのボスが死んでからより強く復活するというストーリー上の制約に対処するのに優れた選択であり、ダンジョンが長くなりすぎてプレイヤーが避けようとすることはありません。このメカニクスは自分自身とうまく組み合わさるため、ボスは単純な攻撃だけを行うか、プレイヤーが学習するチャンスを十分に与えないほど速く進むのではなく、非常に短いフェーズ内で組み合わせを示すことができます。コンセプトによっては、後でより多くのメカニクスを使用するために、このような短いフェーズをすべて導入にするのが良い場合があり、他のコンセプトではこの方法の方が適しています。

(キャストバーなし) 復活シーケンス

これは十分に短いので気になりませんが、魅力を求めてダンジョンをグラインドしているときに一部のプレイヤーがこれにイライラしたと思います。

メカニクス的には、これはヒーラーが繰り返しスタックから回復する機会、プレイヤーがジョブに関係なくローテーションで作業できる強制ダウンタイム、および初めてのプレイヤーがヘビの変身からどのようなメカニクスが生まれるかを考える機会として機能します。 (MSQ を未攻略のままプレイしている場合、野生の魂が何であるかをまだ知らないとしても、ボスのモデルが変更され、「私は大いなる貪食者、蛇の魂を呼ぶ」というセリフを見ることができます。)

シンセスリザー

ボスが向きを変えた後、ボスの前にある単純な 90 度の円錐ですが、プレイヤーがシンセスリザーを全速力で捉えるのに役立ちます。

シンセスリザー

ボスの攻撃経路と 4 回の攻撃のための 4 本の剣を示すインジケーターは良いものですが、それだけではこのシーケンスの速度には不十分です。(プレイヤーが AOE の 1 つを直接見ていない場合、表示されるのが非常に短いため、形状を見逃す可能性があります。) プレイヤーに最初にシンセスリザーの円錐を見せることで、プレイヤーが妥当な時間内に繰り返される攻撃に気づく可能性が大幅に高まり、この攻撃は公平性を保ちながら高速で蛇のようになります。

スリザーベイン フォアガード / リアガード

ソウルベイン セイバーのアップグレード。半部屋の斬撃が即座に発動し、キャスト名に応じてボスの前または後ろに配置されます。このアップグレードがスムーズに自身と組み合わさる点も素晴らしいです。ボスのパワーアップを機械的に売り込んでいます。

全体的な印象 (ボス)

これは楽しいボスで、非常に難しいストーリー コンセプトを扱う優れた方法です。ダンジョン ボスは、ボスを最初の死の前と後に分割する前でさえ、時間に追われています!

全体的な印象 (ダンジョン)

サーチ アンド デストロイがなければ、これは素晴らしいダンジョンになるでしょう。ここにある他のすべては非常によくできています。

--------

オリジェニックス（レベル99ダンジョン）

https://i.imgur.com/O1cj7Jo.png

最初のボス前のモブ

2パック

ここではメカニカルに興味深いものはありませんが、ダブルプルできます。

1パック

退屈な強制シングルプルの後、ストーリー上の理由でオープンセクションが続きます。ストーリー上の理由はすべて結構ですが、モブをそこに存在させて一掃する方法を見つけることができたはずです。

最初のボス - ヘルペカリス

https://i.imgur.com/VgqcmAs.png

血管収縮薬

毒液溜まりを作り、ボスがそれを殴ります。

毒液漏れ (何度も繰り返し)

ボスが腕を上げて側面を殴るのは当然です。プレイヤーはこれまでにも同じ腕のアニメーションを持つ敵をたくさん見てきましたし、直感的に「腕を上げた側面を殴る」と理解しているので、彼らが本当に疑問に思うのは、腕のパンチがどこまで伸びるかだけです。

しかし、このメカニズムが実際に機能する方法は、反応的であるだけでなく、プレイヤーの学習と向上への欲求に積極的に反しています。

ここでの最初の問題は、攻撃が実際に当たる場所はランダムに選ばれた水たまりであり、持ち上げられた腕から最も遠いアリーナの端を除くほぼすべてのエリアをカバーできますが、通常はそのエリア全体をカバーしません。したがって、最も基本的なレベルでは、プレイヤーは最初のテレグラフ セットを見て、このキャストが当たる場所を学習しますが、その最初のセットから確実に間違ったことを学びます。スキル レベルが高くなると、AOE が毎回学習できる一定の場所にないため、AOE のすぐ外側をかわす (したがって楽しむ) 能力が否定されます。プレイヤーはアリーナの端から端まで走るか、キャラクターが攻撃されるリスクを冒す以外に手段がありません。

2 番目の問題は、繰り返される Venomspills が交互にではなくランダムであるため、プレイヤーは先制的にキャラクターを動かすことさえできないことです。事前配置を理解し、毎回キャラクターを中央に戻すプレイヤーであっても、それはまだかなり反応的であり、カジュアル プレイヤーがその知識を持っているとは期待できません。

(さらに、前置詞への移動入力を持続すると、プレイヤーの手 (または入力に使用される他の体の部分) が部分的に固定されるため、プレイヤーが両手で移動入力を分割しない限り、反対方向に走るために旋回するのに少し時間がかかります。コントローラーを使用してこのような手の分割が可能かどうかは関係ありません。コントローラーはキャラクターの移動ジョイスティックを分割できないためです。マウスとキーボードでは可能ですが、直感に反します。レガシー移動では、マウスの前方移動を分割で有効にするには、プレイヤーはボスに対してカメラを横向きに向ける必要があります。また、標準移動では、プレイヤーキャラクターを回して「キーボード方向」のためにカメラを向けるという追加の複雑さがあります。プレイヤーキャラクターを回すのは追加の入力であり、レガシー移動分割方法でさえここで期待するのはばかげているため、標準移動分割方法が実際に改善されているとは確信していません。プレイヤーの入力スキームについて想定しているだけでなく、これはレベルアップ MSQ ダンジョンボスであり、苦戦しているプレイヤーは通常それを考えません。)

(私は実際にレガシー移動を使用しています一部の任務ではスプリット方式を採用しています。こちらの方が少し速く、当然レガシー移動の方が好みだからです。しかし、そもそもそれを行うには任務を知る必要があります。これは、このメカニクスの反学習性よりも前から、ここでは完全に文脈から外れています。)

これはボスのコアメカニクスですが、ひどいものです。(スプレッドと組み合わせる前でさえも。) 代わりに、「AOE の端で片側を真ん中にヒットし、毒を飛ばして反対側を一貫してカバーし、固定パターンでもう少しカバーする」など、興味深いものにできたはずです。それは妥当な 2 ステップの回避だったでしょう。

Writhing Riot

許容できるメモリメカニクスですが、特別なものではありません。プレイヤーは最初のセーフスポットに立ってから、その後の両方のヒットを覚えておくか、2 番目のヒットを覚えておき、それを使用して 3 番目のヒットで共有セーフスポットを計算できます。これにより、3 つのアイテム ルールが適用され、強いプレイヤーは 1 回の移動で回避する計画を立てることができます。

全体的な印象 (ボス)

このボスの核となるメカニクスは非常に反学習的であり、他にはほとんど何も含まれていません。

第 1 ボスと第 2 ボスの間のモブ

2 つのパック

ここでは機械的に興味深いものはありませんが、ダブル プルが可能です。

さらに 2 つのパック

これらはダブル プルとして機械的に興味深いもので、エアロスタット (ドーナツ AOE 付き) とタンクがグループに持ち込む遠距離の敵を組み合わせています。

第 2 ボス - デシーバー

https://i.imgur.com/s7lMwpL.png

バイオニック スラッシュ

アームで示されるように、アリーナの互いに反対側の 2 つの象限を攻撃します。それ自体は特別なものではありませんが、ボスのコア コンセプトと組み合わせて後で戻ってきます。

アンドロイドを初期化

アリーナの前方にいる追加敵を探すようプレイヤーに伝える機能的な警告キャスト。

(キャストバーなし) アンドロイドからのライン AOE

一部の追加敵が完全に出現していないため攻撃しないというコンセプトは良いものです。フロア全体を覆うテレグラフは、少なくとも 1 つのセーフスポットを作成するために何かが起こっているに違いないとプレイヤーに促します。

ただし、ボスのダイアログ「アクティブ カモフラージュ操作中」では、完全に出現していない追加敵が攻撃するように見えます。他の追加敵は通常どおりに出現し、実際には「カモフラージュ」されているとは見なされません。プレイヤーは最初のヒット後に学習しますが、テキストに惑わされるのは面倒です。テキストなしで推測するよりも良いことをテキストで伝えることができていたはずです。

「これは完全に出現していません」の視覚効果も、プレイヤーが追加の列を素早くスキャンするのに十分ではありません。 「ちらつき」効果がアクティブでないときに敵が見ていると、不完全に生成された敵が漠然とした緑のオーラを放っているだけで、ブルームが大量にある環境では目立ちません。

その後、敵は殺されるためにやって来ますが、以前のボスと同じ理由で、これは時々あると便利です。

タレットの初期化

もう 1 つの機能的な警告キャスト。今回はアリーナの両側にある大きなブロックに対して、より大きなライン AOE を発射します。私はこれを「タレット」と呼ぶのは好きではありません。これは長い間、小さな敵とライン AOE に関連付けられてきたからです。しかし、これらを呼ぶのにはっきりとした明確な方法はありません。助けを借りずにキャストバーに何かを選ぶ必要がある場合、少なくともプレイヤーをその AOE サイズから遠ざけないように「大砲」またはそれに似た名前にします。

(キャストバーなし) ブロックからのライン AOE

ブロックは片側ずつ発射され、一連の AOE を回避しながらアリーナ全体をカバーします。プレイヤーは、テレグラフがその側から展開して反対側をカバーするため、どちらの側が最初に発射するかを確認できます。

ただし、どのブロックが発射されるかの視覚的な明瞭さはアンドロイドの場合よりも劣ります。アンドロイドはすべてパーティクル効果があり、チャージアニメーションでかなりの輝きがあり、「完全にスポーンされていない」視覚効果を模倣しています。

サージ

ラインノックバックを伴うこれらのスプレッドは、プレイヤーが先ほど見た AOE を利用する (またはブロックを振り返る) のに適しています。今回は完全にスポーンされたブロックにより、キャラクターが壁にノックアウトされるのを防ぐためです。シーケンスの楽しいフォローアップです。

バイオニックスラッシュ + (キャストバーなし) アンドロイドからのライン AOE

しっかりとしたオーバーラップです。今回は「完全にスポーンしていない」アンドロイドの数が少なくなっています。これは、プレイヤーがどのタイプを探すべきかを知っている必要があるため、概念上は問題ありません。

全体的な印象 (ボス)

良い概念で、導入と視覚的な明瞭さが最高ではないことを除けば、うまく実行されています。プレゼンテーションがもっと良ければ、これは良いボスになるでしょう。

2番目と3番目のボスの間のモブ

1つのパック

退屈な強制シングルプル。

2つのセントリー

これも退屈な強制シングルプルですが、セントリーはタンクが引っ張るのではなく、ドアが開くとすぐに飛び出します。これはあまり変わりません。セントリーは、ホルミンスタースイッチの強制プルシリーズのリソースを消耗させるほど強力ではなく、次のパックに走る間、パーティーはダウンタイムがあります。

2 つのアイボーグ

これらには、オーラム ヴェイルのコインカウンターを彷彿とさせる複数のテレグラフ メカニズムがあり、次のモブ / ミニボスとダブル プルできます。一緒にプルするのは楽しいです。敵は少ないですが、通常のプルよりも複雑な AOE があります。

オートマタ

この 1 つのモブからの大きな AOE ハード ストンプはまったく気になりません。プレイヤーは回避する時間が十分にあり、範囲について戦略を立てることができます。また、それが繰り返されるのを見ると、それについて戦略を立てることができます。特に、パーティーは一緒に回避することで、ダブル プルの場合、タンクの回復範囲とアイボーグの近接範囲の両方で恩恵を受けます。

3番目のボス - アンブローズ・ザ・アンデパーテッド

https://i.imgur.com/ha3VyhV.png

圧倒的な突撃

ボスの槍の持ち方は、最初は正面に当たるのか側面に当たるのかはっきりしません。これは驚くべきことではありませんが、プレイヤーは最初の攻撃から正面に当たることを知るので、許容範囲です。

サイコキネシス (ケージ)

赤いテキスト「近くのケージはサイコキネシスによって緊張しています!」が付いており、長い休憩の後でダンジョンを忘れてカットシーンを再生しなかった場合でも、ケージのあるアリーナの側面を見るようにプレイヤーに促すのに適しています。キャストバーも適度に長いです。

ただし、壊れたケージのドアには、もう少し視覚的なポップさが必要です。それらは揺れ、バーは歪み、わずかに光りますが、これはハイブルーム環境であり、プレイヤーはアリーナの反対側の 2 つの側面をチェックする時間を分割する必要があります。(ドアが揺れるのは、周辺視野が比較的動きに敏感であるため、明瞭さのためにコントラストだけに頼らないようにするための良い工夫です。) ケージのドアが当たる床のラインは明確ですが、ドア自体は、最後に激しく揺れるまではわかりません。

(キャストバーなし) ケージ外の追加

追加を倒すのに非常に論理的な場所です。

超感覚フィールド

プレイヤーは、アリーナの外を指していない 1 つの象限に移動するだけで (またはノックバック耐性を使用)、矢印がパルス中にかなり強くフェードアウトしても、その兆候は十分に明確です。プレイヤーは、矢印が数回明るくパルスするのを見るのに十分な時間があるので、適切な場所を見つけるために複数の目の動きをすることができ、キャラクターを動かす時間もあります。

気づいていないとしても、これが発動するのを見ると、2 つの反対の象限がアリーナの側面を向いており、他の 2 つがアリーナの前後を向いているというパターンに気づくチャンスがもう一度あります。(これは良いことです。プレイヤー キャラクターがアリーナの中央線を通り抜けてそこから吹き飛ばされるという自然な推測ではなく、象限をさらに区別するからです。)

サイコキネティック チャージ

圧倒的チャージと超感覚フィールドを組み合わせ、プレイヤーにノックバックを使用して前方の AOE を回避するように求めます。良い公平な組み合わせです。

エレクトロランス

投げられた槍からの単純な円形 AOE で、プレイヤーをボスの方に押しやります。

サイコキネシス (槍ダッシュ)

槍がどのくらい広く当たるかは明確ではなく、床の参照も簡単ではありませんが、ダッシュ シーケンスはボスから離れたところから始まるため、プレイヤー キャラクターは自然に最初の数回のダッシュを避けて大まかなサイズを確認できる位置に配置されます。これは正確な回避を可能にするのに適していないため、1 回目以降は退屈になります。

(ここから、サイコキネシスの 2 つの形式とサイコキネティック チャージが、スプレッドを追加して繰り返されます。これらのメカニズムはいずれもスプレッドとの相性が悪くなく、これは非常にカジュアルなヒーラーが支援なしでクリアできるはずの MSQ ダンジョンで「ボスが勝ちたがっている」という適切なメカニズム表現です。)

全体的な印象 (ボス)

まあまあのボス。視覚的な鮮明さは標準的でしたが、特別な点はありません。

全体的な印象 (ダンジョン)

最初のボスがひどいなど、退屈なセクションが多く、特に特別なものやエキサイティングなものはありません。

--------

Everkeep (Zoraal Ja、レベル 99 ノーマル トライアル)

https://i.imgur.com/EV8NrhF.png

両刃の剣

これは素晴らしいキャスト名で、そのテレグラフが表示される前からプレイヤーに 2 回目のヒットを予想させる合図になります。また、2 回のヒットはアリーナ全体を 1 回だけカバーするため、プレイヤーは成功するにはキャラクターを移動する必要がありますが、キャスト バーに基づくと 1 回目でヒットする可能性があります。

Calamity's Edge (ビジョンからの至近距離の AOE をアクティブ化)

特別なものではありませんが、妥当です。結局のところ、戦場にあるものは完全に無害である可能性は低いです。

Vorpal Trail (疾走する剣)

これに対する私の最初の反応は、興奮した「初心者の最初のエクサスクエア!」でした。 (「エクサスクエア」は、TOP の最終フェーズでアップグレードされたコスモ アローの俗称です。) 剣がアリーナの端を突進して入ってくるのを見ましたが、かっこよかったです。端から剣をスタートさせるのは、この小さなボスのヒット ボックスと組み合わせると良い選択です。開始時にプレイヤー キャラクターはアリーナの端の近くにいない可能性が高いため、プレイヤーはパターンを確認してから操作できます。

これに対する私の 2 番目の反応は、「ああ、床は役に立たない」でした。TOP は、各ヒットに非常に鮮明な視覚効果があり、アルティメットであるため、正方形のグリッドで動作するメカニズムを円形のアリーナに配置することで、それを回避しています。この任務には、メカニズムを機能させるのに必要なそれらのどちらもありません。

これに一度ヒットするのは問題ありません。床の参照がないことが実際に迷惑なのは、これが繰り返し両刃の剣と組み合わされ、プレイヤーがこのメカニズムのタイミング参照も位置参照も持たず、組み合わせに合わせて動きを計画できないことです。 (特に、このトライアルの最初のセクションには、合計で非常に少ないメカニクスがあるため)

これは、トライアルの残りの部分とエクストリーム (床に適切な線がある) の歪んだ世界の剣の前兆となることを意図しているため、床をきらめかせて、その参照グリッドを事前に用意することをお勧めします。これにより、メカニクスの参照が提供され、ゾラール ジャの歪んだ思考が表現されます。そうすれば、メカニクスのイライラが軽減され、組み合わせの興奮が際立ちます。

(カットシーンとフェーズ遷移まで、それ以上の注目すべきメカニクスはありません。)

スマイティング サーキット

赤いテキスト「ゾラール ジャが剣を構える!」は、アリーナを覗いて環境を知らせる良い合図です。ここでの唯一の不満は、大きな剣に選択されたグロー効果が、アリーナの端にある「死の壁」のものと非常に似ているため、環境からそれらを見分けるのが不必要に困難になっていることです。 (プレイヤーはボスのスタンスを利用できません。Zoraal Ja は至近距離やドーナツで剣を構えておらず、グロー効果で剣の形が歪んでいるからです。) エクストリームでは、プレイヤーは義務の知識と練習でそれを克服する方法を学びますが、ノーマルでは役に立ちません。

グロー効果は、Zoraal Ja が魂を燃やしていることをテーマ的に示しているため、代わりにデスウォールの効果を変更します。インスタンスの背景のようにエレクトロペの紫色に光り、壁のコントラストをシャープにするために明るい部分が立ち上がるようにすると、デスウォールのシャープさを維持しながら剣を視覚的にクリアにすることができます。

Dawn of an Age

これは、アリーナを縮小して変形させる非常に優れた方法で、「歪んだ世界」のミニフェーズを可能にします。 AOE 自体は十分にシンプルで、表示される時間も十分長いため、プレイヤーはほぼ確実にそれをうまく回避でき、アリーナの覆われた部分を消すことができます。素晴らしい作品です。

Vollok

最初に利用可能なプラットフォームに直接剣を置くのは良いことです。このキャストが繰り返されるときに、プレイヤーが剣を探すために遠くを見るように準備します。最初のキャストでは、剣がグリッドに当たったこともプレイヤーに示します。

Sync

「歪んだ世界が同期する」という赤いテキストと組み合わせた、素敵な長いキャスト。これは、どの追加のプラットフォームがアクティブ化され、キャストの終了直後に AOE を実行するかを探すようにプレイヤーに非常に公平に求めます。

Extreme で重複キャストを可能にするために遅延アクティブ化が必要ない場合は、一貫性と明確さのために、タイミングを維持し、キャストバーを長くして、AOE がキャストの最後にヒットするようにします。

Bitter Reaping

ダブル タンクバスターは、Normal ではそれほど強力に攻撃できませんが、カジュアル プレイヤーがコンセプトに慣れておくために、1 つあると便利です。

ゲートウェイ

プラットフォーム間でダッシュソードを移動させるテザーを設定します。このセットアップの各ピースに個別のキャストを与えると、プレイヤーがすべてのピースを出現順に確認するのに役立ちます。

ブレードワープ

ダッシュソードを配置する別のセットアップキャスト。これも良いもので、比較的複雑なメカニズムをノーマルでうまく機能させることができます。

フォージドトラック

キャスト、アクティブ化されるソードへのテザー、および赤いテキスト「ブレードは次元を駆け抜けます。」の組み合わせにより、これは本当にうまく機能します。ダッシュソードはプレイヤーキャラクターから非常に離れているためテザーが必要ですが、以前に設置されたソードには必要ありません。（実際、設置されたソードは数が多すぎて、乱雑にならずにテザーを張ることができません。）

アクチュアライズ

アリーナを元に戻し、ミニフェーズ間で戦闘を切り替えられるようにします。

ハーフ フル

このクリーブは、ボスの上げられた腕を剣までたどることができるため、初めてのプレイヤーにとってはスマイティング サーキットよりもかなり簡単に見つけることができます。これにより、クリーブは、後で「歪んだ世界」アリーナの青い背景の色合いで機能します。

ハーフ サーキット

論理的にはハーフ フルとスマイティング サーキットを組み合わせたもので、変更されたデス ウォールと背景と非常に合理的です。

デューティズ エッジ

大きなマルチヒット ライン スタックは恐ろしいですが、ヒットが遅いため、ノーマルでも妥当です。

全体的な印象

最初のセクション (カットシーンの前) は、全体的な繰り返しと、ヴォーパル トレイルの位置合わせのフラストレーションの間で、このトライアルの重荷になっています。カットシーンの後はすべて、視覚的な磨きをかければ、健全で妥当なノーマルになります。斜め向きのボスと仕事をするのは、近寄りがたい感じがせずに、いい気分転換になります。

--------

アレクサンドリア (レベル 100 ダンジョン)

https://i.imgur.com/PWXpsV4.png

残念ながら、MSQ のキャラクター要件に対する道徳的意見の相違により、多くのプレイヤーがここにいたくないという状況は、必要な背景です。エンドゲームまで残っている場合は、エキスパート ルーレットでこれをもう一度見る可能性がありますが、初回の経験は依然として重要です。

最初のボスの前のモブ

1 パック

退屈な強制シングル プル。これはレベル アップ ダンジョンでは十分に迷惑でしたが、エキスパート ルーレットに表示されるエンドゲーム ダンジョンでは明らかにひどいものです。

プレイヤーがアレクサンドリアの環境を非常に交互に移動することを望んでいるのであれば、問題ありません。シングル プルを機械的に興味深いものにしてください。複数の AOE や、環境と相互作用するミニボスなどを用意してください。

2 つのパック

ここでは機械的に興味深いものはありませんが、2 回プルできます。

免疫プロトコル B9

これは文字通り 1 つのモブで、何も面白いことをしないロボット象であり、パーティーには他の敵と一緒に戦うオプションはありません。(最後のパックを終えるとすぐに出現しますが、ダメージがほとんどない場合はリソースの枯渇にはなりません。) つまらない!

最初のボス - アンチウイルス X

https://i.imgur.com/Wuyhtms.png

免疫応答

テレグラフと光る前面ビーム/側面触手で、ボスの前面または側面を切り裂きます。それ自体はシンプルですが、後でドーナツとクロス AOE のシーケンスと組み合わせると興味深いものになります。

(キャストバーなし) 5 つの AOE シーケンス

このメカニズムには、赤いテキスト「外部エンティティの除去が開始されました。」、雷が流れるとメカニズムを開始する 5 つの紫色のドットの上に 1 つずつ緑色で表示される円とプラス、およびそれらがオフになるとテレグラフとオーディオ キューの間で、何が起こっているかを示す優れた手がかりがあります。ペースも良く、プレイヤーは動きを計画する時間があります。

このメカニズムを本当に優れたものにするには、各 AOE がオフになるまでの時間を視覚的に表示します。(円の上を落下するオブジェクトは塔のように見えすぎますが、2 番目の色で円とプラスを 1 つずつ埋めるなど、そのようなことができます。) これにより、プレイヤーは、各 AOE がオフになったときに部分的に反応するのではなく、キューとグローバル クールダウンに合わせて移動のタイミングを計ることができます (スナップショットを徹底的に練習して習得するまでは)。特に重複する免疫反応のキャストにより、このメカニズムはスキル表現と満足のいく解決に多くの余地がある可能性がありますが、それを可能にするインジケーターがありません。

これはボスのコアコンセプトであり、頻繁に繰り返され、その視覚的な変更により、まともなものから素晴らしいものになります。

隔離

AOE タンクバスターと他の全員のスタック。難しくはありませんが、タンクをスタックから遠ざけるようにプッシュする点で異なります。

全体的な印象 (ボス)

優れたボスになり得る優れたボスには、シーケンスのタイミングの視覚的な表現が必要であり、これにより習熟が可能になります。このビデオをチェックしていたとき、ビデオのパーティーはこのボスで約 3 パーティーの KO を獲得しましたが、間違いを理解し、毎回改善が見られるため、そのまま進み続けました。これが失敗を奨励する様子です。

最初のボスと2番目のボスの間のMob

2つのパック

ここでは機械的に興味深いものはありませんが、2回プルできます。

もう1つのパック

次のパックが出現するまでの短いダウンタイムを伴う、退屈な強制シングルプル。

もう1つのパック

それでも面白くなく、パックが2つだけのシリーズでは、ホルミンスタースイッチのようにリソース枯渇を引き起こすには長さが足りません。特にエキスパートルーレットでは、アイテムレベルの同期がより厳格であるにもかかわらず、プレイヤーが装備を整えると、最低レベルより30高いアイテムレベルで登場します。

さらに悪いことに、このプルの後、パーティーは2番目のボスに到達するまで、木と石の瓦礫の山で長い時間待機します。そこのセリフはストーリーに関連していますが、繰り返しクリアすると配置が退屈になり、クールダウンが戻ってくる時間を与えることでリソースの枯渇が打ち消されます。これはエキスパート ルーレットなので、退屈でイライラすることが多いです。

代わりに、標準的なダブル プルの機会を提供し、最後のパックが交戦しているときにストーリーテリング シーケンスを開始し、そのパックとの戦い中にパーティがプレイします。最後のモブが死んだ直後に瓦礫が撃ち落とされ、パーティは待つことなく先に進むことができます。プレイヤーが Dawntrail MSQ のためにゲームに戻ってきたとしても、前の 5 つのレベル アップ ダンジョンで標準的なダンジョン プルに慣れ親しんでいるため、シーケンスの既存の音声付きセリフにより、プルから目をそらすことなくそれを理解できます。

2 番目のボス - アマルガム

https://i.imgur.com/c2qcAQk.png

分解

ボスは名前の通り、多くの部分に分裂し、ゆっくりとアリーナの半分を囲む三角形を作り、その三角形を AOE で攻撃します。この戦闘全体は紫と紫で、最後の「これには別のインジケーターがありました」というテレグラフまで、危険ゾーンの端だけがマークされていますが、それでもこれは機能します。リードタイムが長く、「アマルガムが部分に分割されます。」という赤いテキストがあり、ボスのモデルが劇的に変化します。

集中流 / 分割流

紫色の輝きが増しながらキャストし、それぞれボスの前後のラインとボスの側面を攻撃します。ベテラン プレイヤーは、この組み合わせを Hot Tail / Hot Wing として認識するかもしれません。

このメカニズムは問題ありませんが、その功績は主にキャストバーにあります。光っている領域は、最初は AOE が明るいセクションを超えてヒットするという手がかりを与えません。また、それぞれの最終的な前後の線は、遅れていることを除けば確実に見るには細すぎます。キャストバーに頼っておそらく一度ヒットするよりも、ビジュアルからこのパターンを推測するチャンスをプレイヤーに与えたい場合は、ビジュアルをヒットする領域にもっと結び付ける必要があります。

(キャストバーなし) エレクトロペ ビーム

円形の大砲が落下するスペースはアリーナの前面に 5 つあり、そのうち 3 つは最初の反復で埋められ、4 つはその後の反復で埋められます。これらは、分解と組み合わせた単純なライン AOE を実行します。

この組み合わせは問題ありませんが、大砲がアリーナの前面にあること (したがって、すでに画面に表示されている可能性が高い)、分解のリード タイム、および赤いテキスト「エレクトロペにエネルギーが蓄積されます!」に大きく依存します。実行可能ではありません。大砲から伸びる細い線は、それだけではプレイヤーの目を引くのに適していません。

これらの後に、コンビネーションが発動する直前にスプレッドが表示され、プレイヤーがコンビネーションのセーフスポットから広がるように促します。タイミングは許容範囲ですが、スカイディープ セノーテの Firearms の同様のセットアップよりも、プレイヤーが調整ループに巻き込まれる可能性が大幅に高くなります。

Ternary Charge

簡単だが楽しい。Ihuykatumu の Prime Punutiy の Shore Shaker と同じです。これは、同じ同心円 AOE のシーケンスです。

全体的な印象 (ボス)

まあまあのボス。分割して再結合するボスというテーマの可能性にもかかわらず、特別なものはありません。視覚的にはかろうじて合格です。これは、イライラする前に実行可能な最悪のビジュアル デザインです。

2番目と3番目のボスの間のMob

4つのパック

ここでは機械的に興味深いものはありませんが、2回引くことができます。

3番目のボス - エリミネーター

https://i.imgur.com/5KaDS0S.png

パーティション

再構成パーティションの準備となるシンプルなサイドクリーブ。

再構成パーティション

キャスト名は、剣が移動して反対側に当たる重要な手がかりであり、アニメーションへの反応ではなく、公平なメカニズムになります。

サブルーチン

ボスは1つまたは2つのパーツを分割し、それらのパーツが後で個別のAOEを実行できるようにします。（クローはラインAOEを実行し、オーブはドーナツAOEを実行します。）理論的には、これは前のボスであるアマルガムを基にして接続する楽しい方法です。

しかし、ボスの背中のパーツの初期位置は非常に高い位置にあるため、プレイヤーキャラクターがデフォルトのカメラの傾きでボスに近づいたとき、またはそれより少し高い位置にあるときは、パーツが適切に見えません。そのため、これは機能的にはボスが何かを設定したという警告に過ぎません。ストーリーではこの任務はコンピューター内で行われるため、キャスト名「サブルーチン」はほぼ何でも設定でき、パーツが別々に表示されていない場合、プレイヤーが得るのはそれだけです。

(このレビューのためにデューティ サポートと一緒にこのダンジョンに入るまで、このボスが分裂することを私はまったく知りませんでした。Krile と G'raha Tia のコメントでそのことが分かりました。友人とデューティ ファインダーのマッチメイキングで初めて挑戦したとき、カメラの傾き、カメラの位置、デューティを学習しようとして目を素早く動かしたことが重なって、パーツが分離しているのが見えませんでした。その後のクリアでは、パーツ メカニクスの潜在的な位置を予測し、スクリプトで指定された時間にそこを見るようになりました。つまり、パーツが分離しているのを見るためにボスを見ていなかったのです。)

プレイヤーが戦闘中にカメラの傾きを調整することを期待するのは現実的ではありません (APM は別として、プレイヤーにはキーバインドの柔軟性がない可能性があります)。また、カメラの傾きはプレイヤーの HUD レイアウトとキャラクターの周囲を見るために重要なので、このボス モデルの取り外し可能なパーツを下に移動し、取り外すとボスのシルエットが大幅に変わるようにすることで、この問題を修正します。そうすれば、初めてプレイするプレイヤーは、デューティ サポートの解説なしで分離に気付く可能性がはるかに高くなり、ボスの統一された配色も変更されません。

(エリミネーション クローによるキャスト) ターミネート

ボスの別のメカニクスと同時にキャストされるワイド ライン AOE。唯一のインジケーターは、敵対リストのキャストと、アリーナの横にあるクローから途中まで伸びる紫色のラインです。

プレイヤーが、クローがボスから離れ、前のサブルーチンからアリーナの横に移動するのを確実に見た場合、このメカニクスは問題ありませんが、組み合わせを処理するには高速です。これは信頼性が高くないため、紫色の輝きは、これが最初に表示されたときに再構成パーティションからの近くの他の紫色と組み合わせると、プレイヤーの目を確実に引き付けるのに十分ではありません。また、キャストが長くないため、プレイヤーが自分に何が当たったのか見えないという厄介な「ゴッチャ」の可能性が生じます。 (特にカジュアル プレイヤーにとって、初めての任務では見るものがたくさんあることを覚えておいてください。)

(Elimbit がキャスト) 破壊の輪

ボスの別のメカニズムと同時にキャストされるドーナツ。インジケーターは敵対リストへのキャスト、アリーナの安全でないエリアの暗転、安全ゾーンの「スポットライト」を締め付ける紫色の縦線です。ドーナツの後にはスプレッドが続き、後にはプレイヤーにキャラクターをドーナツの安全ゾーンから移動させるよう求めるターミネートが続きます。

繰り返しますが、プレイヤーが確実にオーブが外れたのを見た場合は問題ありません。これは確実ではないため、インジケーターは多少不足していますが、ターミネートほどひどくはありません。少なくともアリーナ全体に何かが表示されます。

(サブルーチンが両方のパーツを切り離すことによって開始) 追加フェーズ

6 つの分散型 Lightning Generator は、さまざまなジョブがさまざまな方法で複数のターゲットから順番に恩恵を受けることができるため、興味深い可能性があります。ただし、追加フェーズが非常に短くなければならないダンジョン ボスのコンテキストでは、どの追加を最初にターゲットにするかというパーティーの調整が不足しているため、Duty Finder のマッチメイキングでこれらの追加からのスキル表現が制限されます。

(エリムビットが発動) 救済の光 (ライン プロティアン)

これらはテレグラフと共に登場し、プレイヤーが反応して拡散するために必要なものです。

ここでは、プロティアンの最初の表示は問題ありませんが、良くはありません。エリムビットはドーナツの後、これが最初に現れる前に移動しているため、移動と追加の間に、プレイヤーはエリムビットを見ることができず、テレグラフにのみ反応する可能性があります。(Ktisis Hyperborea で Hermes が True Aero を発動した場合と比較してください。これもライン プロティアンであり、さらに餌付けされたフォローアップがありますが、プレイヤーが確実に追跡しているボスから直接発生します。)

追加フェーズの後、これはノックバック サークル (キャストバーなし) の後にシーケンスされ、ノックバックが解決する直前にプロティアン テレグラフが表示され、ノックバックとプロティアン解決の間に数秒の広がりが残ります。このシーケンス自体は合理的ですが、仲間のプレイヤーの選択に反応する必要があるため、ボスがスプレッド中にキャストを開始するパーティション/再構成パーティションからプレイヤーの目が必然的に逸らされます。

この最後の部分により、初めてのプレイヤーは簡単に個別の KO を受ける可能性があります。キャラクターは、壁にぶつかるのを避けたとしても、ノックバックとプロティアンからダメージを受けたばかりです。回避可能なダメージが来ていること、ましてやそれが何であるかを確認する機会が与えられません。また、繰り返し完了した場合でも、視覚的な要求は、Duty Finder が反応的にしか実行できないメカニズムには大きすぎます。パーティション/再構成パーティションではキャストバーと剣を見る必要があり、プロティアンで仲間のプレイヤーの選択に効果的に反応するには、キャラクターの足元を常に見ている必要があります。その結果、最も実用的な方法は、プレイヤーがパーティション/再構成パーティションに反応するための事前配置として、プロテアンをボスの前後のラインに近づけ、プロテアンが解決したら反応して解決することです。これにより、シーケンス内の反応がさらに押し進められ、面白くなくなります。

プレイヤーが視覚を使用して仲間の選択に反応するように促すメカニズムは、反応時間に対する固有のプレッシャーにより、関連する場所を見続ける必要があるため、その期間中、視覚がプレイヤーの目を完全に占有するという前提で設計する必要があります。周囲のメカニズムも、その基準に基づいて選択および設計する必要があります。

排除

ボスはアリーナを数回切り裂き、そのたびに軽いダメージを与え、地面を切り裂く一連の線を残します。これらの線は同じシーケンスで拡大して爆発し、パターンが非常に密集しているため、さらなるダメージを回避するには複数の動きが必要になります。

それぞれの AOE セットがどこに当たるかがビジュアルでもっとはっきりしていれば、あまり当たらない場所を探してそれに応じて回避する計画を立てるのが楽しくなるでしょう。しかし、ビジュアルは明瞭さよりも輝きやフラッシュに重点が置かれており、床の参照もないので、見ていて不快で、分析するのも面倒です。

全体的な印象 (ボス)

ボスのパーツが分離して再びくっつくのをプレイヤーが確実に見ることができれば、初回体験はずっと良くなるでしょう。繰り返しクリアすればなんとかなりますが、それでもかなり面倒です。

全体的な印象 (ダンジョン)

遅くて退屈で、最初のボスのコンセプトが光りません。遅さに関して言えば、Expert Roulette の他の部分と比べてアイテム レベルの同期がきつくなっていることには少なくとも理由があります。シングル プルと強制待機には理由がありません。

--------

インターフォス（スフェーン、レベル100の通常トライアル）

https://i.imgur.com/pyYpedR.png

アレクサンドリアと同様に、残念ながら、多くのプレイヤーが自分のキャラクターに対するMSQの要件に道徳的な同意がないため、ここに来たくないという状況は避けられません。これはストーリートライアルであるため、デューティサポートがないにもかかわらず、プレイヤーが自分のMSQを完了した後に戻る理由はほとんどなく、統計的には彼らが自分のMSQのためにここにいる可能性が非常に高いです。

一般的な視覚的な注意点として、クイーンエターナルの戦闘ダイアログポップアップのNPCイラストは少し高すぎて、ターゲットバーのデフォルトのHUDレイアウト位置にあるターゲットバフとデバフが見えにくくなっています。最もカジュアルなプレイヤーでさえ、自分の適用されたデバフを追跡するためにターゲットバーを動かすことを期待するのではなく、NPCのイラストを下に移動するか、またはもっと良いことに、これらのポップアップの位置をHUDレイアウトで調整できるようにします。

正当な力

優れたワンツーパンチシーケンス。ボスの4本の腕のビジュアルにより、シーケンスの両方の部分を表示しながら同じ側を2回ヒットできます。十分に明確で、最初の輝きは自然にプレイヤーの注意を引きます。

エーテルティース

最初のレイド全体の多数のヒットは問題ありません。

その後の水たまりのAOEと各セクションの最後のヒットは、そのビジュアルによって完全に沈んでいます。赤いテキスト「エーテルの吸収が加速します！」は警告であり、問​​題ではありません。問題は、危険ゾーンの歪んだグリッドビジュアルです。私の場合はなんとかできますが、各セクションの最後のヒットで捕らえられなくても、引き込みメカニズムであるように見えます。しかし、一部のプレイヤーは乗り物酔いがひどくて嘔吐しましたが、その理由は理解できます。床が歪んだパターンで動いているように見えると、視覚的なバランスのフィードバックがかなり乱れる可能性があるからです。

理論的には、高コントラストのオブジェクトの動きを利用することは、周辺視野の動きの感度とコントラストの要件に合わせて、光るまたは標準的なオレンジ色のテレグラフに頼ることなく確実に見える視覚的な手がかりを提供する方法です。実際には、この結果から、このような手法は床を歪めるために使用すべきではないことが明らかになりました。他の方法で実験したい場合は、ライブゲームにそのビジュアルを導入する前に、さまざまな乗り物酔いに悩むプレイヤーでテストを実行してください。

戴冠式 (円形 AOE)

これらのターゲット AOE が選択されたプレイヤーキャラクターに向かって飛び跳ねる方法は不気味で役に立ちません。AOE でのプレイヤーのリードが減り、ターゲットが誰であるかをプレイヤーが把握するのに役立たないことが多いためです。 (パーティは以前のメカニクスから積み上げられている可能性が高いです。)

これに、他のバージョンのように Prey ターゲット マーカーが付属していれば、適切なリード タイムを備えた妥当なメカニクスになります。

バーチャル シフト

アリーナを別の環境に変化させます。

砂時計アリーナ ミニフェーズ

非常に制限のあるアリーナですが、楽しくて取り組みやすいチャレンジができます。

戴冠式 (ダブルターゲット ライン AOE)

これは本当にクールなメカニクスです。2 人のターゲットが協力してテザーで大砲を引っ張ることができるので、誰もビームに当たらずに済みます。また、テレグラフと大砲の反応を長時間見ることができます。鮮明なビジュアルも優れており、キャラクターがテザーされているかどうかに関係なく、プレイヤーはメカニクスのピースを見ることができます。

デューティ ファインダーは連携していませんが、このメカニクスはパッチ 7.1 で登場するエクストリーム バージョンのために存在しているようですが、リード タイムが長く、攻撃を受けるのに 1 つの脆弱性アップしかかからないため、デューティ ファインダーでも機能します。プレイヤーは反応時間にプレッシャーを感じることなく、それを解決できる十分な時間があり、失敗してもレイズを待つことなく戦い続けることができます。

プレイヤーが他のプレイヤーの選択に反応しなければならない複雑なメカニズムが必要な場合は、これがその方法です。このようなメカニズムは本質的に通常の任務に限定されていますが、これがその方法です。

（このメカニズムは、Aethertithe と組み合わせて「メインアリーナ」ミニフェーズの後のラウンドにも登場します。）

ダウンバースト

アリーナの角または中央からのシンプルで馴染みのあるノックバックですが、アリーナの形状が制限されているため複雑になっています。難しくはありませんが、プレイヤーが初めてこの複雑さについて十分に理解するのに十分な時間があり、楽しいです。

（キャストバーなし）ドローンサークルからのドーナツ AOE

特別なものではありませんが、優れたメカニズムです。ドローンが渦を巻いて円を形成し、危険地帯の端を照らすという視覚的な手がかりが、このゲームを成功に導いています。

パワフル ガスト

もう 1 つのシンプルでおなじみのノックバックですが、今回はアリーナの片側から発生します。ダウンバーストとまったく同じデザインの結果になります。

浮遊ミニフェーズ

プレイヤー キャラクターを動かして下に床がないことを確認することでこのフェーズを開始する方法は重要なトリックであり、プレイヤーはデバフ バーを読まなくても浮遊に気付くようになります。結局のところ、これは通常モードです。強制的な移動は理想的ではありません (特に近接範囲外の場合)。ただし、アリーナの変更中に行うと、それほど不快ではありません。

キャステレーション

プレイヤーに高さの変化ごとに十分な警告を与えることが、このゲームを成功に導いています。空中浮遊を示すオープニングと、「重力が増して落下します」と「重力が減って浮遊します」という変化を示す赤いテキストは重要です。

また、壁が城の一部のように見えるようにしながら、視覚的な明瞭さに配慮していることも称賛に値します。穴の縁が光るだけでなく、穴を通して安全地帯に光が差し込み、プレイヤーが安全地帯に並ぶのを助けます。

このメカニズムに失敗すると一撃でKOになる（壁がプレイヤーキャラクターをアリーナから押し流す）ので、関係のない壁の上部のギザギザや輝きを少なくすることで、視覚的な明瞭さをさらに高めます。また、安全地帯をもう少し広くすることも検討します。これはMSQでは必須であり、非常にカジュアルなプレイヤーは必要な動きの精度を持っていない可能性があるためです。 (特に、2 回の連続した Castellations のより速いシーケンスの場合 - 下から上への遷移により、プレイヤー キャラクターがすでにセーフスポットに近づいていることが保証されますが、最初の移動試行でそこに着地する移動精度に欠けるプレイヤーにとっては役に立ちません。)

Legitimate Force (浮遊フォローアップ付き)

同じメカニズムですが、今回は、キャラクターをアリーナの中央近くに維持すると、プレイヤーは差し迫った落下に失敗することになるでしょう。

プレイヤーは既に浮遊メカニズムを操作したことがあるため、これは合理的であり、2 回目の腕の輝きの直後に赤いテキストを表示するタイミングは完璧です。プレイヤーは既に Legitimate Force を何度か見ており、基本的な形式はすでによくわかっているため、赤いテキストは、2 回目の輝きキューの処理を終えた直後にほぼ正確に表示され、ヒットが交互に発生する場合、プレイヤーがスムーズに「プラットフォームに移動する」ことを移動計画に追加するように促します。

落下のタイミングによって一瞬の躊躇が生まれますが、これはカジュアル プレイヤーが躊躇せずに動作シーケンスを実行する練習をしていないため、必要なことです。

全体として、これは動作シーケンス メカニズムを使用してプレイヤーを成功に導く方法の良い例です。

メインアリーナ再訪

絶対的な権威

このキャストバーの長さ、ドラマチックなセリフ、5秒間のカウントダウンは、通常、非常に高いダメージのメカニズムと関連付けられており、プレイヤーはこれをシールドしたり、リミットブレイクをタンクしたりする可能性が高くなります...キャストはすぐに中程度のダメージしか与えず、代わりに機械的に激しいシーケンスのためにデバフを適用するため、これはほとんど役に立ちません。

デューティファインダーは、これらすべてに対処する前と後に一息つく必要があります。フレア、誰かと一緒に立つ三角形、加速爆弾、プレイヤーキャラクターに次々と石化の視線が付けられ、地面に円形のAOEがラウンドされ、シーケンス全体でダメージのパルスが繰り返されます。ただし、このプレゼンテーションはイライラするほど誤解を招きます。パーティーは回復と軽減ツールを必要としますが、メカニズムの開始には必要ではないからです。

このメカニズムを改善するには、プレゼンテーションを2つのキャストバーに分割します。最初のキャストバーは、すべてのデバフの「非アクティブ形式」を適用し、フレアの「凍結」マーカーを薄くして、カジュアル プレイヤーであっても後で何かが起こることを準備する必要があることを示します。2 番目のキャストバーは、プレイヤーに考える時間を与えるために依然として長く、最後にすべてのデバフを「アクティブ」にしてシーケンスを開始します。このように、プレイヤーがデバフ バーを読んでいなくても、大規模なレイド全体を予想するのではなく、後の回復を含むシーケンスの準備に長いキャストを費やすことになります。

(はい、私の提案は、TOP のコード スメルの根底にある「このデバフはまだ有効ではありません」という同じ概念を Duty Finder に穏やかに導入することですが、期限切れのタイマーではなく、後で有効になるキャストを使用します。これはすべてのスキル レベルに適した概念であり、Ultimate に限定する必要はありません。)

プレイヤーが実際にこのメカニズムが何であるかを知ったら、それは合理的です。デバフは頭上のマーカーで適切にタイミング調整され、プレイヤーはキャラクターの足元を見るだけで円形の AOE を回避できます。シーケンス中に息継ぎの時間がないため、ペースが恐ろしく感じますが、それを行うのに十分な時間があります。

分割して征服

Duty Finder にライン プロテアンを実行させる素晴らしい方法です。素晴らしいメカニズム、優れた作業です。

Alexandria とは異なり、ここではプレイヤーは一般にプロテアンのソースを見ています。その後、ヒットが 1 つずつパーティを分散させます。特に、ヒットごとに鮮明なテレグラフと「何かのために選択」の頭上マーカーがあるため、パーティ メンバー全員に同時に最後のヒットを打つためにかなり分散する可能性が高いため、最終的なプロティアン ポジションを管理できる可能性が高くなります。(調整ループに巻き込まれて失敗する可能性はありますが、プレイヤーが計画ではなく反応に基づいて任務に近づいている場合、それについてできることは限られています。) キャスト名も適切ですが、それだけでは十分ではありません。

これに似たものをもっとお願いします。 デューティ ファインダーの人口を再訓練して、プロティアン ポジションを含む一般的な準備をしない限り、このようなメカニズムから「ガイド レール」を削除することは決してできませんが、それでも多くの概念を穏やかに導入することはできます。 そうすれば、通常の任務がハイエンドの任務へのより良い踏み石になり、ハイエンドの任務の達成度が下がることはありません。

最終フェーズ

ここでの Wuk Lamat の介入は、ストーリーテリングと、パーティの努力がほとんど無意味になるほどのダメージを与えることの両方で、プレイヤーからスポットライトを奪うため、やる気をそぐものです。プレイヤーの努力は重要ではないと伝えたら、どうしてプレイヤーが努力することを期待できるでしょうか? さらに悪いことに、カットシーンは 2 番目の絶対権限解決を急に中断し (パーティがそれを可能にする十分なボス HP しきい値をまだ満たしていない限り)、プレイヤーが任務の最も恐ろしい機械的なシーケンスを実行するのを本能的に妨げます! 全員が準備万端で動いているのに、突然 Wuk Lamat によって制御が奪われ、ステージを譲らざるを得なくなります。

任務への NPC 介入はうまく行うことができますが、プレイヤーの行動の重要性を奪ってはなりません。たとえば、The Final Day (Endsinger Normal) にも NPC 介入カットシーンがありますが、光の戦士の成功を祈るサイオンによって、プレイヤーの行動の重要性が下がるのではなく、高まります。これは拡張パックの MSQ のクライマックスに非常にふさわしいものですが、これはそうではありません。同様に、The Final Day の NPC 介入は、以前はタンクの限界突破 3 が必要だったレイド全体の到来を予期し、限界突破が利用できないことに気付いたため、プレイヤーが自分のキャラクターを制御できないと予想しているときに行われます。(結果として、The Final Day でのサイオンの祈りの動機付け効果は、The Unending Coil of Bahamut (Ultimate) (UCoB) でのフェニックスの介入の効果に似ています。)

友情の力に関連する NPC 介入があることは、まったく問題ありません。スフェーンの記憶を復元しようとすることに基づいていてもかまいません。代わりに、2 回目の絶対権限が完了した後、クイーン エターナルがオーバーロードを続け、カットシーンが始まったと仮定します。このバージョンでは、ウク ラマットはグラハ ティアとクリルを伴ってスフェーンの記憶を復元しようとし、その成功によりスフェーンはアレクサンドリアから 1 人以上のボスを召喚します。NPC はそれらのボスと交戦し、プレイヤー パーティーはスフェーンと戦い続けることができます。3 人の NPC では軽量パーティーを組むのに十分ではないため、プレイヤーは最終フェーズ中に NPC のリソースが自身の戦いで徐々に枯渇していくのを目にすることになり、テーマに沿った緊迫感と、場合によっては (寛大な) ハード エンゲージさえも生み出します。

ダイナスティック ダイアデム

小さなセーフスポットを備えた基本的なドーナツで、プレイヤーが自分のキャラクターをアリーナの中央前部に移動するように促します。これは Royal Banishment の事前配置には役立ちますが、そこにある円錐の共有ポイントはここの円の中心と一致しないため、あまり良くありません。

Royal Banishment

レイド全体でのスパムと、適度に遅い交互の円錐シーケンス。シンプルな動きと避けられないダメージの小さなヒットが多数あり、Sphene の過負荷状態を伝える優れたメカニズムです。

全体的な印象

このトライアルは、最終フェーズとストーリーのコンテキストによって本当に台無しになっています。他のフェーズの実際のメカニズムは、いくつかのメカニズムを少し磨けば、非常に良いものから素晴らしいものになるでしょう。

----------------

多くの強力なコンセプトといくつかの本当に優れたメカニズムにもかかわらず、すべてのグループ MSQ デューティには少なくとも 1 つの大きな設計上の欠陥があり、そのうちのいくつかはビジュアル デザインにあります。これは、すぐに楽しむためにも、プレイヤーが楽しめるという自信を持ってエンドゲームにたどり着くためにも、良くありません。

私はアルティメット レイダーですが、初めて新しい MSQ 任務をこなすときはいつも不安になります。いつも不安になります。MSQ の続行がほぼ完全にブロックされたことがあり、新しい任務は私を止めてしまうものになる可能性があり、またはゲームが私が上達できないものを要求する方向に進んでいることを私に知らせるものになる可能性があります。パッチ 7.0 の MSQ では、その点で多くの懸念が残りました。グループ任務のどれもプレイしていて気持ちが良かったとは思えなかったからです。

As with previous parts of this letter, for your convenience I have attempted to provide a Japanese version using Google Translate, since I unfortunately do not speak Japanese myself. Please pardon any mistakes so generated. In case of any confusion, please refer to the English text if at all possible, as I cannot verify the accuracy of the machine translation.

この手紙のこれまでの部分と同様に、残念ながら私自身は日本語が話せないため、皆様の便宜を図るため、Google 翻訳を使用して日本語版を提供するよう努めました。間違いがあった場合はご容赦ください。機械翻訳の正確さは確認できないため、混乱が生じた場合は、可能な限り英語のテキストを参照してください。

--------------------------------

パッチ 7.0 の任務の設計のレビューを続け、このサブセクションでは、MSQ で必須ではない 2 つのオプション ダンジョンについて説明します。概念的には、AAC ライト ヘビーウェイト ノーマルもこのサブセクションに含めますが、両方のダンジョンに詳細な分析に値するメカニズムがあるため、長さを考慮して分割しました。

エキスパート ルーレット ダンジョンであるため、これらの任務は、ストーリーを完了した後、プレイヤーが何度も戻ってくるように促されることで恩恵を受けることができますが、繰り返しを維持しながら、カジュアル プレイヤーに励みになる初回体験を提供する必要があります。結局のところ、プレイヤーが任務に戻るのを躊躇すると、繰り返しベースのゲームプレイ ループの報酬は苛立たしいものになります。

これらはエンドゲームの任務であり、戻ってきたプレイヤーは、後のパッチで戻ってきたとしても、最初に MSQ でウォームアップすることが多いため (最新の MSQ がこれらの任務に必須ではないパッチ MSQ である場合)、これらの任務は、最も弱いプレイヤーをあまり気にする必要がありません。しかし、エンドゲームで活躍するプレイヤーのほとんどは、まだカジュアルプレイヤーです。

(Duty Support にはこれらのダンジョンが含まれていないため、他のプレイヤーに迷惑をかけないように、アリーナの写真はエクスプローラー モードで撮影する必要がありました。アリーナにボスがいないのはご容赦ください。)

----------------

テンダーバレー（レベル 100 ダンジョン）

https://i.imgur.com/Sy1OWH5.png

最初のボス前のモブ

4 つのパック

これらはダブル プルできるだけでなく、ダブル プルごとに 1 ～ 2 体のより危険なモブが HP が高くなっています。これはレベル アップ ダンジョンと同じようなスキル表現の機会であり、エキスパート ルーレットの激しい繰り返しとフル ジョブ ツールキットにより、さらに歓迎されています。

最初のボス - バレルテンダー

https://i.imgur.com/NXfubzu.png

ヘビーウェイト ニードル

これはボスから発生する通常の円錐 AOE のセットですが、最初のテレグラフは実際の AOE よりも狭く、テレグラフは最後の瞬間に実際にヒットする領域を示すためにのみ拡大されます。実際にセーフスポットがあるのは大きなオープン エリアのみで、より多数の小さなオープン エリアは拡大されて 2 回ヒットします。

その結果、初めてプレイするプレイヤーはテレグラフに従ってキャラクターを小さなオープン エリアまたは大きなオープン エリアの端に移動させる可能性が高く、その結果メカニズムに失敗します。しかし、次回何をすべきかわかっていても、励まされるのではなく、誘惑されたように感じます。

なぜでしょうか?初期のテレグラフが実際の AOE と一致していないことを示す他の兆候はなく、拡大するテレグラフもプレイヤーに有利に使用されていないためです。(Sigmascape V4.0 (O8N) とその Savage バージョン (O8S) で Kefka と戦うのは楽しいです。なぜなら、彼がトリックスターであることを示す兆候があり、テレグラフが実際の AOE と一致していないからです。プレイヤーがストーリーで彼がトリックスターのボスであると知らされるからだけではありません。Dispatch (Vanguard Commander R8) の円形バージョンの拡大するテレグラフは、早い段階で拡大し始め、プレイヤーを真の安全地帯に導くように注意深く配置されているため、許容できます。) ここで習得すべきスキルはなく、プレイヤーは通常のダンジョン ボスの進行プロセスを実行することを期待していません (または進んで実行していません)。これは「落とし穴」であり、プレイヤーは、100 レベルのゲームプレイでテレグラフ システムを信頼するように教えられているにもかかわらず、テレグラフ システムがここにあることを知っていて、覚えていなければなりません。 (テレグラフはエコーによって付与されることをプレイヤーが覚えていればなおさらです!)

これはボスであり、ゲームを拡張して、ゲームがプレイヤーに教えたことをそのまま実行したために罰を与えるものです。これは決して楽しい体験ではありません。むしろ、プレイヤーを劣ったものとして扱っています。

したがって、実際の AOE の形状からは遠いものの、Black Cat Crossing (AAC ライトヘビーウェイト M1 (M1N)) の光る線のようなものを針として使用すれば、これはより優れたメカニズムになります。少なくとも、実際の AOE はピクセル線ではないことは直感的に明らかなので、プレイヤーはそれがどのくらい広がり、どのような形状になるかを発見することを期待します。

このメカニズムの仕組みをすでに知っているプレイヤーはどうでしょうか?

「広い空き地を見つける」だけでは、プレイヤーが隠されたパターンを知らなくても面白くありません。 (2 つの可能性しかありません。セーフスポットは北と南東、または北西と南のいずれかです。) カジュアル プレイヤーにとっても、これは 90 回目には退屈であり、他のメカニズムと組み合わせる前に、この戦闘の開始時に 2 回キャストされます。

その退屈さを回避するには、プレイヤーにスキル表現の機会を提供する必要があります。このメカニズムでは、床の形状に AOE がどこに広がるかをプレイヤーが計算するための参照があれば、必要な範囲だけ回避することでスキル表現を許可できます。しかし、床にはそのような参照線がないため、その機会はプレイヤーに与えられず、メカニズムは退屈なものになります。

伝承のコンセプトが「定住」エリアではなくオープンな砂漠地帯だからといって、スキル表現を容易にするための優れた床の形状がないわけではありません。このボスは、この攻撃を含むいくつかの攻撃の一部としてアリーナの中央にジャンプしたり、中央を攻撃したりしますが、床にはすでに乾燥した砂漠の環境を伝える亀裂があります。ボスが繰り返しその場所にぶつかる衝撃によってアリーナの中央から大きな亀裂が放射状に広がるようにするのは、まったく理にかなったことだったでしょう。結局のところ、戦いはボスの家で行われるのですから。そして、それを使ってセーフスポットの可能性のある端をマークするのです。

現状では、この仕組みは楽しいところがまったくありません。まず不公平で、次に退屈です。

テンダードロップ

多くのサボテンを落とします。次のキャスト（ヘビーウェイトニードルまたはバレルブレーカー）の解決後すぐに爆発します。サボテンには2種類あります。小さな爆発のものは裸で、中程度の爆発のものは薄茶色のトゲとピンクの花があります。

2種類のサボテンを区別する手がかりは不十分でコントラストが低いとだけ言っておきます。トゲは小さく、ピンクの花はオレンジ色の床に簡単に溶け込んでしまうからです。しかし、プレイヤーがこれらすべてのサボテンを見回すときに簡単に目にするものをお見せしたいと思います。

以下は、サードパーティツールを使わないゲームプレイ動画からのフリーズフレームのスクリーンショットです:

https://i.imgur.com/1UxQQL5.jpeg

https://i.imgur.com/gaAbRwM.jpeg

https://i.imgur.com/frPYkiB.png

動画は完全には読み込まれていませんが、さまざまな視覚効果によってサボテンがぼやけているため、プレイヤーのインターフェースがサボテンよりもはるかに鮮明であることがわかります。最初のスクリーンショットでは、ぼやけはボス自体からのみ発生しており、サボテンが爆発する直前に解決される後続のキャストのアニメーションの一部です。 2 枚目のスクリーンショットでは、ぼかしは視点プレイヤー自身のジョブ アクションから来ています。これは、ジョブ ファンタジーやゲームプレイ フィードバックの重要なソースとしてオンのままにされることがよくあります (プレイヤーがそれらを制限またはオフに設定するオプションについて知っている場合でも)。3 枚目のスクリーンショットでは、両方のぼかしソースが同時に存在していますが、その後のキャストは異なります。

これは視覚的な明瞭さにとってすでに悪いです。しかし、プレイヤーが見るものはこれよりもさらに悪いです。特に、花とトゲが異なる爆発サイズを示していることをプレイヤーがまだ知らない場合はそうです。

プレイヤーがサボテンを見渡す時間枠を考えてみましょう。ボスがサボテンを空中に投げたばかりなので、テンダー ドロップのキャストの最後にすぐに見ることはできません。サボテンはトゲを覆い隠すほこりの雲の中に着地し、花に届くので、プレイヤーはその時間を使ってサボテンの間を素早く見ることも事実上できません。プレイヤーは、騒ぎが収まってからトゲや花のチェックを開始し、そのキャストのためにどこに移動するかを決める前に、すべてのサボテンと次のキャストに目を通す貴重な数秒しか残っていません。

その結果、初めてプレイするプレイヤーは、すべてのサボテンを次のキャストと一致させるのに十分な精度で見つけようとしたり、次のキャストを待ってから、その潜在的な安全地帯の近くのサボテンをチェックしたりして、目を素早く動かさざるを得なくなります。(「試みる」というのは、サボテンが多すぎて、プレイヤーが利用可能な時間内にすべてのサボテンの正確な位置を特定するのは不可能だからです。人間の目には限界があります。) この急速な目の動きは、プレイヤーが目が動いている間の視覚情報を使用しているか、目の動きを減らすために完全に周辺視野を介して一部のサボテンを見つけているか、またはその両方であるかにかかわらず、効果的にさらなるぼやけを生み出します。 (障害によっては、プレイヤーの目の筋肉が健常者よりもゆっくりと動くこともあります。)

スクリーンショットをさらにぼかして、各部分をほんの一瞬だけ見ると... サボテンの花やトゲはもはや判別できません。プレイヤーが実際に見るのは、美しくレンダリングされたサボテンの「ぼやけた塊」と表現できるかもしれません。

そのため、プレイヤーは、メカニクスが解決される前にサボテンを見ても、花やトゲが存在し、関連していることに気付かないかもしれません。自分で気付くには、初期のテレグラフが横たわっているという「特徴」を核とするボスとの戦いの最中に、花やトゲのあるサボテンとトゲのないサボテンを、中心視野で十分長く捉えて、それらの要素に気付く必要があり、もう一方のタイプがどのようなものか思い出せるほどの間隔で連続して捉える必要があります。

メカニクスが初めて解決したときに爆発を見たらどうでしょうか。その後、プレイヤーは、初期のテレグラフがどれだけ広がっているか、また、キャラクターがすべてのサボテンから十分離れているかどうかを確認するために、キャラクターの足元を見るように促されます。キャラクターをサボテンの上に立たせない限り、それは、どちらのタイプのサボテンも見るには最悪の場所です。ましてや、両方を見るのは無理です。 (そして、爆発の大きさの違いは、中心視野の小さな部分であれ、周辺視野のオレンジ色の海の中の円全体であれ、円の曲率を見ても簡単には見えないほど大きい。) そのため、プレイヤーが花とトゲの存在とその時点での関連性に気付く可能性はさらに低くなります。

その結果、プレイヤーは 2 種類のサボテンが存在することに気付かずにメカニクス全体を簡単に通過し、安全地帯を見つけるのに十分な速さと正確さでサボテンを見つけられなかったために失敗したという印象を受けることになります。(または、同じ理由で偶然成功したという印象を受けることもあります。) 今度は、この印象によって、次回はサボテンを見つけることにさらに集中するようになります。これは不快なだけでなく、サボテンの種類の違いに気付く可能性が非常に低くなる、まさに急速な眼球運動戦略にさらに踏み込むことになります。

かなりの数のプレイヤーにとって、このメカニクスの視覚的デザインは、学習に費やした努力を無駄にして、行き止まりに導いてしまいます。プレイヤーが行き詰まってしまったら、抜け出す方法は仲間のプレイヤーに教えてもらうか、メカニックを上達させるためにビデオ映像をゆっくり見るかのどちらかしかありません。特にカジュアル プレイヤーにとっては、どちらもあり得ません。彼らは通常、そのメカニックは不公平だとか、自分にはできないと結論付けるだけです。特に、すでに不公平なことをする意思を示している「ガッチャ ボス」の状況ではそうです。

さらに悪いことに、障害のあるプレイヤーは、障害の種類に関係なく、このサブセットに入る可能性が不釣り合いに高くなります。視覚障害のあるプレイヤーは、明らかな理由で苦しみます。しかし、視覚障害のないプレイヤーは、障害を補うために視覚にさらに圧力をかけ、このダンジョンのロックを解除するために必要な MSQ から十分な練習を積んでいます。これにより、彼らは急速な眼球運動戦略に駆り立てられます。特に、彼らのメカニカル語彙には過去のメカニクスの長い歴史があり、多くの爆発源を正確かつ迅速に特定することが必須の方法であるためです。

(私はまさにこの行き止まりに陥っていましたが、友人から花と棘の存在を聞きました。すると別の友人が「花と棘って何?」と叫びました。その会話には3人しかおらず、傍観者もおらず、全員が少なくとも最終フロアのサベージをクリアできるレイダーでした...そして私以外の2人はグラフィックアップデートによる遅延がなかったため、花と棘が画面に何度も表示されていました。)

これはまったく受け入れられません。プレイヤーが行き止まりに陥らないようにするのがあなたの仕事であり、それを奨励するものではありません!

2 種類のサボテンについて、そしてセーフスポットが常に花やトゲのないサボテンの近くにあることを知っているプレイヤーの経験はどうでしょうか?

バレル ブレーカーのシーケンスについて考えてみましょう。この知識を持つプレイヤーは、キャラクターがノックバックに乗ってコーナーに入り、それに応じてセーフスポットのあるコーナーを探すように計画できます。したがって、各コーナーでは、そのコーナーに最も近いサボテンだけを自然に見ます。また、隠されたパターン (反対側のコーナーに 2 つのセーフスポットがある) により、中央視野で最大 2 つのサボテンを見た後に 1 つ見つけられることが保証されます。これは、与えられた時間に対して妥当なタスクであり、したがって妥当なシーケンスです。

(ここでは、隠されたパターンがプレイヤーの助けになることに注意してください。なぜなら、プレイヤーは、その恩恵を受けるためにその存在を知る必要がないからです。隠されたパターンについて知っているプレイヤーは、このシーケンスのセーフスポットを見つけるために 1 つのサボテンを見るだけでよく、それはカウントされません。)

しかし、ヘビーウェイト ニードルのシーケンスについて考えてみましょう。 2 つのメカニズムの間では、必ずしも共通の安全地帯があるわけではないので、プレイヤーはアリーナのほとんどの場所でサボテンをチェックして、小さな爆発を起こすものを探し、それに応じて回避する計画を立てなければならないかもしれません。また、たとえ常に共通の安全地帯があったとしても、プレイヤーは、初期のテレグラフが表示されたらヘビーウェイト ニードルの安全地帯の近くにあるサボテンをすべてチェックするか、初期のテレグラフが表示される前にサボテンをすべてチェックし始めて、いくらかの労力を無駄にする必要があります。(どちらの場合も、チェックするサボテンのほとんどには花とトゲがあるため、検索は停止しません。) 各サボテンを中央視野でチェックして爆発の大きさを見極める必要がある場合、これはまったく合理的ではありません。

あなたの担当デザイン チームは 3D アート チームでもストーリー ライティング チームでもないことは理解していますが、各ビジュアル インジケーターの必要な明瞭さの基準がチーム間で伝達され、リリース前に徹底的にテストされていることを確認する必要があります。 これは、O11N および O11S の Starboard / Larboard Wave Cannon から学んだ「このキャスト名はメカニクス キューです。派手な翻訳はスピードの明瞭さを犠牲にしてはいけません」というマークのビジュアル デザインに相当します。そうしないと、各バトルで周囲のストーリーを表現するために費やした作業が、ここでのようにゲームプレイの問題を引き起こす可能性があります。

このメカニズムを救済するには、ゲームによって画面に生じるぼやけや、プレイヤーの目の動きによる実際のぼやけにもかかわらず、2 種類のサボテンを視覚的に区別する必要があります。バトルの伝承コンセプトを犠牲にして高コントラストと輝きだけに頼るのではなく、サボテンを徹底的に区別し、プレイヤーが確認する時間を増やす方法でメカニズムを提示します。

ヘビーウェイト ニードルの床の形状を変更すると言ったことを覚えていますか? そこから作業を進めていくと、あらゆる位置に対応する小さな緑のサボテンのつぼみが常に存在し、このキャスト (適切に名前を変更) が、選択したサボテンをつぼみの形から成長させてから、爆発的にトゲを放出し、再びつぼみの形に戻ります。これにより、サボテンが着地して塵が落ち着くまでの期間が自然になくなり、プレイヤーは現在サボテンを効果的に確認できません。より微妙なことに、これにより、可能な位置のセットは限られているが大きく、メカニクスの合間の「空き時間」や戦闘の外でセットを確認できるようになります。

これにより、異なるタイプのサボテンを区別するトゲもなくなりますが、トゲは戦闘の最中にはほとんど見えないため、それほど大きな損失にはなりません。代わりに、小さな爆発のサボテンは花も枝もなく細く高く成長し、中程度の爆発のサボテンは成長するために取り込んだ水エーテルで輝く青い花を咲かせて球状に成長するようにします。完全に異なる形状と中程度の輝きにより、初めてプレイするプレイヤーはぼやけた画面越しに実際にある程度の視認性を確保できるだけでなく、背の高いサボテンの爆発が小さいことにプレイヤーが気づけば、周辺視野で背の高いサボテンを探し、中心視野で背の高いサボテン 1 本とその周囲の正確な位置を把握することで、今後の繰り返しに対処できます。(これにより、特に Heavyweight Needles シーケンスでは、現在可能なよりもはるかに速く視覚検索を停止できます。これが、中程度の爆発のサボテンを単純に大きくすることを提案しなかった理由です。それらはプレイヤーが最初に見つける必要のあるものではありません。) そのため、学習に費やす努力により、プレイヤーは行き止まりから遠ざかり、必要な視覚情報を見つけるための合理的な方法に向かうことになります。

バレル ブレイカー

ボスはアリーナの中央にパンチを繰り出し、ノックバック耐性を使用しない限り、すべてのプレイヤー キャラクターをノックバックします。それ自体は単純です (コーナーを狙う) が、常にテンダー ドロップの直後に行われます。

トゲトゲ左 / トゲトゲ右

初期のテレグラフとボスの光る側面は、これをその側面への 180 度のヒットとして示していますが、AOE の実際のサイズについてはここでも嘘をついています。しかし、ここでヘビーウェイト ニードルの「ゴチャ」を適用し、そのために必要な距離だけ回避しようとする初めてのプレイヤーであっても、このメカニズムは AOE がさらに拡大するため、キャラクターにヒットするように誘います。 (唯一のセーフスポットは、示された側と完全に反対側にあります。)

プレイヤーがこの仕組みの仕組みを理解すれば、ボスのターゲットリングの側面にある向き合った矢印を使用して、最適化のためにセーフスポットを正確に見つけることができます。そのため、これはヘビーウェイトニードルほど悪くはありません。しかし、初期のテレグラフが他の手がかりなしに嘘をつくのは依然として迷惑であり、この仕組みは一度知ってしまえばそれほど面白くありません。

全体的な印象 (ボス)

このボスのコアコンセプトは、最初は不公平であり、ビジュアルデザインが貧弱であるため、繰り返し完了しても挽回できません。ビジュアルがクリーンアップされたとしても、面白くありません。

最初のボスと2番目のボスの間のモブ

4つのパック

これらはダブルプルでき、ダブルプルごとに1つの大きなモブが再び現れます。各ダブルプルの前には水道管スライドがあり、遭遇を分けると同時に、かつて定住していたエリアに到達するストーリーを示しています。

2 番目のボス - アンスラサイト

https://i.imgur.com/c78bVc6.png

アンスラボム (基本形態)

アリーナの床に 3 つの爆弾を投げ、黒い矢印の弧が爆弾が着地する場所の画像につながります。(その後、爆弾はそれぞれ 1 つの至近距離 AOE で爆発します。) 弧は視覚的なインジケーターの優れた重要な部分であり、プレイヤーはボスから弧をたどることですべての爆弾を見つけることができます。キャストの長さも適切です。

アントラボム (ライン AOE チュートリアル)

アリーナの床にあるパイプの開口部に爆弾を 1 つ投げ込みます。黄色の矢印の弧がパイプの開口部の上にある黄色の矢じり (爆弾の画像はありません) につながります。(アリーナの片側にある対応するパイプは、プレイヤー キャラクターが当たったものをノックバックするライン AOE を実行します。2 種類のパイプは、アリーナを建設した以前の文明によって装飾として使用された色、形状、パターンによって区別されます。) キャストの長さも適切であり、このキャストだけを表示することで、この AOE が爆弾が着地した場所に直接当たらないことをプレイヤーに気付かせることが重要です。

ライン AOE の端は、プレイヤーが床の形状、具体的にはアリーナを横切る各サイド パイプの端に接する小さな石のわずかに赤くなった線によって理解できるはずです。しかし、色の違いと石の間の隙間は非常に小さいため、ジョブアクション、ボスアニメーション、プレイヤーの目の動き、周辺視野での作業、プレイヤーのカメラの動きなど、効果的なぼかしが導入されるとすぐにこれらの線が見えなくなります。また、ぼかしがかかっていない場合でも、地面のパイプによって中断されます。地面のパイプはアリーナの入り口付近のスペースの多くを占めており、そのためこの時点では初めてプレイするプレイヤーのカメラの近くにある可能性が最も高いです。

この戦いは以前の文明の遺跡で行われることは理解していますが、視覚的な明瞭さのために適切に鮮明な床またはそれらを必要としないメカニズムのいずれかを実現する方法を見つける必要があります。（そして、それが可能であることはわかっています。Alexander - The Cuff of the Son（A6N）、Alexander - The Burden of the Son（A8N）、およびそれらのSavageバージョン（A6SおよびA8S）の床の配色は制限されていますが、Swindler's Heightデバフの隆起したパネルと下げられたパネルは非常に明確です。）

たとえば、以前の文明では、色あせたペンキではなく、さまざまな石の自然で長持ちする色を使用して床やパイプを着色することができました。(おそらくボスはほこりを吹き飛ばすでしょう。) そうすれば、Shaaloani のベースとなっている米国西部で容易に入手できる硬い建築用石材の中でも、はるかに制約の少ないパレットが得られます。(建築文明の論理的な装飾パターンの選択肢は、石から直接構築されたと見なせる比較的大きくてブロック状のオプションに制限されますが、現在の装飾パターンは確実には見えません。必要に応じて、石を彫刻することで、伝承目的のより細かい詳細を提供できます。) 建築文明に、透明で耐久性のある素材で柔らかい石を保存する能力を与える場合は、これらのライン AOE の端を装飾用の黒曜石でマークすることもできます。これは非常にコントラストが高く、地理的なエリアで入手できます。

対照的な色の正当化の理由が何であれ、線状 AOE の端を示す床の形状は、床パイプの上に広がる必要があります。現状では、床パイプが線状 AOE の端の 1 つによってそれぞれ二分されていることにプレイヤーが気付いたとしても、赤い円のパイプでは、あまり意味がありません。

アントラボム (最終形態)

最初の 2 つの形態を組み合わせた 3 つの爆弾を投げます。2 つはアリーナの床に着地し、1 つはパイプに入ります。床とパイプの視覚的なデザインが適切であれば、完全に合理的です。

カーニフラグレーション

最終形態のアントラボムを 3 つ連続して投げ、その後に基本的なスプレッドが続き、間にキャストバーはありません。各アントラボムに与えられた時間はそれほど短くありませんが、2 つ目と 3 つ目は、前の爆弾の配置と爆発中に表示されます。スプレッド マーカーは、最後の爆弾セットが投げられるときに表示されます。

ここで、床のパイプとライン AOE の端を高コントラストでマークすることがさらに重要になります。なぜなら、プレイヤーが前のセットをかわすためにキャラクターを動かすと、カメラの動きによって、2 番目と 3 番目のセットの爆弾のインジケーターがいくつかの点でぼやけるからです。

しかし、このメカニズムには別の問題があります。

この種のメカニズムでは、プレイヤーは長年の経験から、安全地帯を 1 つ見つけるだけでよく、アリーナ内のすべての安全地帯をマップするよりも速いことを知っています。そのため、特にプレイヤーが遠距離で最大限のダメージを与えることができるジョブをプレイすることに慣れている場合、ペースが速くなったことに対する自然な対応は、まずキャラクターをすべての爆弾から遠ざけることです (そのためには周辺視野から大まかな場所を知るだけで十分です)。次に、パイプに当たらない最も近い場所にキャラクターをさらに移動します。 (この 2 回目の移動は、プレイヤーがライン AOE チュートリアル Anthrabomb で学んだパイプの交互配置によって小さくなることが保証されており、プレイヤーは最初の移動中に黄色の弧を見つけ始めることができます。) この種の他のメカニズムでは、これは特定のメカニズムの視覚的インジケータの練習不足や、単に高速要求に対処するための優れた方法です。

ただし、ここで適用されたこの方法は、通常、プレイヤーがキャラクターをパイプの横のサイド ウォールの 1 つに近づけるように促します。 (ライン AOE のエッジをマークする際のコントラストが不足しているため、プレイヤーはサイド パイプ自体を参照として使用するように促されます。) その時点で、プレイヤーのカメラの動作を考慮してください。カメラをサイド パイプの上に配置しようとすると、カメラはサイド パイプの上の見えない壁にぶつかり、キャラクターに押し付けられ、視界がほとんどなくなります。カメラをサイド パイプから離して配置すると、キャラクターから最大距離であっても、その後の爆弾セットでアリーナ全体を表示することはできません。 (プレイヤーが主にマウスで移動する場合、後者を選択する傾向があります。)

その結果、プレイヤーはボス自体ではなくカメラと戦わなければなりません。これは、インターフェースと戦うことで没入感が失われることと、インターフェースのせいでプレイヤーの事前学習が失敗することの両方で煩わしいものです。プレイヤーは結局、このボスで回避するこの方法を避けなければなりませんが、それは自分のキャラクターにとって効果的ではないからではなく、後の爆弾が見えなくなるからです。

もし横のパイプに見えない壁がなければ、プレイヤーはそこからアリーナにカメラを向けることができ、ほとんどのアリーナと同じようにカメラを使えるようになり、問題は解消されるでしょう。最初の攻撃の後の攻撃が、キャストバーで回避できるような、アリーナ全体をはっきりと見る必要のない別の形式であれば、最初の回避後にプレイヤーが利用できる視界で十分でしょう。この仕組みが、Anabaseios: The Twelfth Circle (Savage) (P12S) の Gaiaochos ミニフェーズのように、アリーナ全体を見るためにズームアウトしたカメラをプレイヤーに効果的に付与するコンテキストで提示されれば、この問題も解消されるでしょう。 (レイダーたちは、これらのミニフェーズを口語的に愛情を込めて「UAV」と呼んでいます。これは「無人航空機」の略で、完全にデザイナーが意図した独特の感覚ですが、TOP ワールド レースで使用されるカメラ ズームハックのように機能します。) しかし、ボスが立っているこの場所では、これらのオプションはどれも当てはまりません。

プレイヤーが粘り強く、混乱していると感じるメカニクスでも隠されたパターンを探し出すのに十分な強さがあれば、状況は最終的に改善されます。 爆弾の可能性のあるセットの多くは実際にはブロックされており、常に近くにセーフスポットがある側壁から離れた場所がいくつか作成されます。 プレイヤーがキャラクターを側壁から遠ざけることを学ぶと、繰り返し行うことで、アリーナの視界が得られ、これらの場所の 1 つまたは複数に気付くことができます。 逆に、これらの場所の 1 つに引き寄せられた場合、そのエリア内に留まるために必要な動きが小さくなるため、メカニクスの混乱が軽減され、最終的に、メカニクス中に現在のフロアの低コントラストのマークを確実に参照できるようになります。ジョブの稼働時間に合わせて動きを最適化することさえできるかもしれません。しかし、そこまでの道のりは不必要に気が滅入るものであり、苦戦しているプレイヤーはそれをまったく追えないかもしれません。

全体的な印象 (ボス)

このボスは、そのビジュアル デザインとゲームの標準カメラの不幸な犠牲者です。表面下には最適化の可能性がありますが、インターフェイスとビジュアル デザインとの戦いを避けるためには、隠れたパターンを見つける必要があります。

2 番目と 3 番目のボスの間のモブ

2 つのパック

これらは 2 回引き出すことができ、2 番目のパックには、自動攻撃でターゲットの周囲にマークされていない円を切り裂く 2 つの Yok Huy 船が含まれています。これはメカニズム的に興味深いものです。さらに、レベル 100 のすべてのヒーラー ジョブには、この量のダメージを簡単に処理できるさまざまな AOE ヒーリング ツールがあるため、パーティーがより良いダメージ出力のために自動攻撃を選択する場合でも、カジュアル ヒーラーにとって過度の負担にはなりません。

確かに、これは良いことだと私は考えています。スキルの低いヒーラー プレイヤーが、雑魚のプルでパーティー ヒーリング クールダウンをいくつか使用し、次のボスでそれらのクールダウンを再び使用できることに気付くように促すためです。これは、タンクだけがダメージを受けている場合でも、雑魚のプルで AOE ヒーリング クールダウンを効果的に使用できることを、彼らに穏やかに示唆します。

この 2 つのパックの後に、矢印スライドがあります:

https://i.imgur.com/iqKtz4L.png

この矢印スライドは、次のボスが Greatest Labyrinth に配置する矢印をプレイヤーに紹介することを明らかに意図しています。ただし、完全にまっすぐなので、プレイヤーは前のセクションの水道管スライドとまったく同じように操作します。これにより、初めてのプレイヤーは矢印の動作を見逃しやすくなります。特に、パーティー メンバーがすでにダンジョンに慣れていて、突進している場合はそうです。(パーティー全体が初めてそこにいる場合でも、宝箱を見ているか、前のゴミの引き出しのことを考えている可能性があります。)

これを見逃しにくい導入にするために、矢印スライドに物理的なパイプと同じ時間、または最大で物理的なパイプにかかる時間をかけますが、廊下全体を回転させずにいくつかのターンを含めます。(乗り物酔いを避けるために、正確な設定を調整する必要があるかもしれません。) この変更により、プレーヤーが矢印をすでに理解している必要も、繰り返し完了にイライラする遅延を追加することなく、以前の物理的なパイプとは異なる動作であることが明確になります。

1パック + ミニボス

スライド後の2パック目は、このアリーナに立つヨク・フイの証人1体です:

https://i.imgur.com/yYyAmZH.png

Shadowbringersシリーズで多くのジョブをレベルアップしたベテランプレイヤーは、このアリーナをキタナ・ラヴェルから認識し、レベルアップ方法に応じて同様のメカニズムを期待するでしょう。そして、それは正しいでしょう。ヨク・フイの証人のメカニズムは、そこのロンカン・ドリーマーのミニボスのものとほぼ同じですが、唯一の違いは、ここでは、ランダムなプレイヤーキャラクターにドロップされたサントス地上AOEが、像がエンシェント・レイスからラインAOEを発射するとすぐにターゲットになることです。そのようなプレイヤーは、すでに像の仕組みを知っており、そのアニメーションが始まる前に像の回避を確認することで、このタイミングを管理できます。

しかし、断続的にしかサブスクライブしないプレイヤー、または比較的新しいプレイヤーで、多くのジョブをレベルアップするために MSQ に追いつく (または追いつくのにほぼ追いつく) 期間が長くないプレイヤーは、MSQ の 2 つの拡張版前の 1 つのダンジョン完了時に Ronkan Dreamers を見たことがあるだけかもしれません。彼らはミニボスや彫像を覚えている可能性は低く、そのため、その参照なしで Yok Huy Attestant がどのように機能するかを学習する必要があります。したがって、彼らは Ancient Wrath の推測を確認するために彫像の AOE のアニメーションを見る必要があり、そのために Sun Toss テレグラフから目を離します。彼らはおそらくオレンジ色のテレグラフに気付くでしょうが、それを見るのに時間がかかるため、彫像の AOE アニメーションを見るのをやめるまで、彼らにとっては反応的になります。

その結果、どちらのメカニズムも一度学習すれば反応的ではなくなりますが、プレイヤーの 2 つのメカニズムの学習プロセスは互いに干渉します。ベテラン プレイヤーでも、レベル アップ ダンジョンを頻繁に繰り返さない場合、または各ジョブがアクセスできる最高レベルのダンジョンに直接キューイングすることで各拡張ごとに各ジョブのレベルを上限まで上げ、2 つの拡張で Qitana Ravel を見たことがない場合、この状況になることがあります。または、これは別のダンジョンであり、このミニボスは Ronkan デザインとの類似性を示す最初の強いヒントであるため、単に理解を再確認したいだけの場合です。

ハイエンドのデューティ デザインでは、このような干渉は、プレイヤーが急速な機械的なシーケンスを通じて注意を管理することを強いることによる、許容できる副作用であることがよくあります。プレイヤーは進行のプロセスを期待し、実行する意思があるため、現在の攻撃を習得するためにシーケンス内の次の攻撃でキャラクターを攻撃できるため、許容されます。しかし、これは通常のダンジョンであり、カジュアル プレイヤーにはそのような忍耐力はありません。プレイヤーは、Sun Toss が追加される前に Ancient Wrath を 2 回以上使用して、以前のメカニズムを十分に理解し、その解決方法を見なくても問題ないと思うようになるまで、簡単にそれ以上の努力が必要になるでしょう。そして、その結果として生じる悪い第一印象で、ダンジョンのこのセクションを判断するでしょう。(特に、パッチ 7.0 MSQ のメカニズムで悪い経験をしてここにたどり着いた場合や、Barreltender の「ガッチャ」メカニズムの後で特に注意を払っている場合)

プレイヤーがすでに Ancient Wrath をよく知っている場合でも、Sun Toss を気にせずに近接範囲/視線内に留まるアップタイム ドッジを実際に見ることができるのは、少し満足感があります。プレイヤーにそのような瞬間を与えてください。

このセクションを Qitana Ravel とまったく同じにすることなく強く思い起こさせるには、Sun Toss を古いミニボスのスムーズなタイミングで別のメカニズムに完全に置き換えます。置き換えの仕組みは、The Qitana Ravel よりも少し難しいかもしれません。これは、このダンジョンのプレイヤーはより経験豊富であり、このダンジョンは MSQ に厳密に必須ではなく、Expert Roulette ダンジョンは Leveling Roulette ダンジョンよりも繰り返し行われるためです。たとえば、Yok Huy Attestant は、Lozatl (The Qitana Ravel の最初のボス) と同じ 3D モデル構造を持ち、各肩のクリスタルが光ってアリーナのその側を攻撃したことを示します。これらの側面の裂け目は、それ自体は単純ですが、破壊された壁の部分と組み合わせることでプレイヤーに新しい考慮事項を提示し、モデルと彫像のコンテキストで認識できる Ronkan の仕組みであるため、ここではうまく適合します。

1 つの残念なタイミングの選択を除けば、私はこのセクションが好きです。これは、ダンジョン後のカットシーンで詳しく説明する前に、仕組みを通じてストーリーをほのめかす良い方法であり、前のパックをアリーナに引き込むことができるのは Expert Roulette では歓迎されます。 (Qitana Ravel の 2 番目の Ronkan Dreamer は、追加のパックとより厳密なアイテム レベルの同期により、より強力なプルを提供します。)

3番目のボス - トゥラルの最大の蛇

https://i.imgur.com/7UzBGhe.png

弾む評議会

このボスの最も頻繁に繰り返されるメカニズムは、小さな蛇のセットを生成し、それぞれが地上のAOEを1つ（同時に）実行してから去ることです。

(最初のキャストは、細い前後の線のみ、2 番目のキャストは、1 つのコーナーが開いた至近距離の円のみ、その後のキャストはすべて 2 つのタイプを混合します。これらの後続のセットには、隣接するコーナーを占める 2 つの円の蛇と、残りのコーナーとアリーナの中央を占める 3 つの線の蛇が常に存在します。線の蛇は、南北、東西、または真斜めにしか向けることができません。)

これらの蛇には、跳ね返る方向を示す明るい白青の矢印が付いているのが気に入っています。これは、蛇が巻き上げられた後に大きく跳ね返って AOE を作成するという伝承のアイデアを不明瞭にせず、グロー効果にも頼らない、鮮明でコントラストの高いキューです。これは、各 AOE がどのような形と方向をとるかをプレイヤーが自分で判断するために必要なものを正確に提供します。素晴らしい作品です。個人的には、矢印がすぐに表示されることを望んでいますが、非常によく構成されています。

残念ながら、これらの AOE の配置は床タイルに対して適切ではありません。この床は、Yok Huy の好意により幾何学模様が刻まれた正方形のタイルで構成されており、各コーナーの蛇はタイルの中央に正確に配置されていますが、線 AOE の幅はタイルの辺の長さよりわずかに大きいため、タイルのグリッドに沿っている AOE であっても、プレイヤーは正確な参照ポイントを見つけることができません。(また、これを示すテレグラフは 1 秒未満しか表示されないため、プレイヤーはタイルの幅とまったく同じであると簡単に誤解する可能性があります。その後、タイルの端を誤って参照ポイントとして使用してキャラクターがヒットした場合、真の原因をまったく知らないまま、ヒットをスナップショットまたはラグのせいにする可能性が高いです。) 対角線と円形の AOE には、床に参照ポイントに似たものがないため、誤解を招くことはありませんが、プレイヤーが正確に回避する機会は依然としてありません。そのような参照ポイントは、可能性のある端を彫刻パターンに組み込むことで作成できます。

これらの AOE と床が適切な参照ポイントを提供するように調整されれば、これは本当に楽しいメカニズムになる可能性があります。特に、最終形態では小さなセーフスポットが作成されるためです。プレイヤーがすべての蛇を見て、それらがヒットする場所を大まかに把握するのに十分な時間があり、床の参照を使用して正確な場所に移動します。そのため、プレイヤーの努力に対して、小さなセーフスポットにもかかわらずすべての AOE がキャラクターを逃すのを見るスリルと、前述のジョブ最適化の機会が報われることになります。しかし、現状では、これらの機会は実際にはプレイヤーには提供されておらず、セーフスポットの大まかな場所しかわからないため、テレグラフに反応してキャラクターの位置を調整することを心配することになります。

最大の迷宮

このメカニズムは、すべてのパーティーメンバーをアリーナの正確な中心に引き寄せることによって開始されます。これは、状況によっては役立ちます。しかし、ここではプレイヤーは視覚的に要求の厳しいタスクにすぐに取り組まなければならないため、引き込みはアリーナの視界とキャラクターの位置の個人的な追跡を妨げます。

プレイヤーが自分のキャラクターを動かすと、その動きをアリーナに対するキャラクターとカメラの動きの追跡に自然に含めます。これにより、地球上での移動のために進化したのと同じ脳システムの多くを使用して、カメラの動きを部分的に自動的に補正できます。強制移動メカニクスではこれが許可されず、代わりに、キャラクターが新しい場所に着くと、プレイヤーは追跡を再度確立する必要があります。(特に、この例のように、プレイヤーに事前の指示を与えない強制移動の場合。プレイヤーに強制移動の計画を立てるように求めるメカニクスは、少なくとも混乱に備えさせます。特に、Eden's Verse: Iconoclasm (E7N) とその Savage バージョン (E7S) のように、プレイヤー キャラクターの新しい場所を示す場合はそうです。)

これは、強制移動メカニクスが本質的に悪いということではありません。まったく違います。強制移動は、The Interphos の浮遊ミニフェーズの開始や、E12S の各 Relativity の終了など、一部のメカニクスで見事な効果を発揮するために使用されています。

ただし、強制移動メカニクスと視覚的に要求の厳しいタスクを配置する場合は、強制移動がプレイヤーの視覚追跡に与える影響を考慮する必要があります。矢印の配置を操作する視覚的に要求の厳しいタスクがメカニクスの中核であるこのコンテキストでは、直前に強制移動を行うと、面白くも必要もなく、タスクが中断されます。

アリーナの変形を損なわずにこの混乱を回避するには、Everkeep が Dawn of an Age で行っているのと同じ手法を使用します。つまり、変形しようとしているアリーナの部分に十分な時間をかけ、その AOE で、キャストの最後にそこに残っているすべてのプレイヤー キャラクターを KO (および/または移動) させます。プレイヤー キャラクターが形成されたばかりの矢印に足を踏み入れないように、その AOE をもう少し長く延長したり、アリーナの変形中にバインドしたりしても問題ありません。

アリーナの変形後、プレイヤーには「あなたは定命の苦しみから逃れたいと感じています...」という赤いテキストが表示されます。これは、初めてのプレイヤーやカジュアル プレイヤーに、正しいコーナー タイルで浄化されないとキャラクターが KO になる Greatest Curse デバフを警告するためのものだと理解しています。しかし、これは赤いテキストの適切な使用方法ではありません。

まず、このテキストは失敗の結果のみに焦点を当てており、プレイヤーが成功するために何をしなければならないかに焦点を当てていません。このような焦点は、成功への道筋について多くを語るとパズルが台無しになってしまうような、一部のパズルの仕組みには適しています。しかし、これはパズルの仕組みではなく、高速な視覚処理の仕組みです。第二に、このテキストは結果について曖昧で、飾り立てています。プレイヤーがデバフ バーの読み方や、失敗した仕組みのさまざまな結果に十分慣れていても、「失敗すると KO になる」という結論に達する前に、頭の中でこれを解釈する必要があります。ここでも、時間が重要です。飾り立てたテキストは、伝承パズルでは役立つことがありますが、これはそうではありません。

テキストを「迷宮を進んで呪いを解け!」に変更します。これは単純明快で成功への道筋に焦点を当てており、デバフについては、プレイヤーにその存在を知らせるために必要な場合にのみ言及しています。

では、この仕組みで成功するには、プレイヤーは何をしなければならないのでしょうか。 (「迷宮を進む」よりも詳細です。)

デバフを解消して KO を回避するには、プレイヤーは青い円でマークされたアリーナの 1 つのコーナー タイルにキャラクターを移動させる必要があります。矢印タイルは、プレイヤーが通常どおりにキャラクターをそこに移動することを妨げます。そのため、プレイヤーは矢印 (迷宮) を調べて、どの直接到達可能なタイルがキャラクターを目的のコーナーに移動させるかを確認する必要があります (変更されていない中央エリアに隣接する 8 つのタイルのうち、そのようなタイルは 1 つだけです)。

このタスクの手順を分解してみましょう。特に、メカニズムが何であるかは理解しているが、隠されたパターンについて何も知らないプレイヤーに解決策を保証する最も効率的な一般的な方法を分解してみましょう。

まず、プレイヤーはアリーナのコーナーを調べて、青い円でマークされた 1 つのタイル (迷宮の概念的な終点) を見つける必要があります。 (Greatest Curse は残り 16 秒になるまで、コーナーのマークははっきりと見えませんが、開始時は 18 秒です。) どこから見始めるかによって、4 つのコーナーすべてを確認する必要がある場合があります。初めてプレイするプレイヤーは、そのタイルが正しいものであることを知るために円を見る必要がある可能性が高いため、中心視野でコーナーを個別に確認する必要があります。(以前にダンジョンをクリアしたことがあっても、青と紫の区別が難しい色覚異常のあるプレイヤーにも同じことが言えます。) 前述の色覚異常を除けば、以前にこれを行ったことがあるプレイヤーは、周辺視野を使用して一度に複数のコーナーを確認できます。いずれの場合も、カメラを回転させる必要がある場合があります。キャラクターを正確に中心に配置してアリーナの 4 つのコーナーすべてを見るには、カメラの傾きと角度を正確に調整する必要があり、その場合でも、HUD 要素で覆われていることが多い画面の端を使用するためです。 (皮肉なことに、これを可能にするカメラの傾き設定は、ヴァリガルマンダのくちばしが稲妻のくちばしのアニメーション中に画面から外れるようにする設定と同じです。さらに、プレイヤーがキャラクターを自分で変更されていない中央領域の端に移動させると、カメラの回転をなくす実行可能な設定と角度の範囲が広がります。)

次に、プレイヤーはエンドポイントから中央エリアまで逆方向に作業する必要があります。プレイヤーはタイルを 1 つずつ調べてパスを見つけ、最小のスケールで「このタイルにどうやって行けるか」という質問を何度も繰り返します (コンピューター サイエンスに詳しい読者なら、この方法が幅優先探索であることに気付くでしょう)。そこで、エンドポイントに隣接する 2 つのタイルを調べ、それぞれの矢印がエンドポイントを指しているかどうかを確認します。プレイヤーが近接戦闘のアップタイムのためにキャラクターをアリーナの中央に戻すことを可能にする隠されたパターンにより、それらのタイルの 1 つだけがエンドポイントを指しています (プレイヤーは必ずしもそれを知っているわけではありませんが、隣接する両方のタイルをチェックした後、両方へのパスを探す必要がなくなるという利点があります)。エンドポイントを指している 1 つのタイルを中間目標として設定します。そこにあるパスはすべてエンドポイントにつながるからです。

次に、中間ゴールからその隣の 2 つのタイルを見て (エンドポイント自体はすでにチェック済みなのでカウントしません)、それぞれの矢印が中間ゴールを指しているかどうかを確認します。そのうちの 1 つまたは両方が指し示していれば、中間ゴールをその新しいタイルに設定します (はい、最終的には中間ゴール タイルを指し示すタイルが複数存在する可能性があります)。次に、新しい中間ゴールに隣接している、まだチェックしていないすべてのタイルを見て、再び矢印が内側を指しているかどうかを確認します (一部の中間ゴール タイルは、アリーナのどの端にもないため、最終的に 3 つのチェックされていない隣接タイルを持つことになります)。このプロセスを繰り返し、変更されていない中央エリアに隣接する中間ゴール タイルを見つけます。これは、通常のキャラクターの移動でそのタイルに到達できることがわかっているためです。

最後に、キャラクターを直接到達できる中間ゴール タイルに移動し、矢印がキャラクターをエンドポイントに運ぶのを待ちます。パスを移動する際にプレイヤーからのさらなる入力は必要ありませんが、デバフが期限切れになる前にそれが発生するまで時間を残す必要があります。

(直接到達可能なタイルから前向きに進むのはどうでしょうか。それは魅力的ですが、そうすることでプレイヤーはアリーナのあちこちを巡って長くて無駄な道を調べずに済むという保証はありません。(それを保証するさらに隠されたパターンがあったとしても、プレイヤーは多くの失敗を犯すか、多くのビデオを参照するかのいずれかのリスクを負わなければそれを見つける方法がありません。通常のダンジョン ガイドは、通常、そのような詳細な知識なしに、任務のリリース直後に作成されます。また、後ろ向きに進むと、プレイヤーは直接到達可能なタイルのほとんどから目をそらすことになります。さらに、この時点までのパッチ 7.0 任務では、プレイヤーはメカニクスが一貫して視覚的に合理的であるという確信を持てませんでした。) また、キャラクターを間違った道に置いた場合、強制カメラ モーションによってその間画面が実質的にぼやけてしまい、再試行する時間が短くなります。)

(後ろ向きに進むが、現在の中間目標タイルに隣接する各タイルをチェックし続けるのではなく、新しい中間目標タイルが特定されるとすぐに検討するのはどうでしょうか。そうすれば成功は保証されます。実際、コンピュータ サイエンスでは、これは深さ優先探索と呼ばれます。しかし、これでは、プレイヤーの作業記憶に対する要求が大きくなり、未チェックの各タイルをより長く保持して、より大きなリストを蓄積することになりますが、最悪のシナリオは改善されません。)

これには、全体として多数の視覚チェックが必要であり、そのほとんどは矢印の方向を確認することです。矢印は茶色の上に緑色で、緑色は暗い色から、矢印の方向とはまったく関係のない方法で軽く光る色まで変化します。光は視覚的な乱雑さであり、プレイヤーの助けにはなりません。したがって、矢印の方向を確実に識別するには、プレイヤーはその矢印を中央視野に入れるだけでなく、その矢印の上で目の動きを少し止めて、周囲の茶色のタイルから矢印を見つけ出す必要があります (特に、グラフィック更新による強制ブルームに対処するために後処理を使用する場合)。

(矢印自体ではなく、隣接するタイルを見ると矢印の方向を確実に識別できる場合があります。しかし、私はそうではありません。最近、周辺視力と色覚検査を含む視力検査を受けました。したがって、その能力がすべてのプレーヤーに備わっているわけではないことは明らかです。視覚障害のあるプレーヤーの場合はさらに悪いとしか思えません。そのため、前の段落で説明した方法で矢印の方向を確認することに基づいて進めます。これはプレーヤーのより現実的な想定だからです。)

また、プレーヤーが各ステップで確認するタイルは互いに隣接していないため、さまざまな方向に短い目の動きが生じます。その結果、プレーヤーはフリック、ストップ、フリック、ストップの繰り返しの目の動きパターンを強いられ、障害がなくても時間的プレッシャーが増すため、身体的に不快な場合があります。

悪いケースでこれがどの程度の目の動きになるか、例を挙げて確認してみましょう。

プレイヤーが探索中に 9 つの目標タイルを通過し、その中にはエンドポイントは含まれるが直接到達可能なタイルは含まれず、各タイルにはまだチェックされていない隣接するタイルが 2 つあるとします。(長いパスや複数の中間目標タイルを同じステップで処理する場合、これは非常に起こり得ます。私はこれが最悪のケースであるとさえ確信していません。) また、プレイヤーがタイルを最悪の合理的な順序でチェックすると仮定します。そのため、エンドポイントを見つけるためにカメラを回転させ、最後の目標タイルまで各目標タイルの隣接するタイルの両方をチェックする必要があります (最後の目標タイルで直接到達可能なタイルを優先するように推論できます)。

カメラを回転させることは、片方の目の動きと同じように機能します。その後、プレイヤーは、たとえ周辺視野でエンドポイントを特定できたとしても、エンドポイントに隣接する 1 つのタイルに目を移動し、次にもう 1 つのタイルに目を移動する必要があります。同じことが、最後の目標タイルまで、残りの各目標タイルにも当てはまります。つまり、カメラの向きを変えることで 1 回目を動かし、最初の 8 つの目標タイルでそれぞれ 2 回目を動かし、最後に直接到達可能なタイルをチェックするために目を動かします。18 回の目の動きがあり、それぞれに短い停止が必要です。

突然、16 秒はそれほど長くありません。プレイヤーが方向転換する時間、プレイヤーがキャラクターを移動する時間、キャラクターが矢印のパスに沿って移動する時間、およびクライアントとサーバーの遅延を考慮してもです。実際、これらを含めると、このメカニズムでは成功を確実にするために 1 秒あたりほぼ 2 回の目の動きが必要です。

これは、TOP の Party Synergy (最初のクローン AOE) や Predation Dodge などの Ultimate メカニズムよりもさらに高い目の動きの頻度です。これらのメカニズムでは、プレイヤーは計画されたスプリントでキャラクターの移動時間を短縮でき、視覚処理速度を向上させるための十分な練習が期待されます。 (キャラクターの移動時間の一部は矢印のパスに沿っているため、ここではスプリントはあまり役に立たないことに注意してください。ここでは移動速度バフが役に立ちません。) 確かに、これには大きな目の動きが含まれており、それは重要ですが、反復的な動きと急速な停止と開始はどちらもそれ自体が身体的な負担です。

(パーティシナジーの方法: クローンが表示される前にプレイステーション マーカーを確認します。そのため、クローンではなくマーカーに目を向ける必要があります。中央のクローンは無視し、スポーン音と周辺視野で大まかにサイド クローン (と目) の位置を特定します。プレイステーション マーカーを確認した後、視覚的に最も近いサイド クローンの腕を確認します。腕が空の場合は、足を見下ろして、スケート靴を履いているか、通常の足を持っているかを確認します (スケート靴はスーパーリミナル スチール、通常の足 + 空の手はデコイです。デコイは 1 つだけです)。デコイを最初に見つけた場合、近くにある次のクローンの腕を確認します。武器を特定したら、デコイが見つかったかどうかに関係なく、キャラクターを 2 つの潜在的なセーフスポットの大まかな領域に移動し始め、アリーナの反対側にもう 1 つのクローンを探します。再びクローンの腕を確認します。腕が空でデコイをまだ見つけていない場合は、足を確認します。2 番目の武器を特定すると、検索が停止します。)

(キャラクターの移動開始までの最悪のケース: 3 回の目の動き (腕、足、腕)。解決を完了するまでの最悪のケース: 5 回の目の動き (腕、足、腕、足、腕)。目は 1 回だけ画面を横切る必要があります。クローンの出現から AOE スナップショットまでの時間: 5 秒。これは 1 秒あたり 1 回の目の動きをわずかに上回り、デコイによる視覚的な乱雑さはありますが、Greatest Labyrinth よりも少ないです。)

(捕食回避法: 最初に出現したクローンの 1 つの腕を見ます (ここにはデコイはありません。空の腕には常にスケートが付属しています)。次に、その反対側のクローンの腕を見ます。キャラクターを 2 つの潜在的な安全地帯のほぼ間に配置して、次のクローンの腕がある残りのインターカーディナルの 1 つに目を向けます。そのクローンが出現してその武器を特定したら、その反対側の最後のクローンの腕を見ます。次に、砂時計の切り裂きの輝きが表示されるまで Final Omega を見ます。キャラクターを最初のクローンに移動します。正確な位置合わせのために最初のセーフスポットを見ながら、セーフスポットを狙います。また、位置合わせを助けるために、2 番目のセーフスポットを事前に見ます。)

(最初のクローンの出現から最初の AOE スナップショットまでの 11 秒間に、毎回 7 回の目の動きがあり、視覚的な混乱はありません。目は画面を 2 回横切るだけで済みます。Party Synergy を行える場合は、Final Omega の重複を処理するための余分な時間があれば、ほぼ確実にこれを行うことができます。)

一部のプレーヤーは、各 AOE ペアを識別する際に常に同じ性別のクローンから開始することを好むため、TOP で今説明した方法よりも視覚的に要求の厳しい方法を使用することを選択します。ただし、これはオプションです。これを行うと、より困難な目の動きのタスクと引き換えに、より簡単なパターン マッチングのタスクを選択していることになります。より困難な目の動きに苦労している場合は、より簡単な目の動きのタスクを選択し、練習と記憶によっていずれかのクローンからパターン マッチングを学習することで、成功への実用的なルートを得ることができます。

Greatest Labyrinth には代替オプションはありません。 (試行錯誤や仲間のパーティメンバーを追うことは別ですが、どちらも信頼性が低く、メカニズムを実際に解決するものではありません。) フリック、ストップ、フリック、ストップという繰り返しの目の動きが身体的に不快、怪我、またはまったく不可能だと感じるプレイヤーは、そのままでいるしかありません。

なぜ通常のダンジョン メカニズムは、生の目の動きの頻度とその頻度の維持においてアルティメットよりも要求が厳しいのでしょうか。どちらも、プレイヤーが制限されていると感じても変更できません。

これはばかげています。

この視覚タスクの強度を別の方法で調べると、人間の体 (脳を含む) が「フリック」目の動きを開始するのにかかる時間は、同じ人を異なる時間にテストしてもかなり異なりますが、通常は約 200 ミリ秒、つまり 5 分の 1 秒です。プレイヤーが見たものに応じてどこを見るかを決めるのに費やす時間は考慮されていません。したがって、プレイヤーに 1 秒間にほぼ 2 回の目の動きを求めるということは、プレイヤーに、平均して検索時間のほ​​ぼ 40% を文字通り目が動くのを待つことに費やさせることになります。これもまた不合理です。

(読者の中には、顔を調べるなどの状況では、人間が 1 秒間に定期的に複数の目の動きをすることをご存じの方もいるでしょう。顔を見るというタスクでは、人間の脳がさまざまな顔の特徴を見るために事前に複数の動きを計画できるため、遅延に陥らないと考えられます。これは、Greatest Labyrinth ではほとんど許可されない計画の一種ですが、Party Synergy ではある程度許可され、最初のクローンが現れると Predation Dodge で完全にサポートされます。また、このような頻繁な目の動きが一般的なタスクでは、このようなハード ストップもないため、読むときに誤って繰り返された単語などの詳細を見逃すことがよくあります。自然な速度で読む方がはるかに速いにもかかわらず、Greatest Labyrinth よりも読書の方がはるかに快適であると断言できます。)

この要求が明らかな災害をもたらさなかった唯一の理由は、プレイヤーが周辺視野でパーティ メンバーの試みに気付くことができ、通常のダンジョン ボスではパーティの KO で特に多くの時間を失わないことです。

特に隠しパターンを考慮すると、時間制限を延長すると状況が改善される可能性があります。(隠しパターンは適切に設計されており、より速いプレイヤーは正しい矢印のパスを移動するのに要する時間を見積もる必要がなく、ジョブの稼働時間を最適化できます。) そのため、Deltascape V3.0 (O3N、ハリカルナッソス) は、比較的低いコントラストにもかかわらず、視覚的な迷路に悩まされることはありません。プレイヤーがタイルの動きを利用して周辺視野でタイルの方向を見つけることができない場合でも、探索時間ははるかに長く、途中でいくつかの静止タイルに向かって作業する必要があります。プレイヤーは、余裕のあるタイマーを使用して、最悪のシナリオの長さを制限するいくつかの小さな探索を要求します。

しかし、もっとエレガントな解決策があると思います。そのようなメカニズムから期待されるように、輝きがプレイヤーの視覚を助けるようにします。

各矢印は、すでにその先端となる 1 つの大きな三角形、先端に向かって 2 つの V 字型、そして中央を埋める小さな三角形で構成されています。私は、先端の三角形をブルームに関係なくはっきりと輝かせ、各矢印の残りの部分を 1 つの輝かない中程度の緑色にします (ダンジョンの前の矢印スライドを含む)。アイデアは、プレイヤーが中央視野でエンドポイントまたは中間目標タイルを見ているときに、中央視野ではない視野で、その方向を指している隣接する矢印の輝きを確実に見ることができ、他の方向を指している矢印とは明確に区別できるようにすることです。これを確実にすることで、中間目標タイルにならない隣接するタイルを直接見る必要がなくなります。これにより、必要な目の動きがほぼ半分に減り、動きの方向がはるかにスムーズになるため、視覚的な迷路のアイデアや複雑さを変えることなく、視覚的な要求をはるかに管理しやすくなります。

プレイヤーが視覚的なインジケーターをより速く見て処理する必要があるほど、より明確にする必要があります。 ここで必要な基準は、プレイヤーがより速く作業する必要があるため、O3N よりもはるかに高くなります。

湿った召喚

これは単純なマルチヒット スタックですが、ヒットごとにスタック領域にヒットするサーペントの追加による水たまりが残ります。その結果、初めてプレイするプレイヤーは水たまりの存在に反応する必要があるだけでなく、毎回、パーティー メンバーが水たまりからどこに移動するかの選択に反応する必要があります。

仲間のプレイヤーに反応する必要があることは大惨事ではありませんが、十分な遅延が発生し、各水たまりがアクティブになるまでの時間がイライラするほど短くなります。これは、回転する円錐状の AOE など、パーティに動きの方向を示唆するものと重なると、反応性が低くなりますが、技術的には重なりによってメカニズムが難しくなります。

Greatest Flood

コーナー タイルからの単純なノックバック。具体的には (アリーナの中心に対して) 最も近いコーナーよりも少し外側です。これは、Moist Summoning の水たまりが残っているため複雑になります。

優れたメカニズムです。水たまりは、難しすぎることなくプレイヤーに配慮を示し、プレイヤーはノックバック耐性をアップタイムや水たまりの悪い配置に対処するために使用できます。

Great Torrent

本当に楽しいメカニズムです。プレイヤーはテレグラフを使用して、シーケンスの最初の AOE の外側に正確に配置することができ、シーケンスが完了した後に広がるのに十分な時間があります。

全体的な印象 (ボス)

このボスには、優れたフロア ジオメトリとそれにマッチするメカニズムが組み込まれていますが、正確な回避を可能にするフロア ジオメトリは実際には提供されていません。Greatest Labyrinth は、一部のプレイヤーに成功への実行可能なルートが残されていないため、低ポイントでもあります。

全体的な印象 (ダンジョン)

雑魚モンスターは面白いですが、全体的にそれほど難しくはありません。しかし、すべてのボスに深刻なビジュアル デザインの問題があります。

このダンジョンのすべてのボスにビジュアル デザインが問題になっているという事実は、1 つのメカニックや 1 つのボスだけではなく、視覚的な明瞭さを下げれば難易度が楽しくなるかどうかのテストのように感じられます。あなたがそのようなテストとして意図していたかどうかにかかわらず、私は答えます。いいえ、それは大多数のプレイヤーがあなたに有益なフィードバックを与えるものではありません。

平均的な Ultimate レイダーは、メカニックに必要な能力の種類や、メカニックを実行するために目をどのように動かすかについて意識的に考えません。一般的なカジュアル プレイヤーは、そのようなことについて考える可能性はさらに低くなります。必要な視覚的指標や参照が不明瞭なためにメカニクスに苦戦している場合、「この任務はオレンジの上にオレンジ」のような明らかなケースを除いて、彼らは通常、視覚的な明瞭さの欠如が問題であるとは認識しません。

代わりに、彼らは、ある時点を超えて改善することなくメカニクスに失敗し続けていることにのみ気づき、それを理由に単に「これは難しすぎる」と言うかもしれません。彼らは、バレルテンダーが作成したサボテンのように、視覚的指標の核心部分を完全に見逃し、デザイナーの意図に関する誤った考えに基づいて評価を行うかもしれません。彼らは、必要なすべての情報を時間内に入手できないことに気づき、メカニクスまたは任務を「不公平」と見なすかもしれません。特に、バレルテンダーはダンジョンの最初から不公平になる意思を示しているからです。彼らは、必要な情報を見分けるのが遅れていることに気づき、その情報で決定を下す時間が非常に短かったことを正しく認識しますが、遅延の原因が視覚が不明瞭であることではなく視覚がないことであると意識的に認識していないため、メカニクスを「反応的」と宣言します。あるいは、単に気に入らないだけで、その理由を突き止められないだけかもしれません。

これらすべては、「Dawntrail は難しすぎるのか?」という怒りのやり取りだけでなく、文句を言いやすいことでもかき消されてしまいます。平均的なカジュアル プレイヤーは、Strayborough 人形が迷惑だと認識できるので、それについて多くの苦情が寄せられます。平均的なカジュアル プレイヤーは、私がここで行ったように、視覚デザインの選択がメカニクスの視覚処理にどのように影響するかを説明することは、労力や知識の面ではまったく不可能です。

一方、不明瞭なビジュアルを処理できるプレイヤーはまったく影響を受けません。彼らにとって、メカニクスはそれほどエキサイティングではありません。なぜなら、その背後にあるメカニクスは同じだからです。アリーナとアニメーションは、伝承やストーリーにそれほど適合していません。なぜなら、必要なメカニクスの指標として明確なビジュアルで、多種多様な伝承コンセプトを提示できるからです。私はこのダンジョンのビジュアルデザインをどのように変更するかを説明し、伝承の適合性と視覚的な明瞭性は相互に排他的ではないことを実証しました。そしてもちろん、メカニクスに関係のないビジュアル要素は、アーティストや伝承ライターが望むほど微妙なものにすることができます。

このレビューのデザイン原則のセクションで書いたとおりです。必要な視覚的インジケーターを見にくくすることは、楽しさを損ないます。

時間をかけてこの詳細な内訳を説明したのは、このゲームのリアルタイム性に適したビジュアルの明瞭性とアクセシビリティの標準を採用し、それに従ってほしいからです。プレイヤーが Eden's Verse: Furor (E6N、Ifrit および Garuda)、Asphodelos: The Third Circle (P3N、Phoinix)、およびそれらの Savage バージョン (E6S および P3S) について語ったように、オレンジの上にオレンジを重ねることはもうありません。これらは 2 つのレイド ティアが離れているため、Asphodelos の制作パイプラインで Eden's Verse からのフィードバックに対応する時間がありました。色覚を必要とする、または色覚がないと非常に見えにくいインジケーターはもう不要です。また、プレイヤーが見る必要のあるものは、ぼやけた塊のままになってはいけません。

視覚的な明瞭性とアクセシビリティの基準を設計プロセスに組み込みます。各視覚要素を使用するメカニズムの要件を反映させます。真剣に。

--------

ストレイボロー デッドウォーク (レベル 100 ダンジョン)

https://i.imgur.com/XhhwIG1.png

最初のボス前のモブ

4 つのパック

ここではメカニカルに興味深いものはありませんが、2 回引くことができます。

最初のボス - レオノッグ王家の長 I

https://i.imgur.com/TRnfomJ.png

落下する悪夢

落下する頭を各プレイヤー キャラクターの上に置くことで、人形の失敗に対する罰の一部を導入しますが、プレイヤー キャラクターは人形に縛られていないため、回避する機会があります。これは良い紹介になるでしょうが、頭と床のコントラストがほとんどなく、人形の扱いがひどいです。

(頭は 10 秒間の Benoggined デバフを与えます。これは「はは、失敗した」というビジュアルで、Out of the Action と Heavy を組み合わせたものと同じです。)

頭にはコントラストの高い緑のセクションがありますが、画面上の 2D 形状は、ボスがプレイヤーのカメラに対して頭をどの方向に向けるかによって大きく異なるため、プレイヤーはこれらのセクションの使い方をすぐには習得できません。

Team Spirit

アリーナの端に並んで行進の準備をしている人形に短い警告を投げかけます。それ自体は良いことです。

Spirited Charge

人形がアリーナを行進し始めます。

これらの特定の人形は、パッチ 7.0 で最も苦情の多いメカニズムの 1 つであるため、詳細に説明しましょう。プレイヤーはこれまで、The Grand Cosmos のほうきなどの動くオブジェクトの仕組みを楽しんできたのに、なぜこれがこんなにも迷惑なのでしょうか?

まず、人形がかなり速く移動していることに注目してください。Honey B. Lovely の動くハートよりも速いのです。人形のスピードは、The Skydeep Cenote で Feather Ray が作り出した弾幕の泡に匹敵するように感じられます。そのため、それらの泡と同じように、プレイヤーの画面には人形に当たったことが伝えられると同時に、その位置が正確であればうまく回避できたであろう位置も表示されます。Expert Roulette で人形に何度も遭遇すると、結果として生じる「私はそれから離れていた!」という感情は、さらに不快なものになります。

次に、人形が並んだ後の利用可能な時間は、すべての人形の開始位置を確認するのに十分ではありません。人形がプレイヤーの画面上で最初に開始位置に静止してから、人形が視覚的に動き始めるまでの間にはわずか 4 秒しかありません。つまり、プレイヤーが最初に人形の開始位置を見る機会を得てから、人形が実際にサーバー上で動き始めるまでの時間は、さらに短いということです。これでは、各辺に 4 体ずつ、ほぼ正方形に並べられた 16 体の人形を確認するには、まったく不十分です。(また、重なり合うメカニズムのため、プレイヤーは、自分のキャラクターや 1 つの潜在的な安全地点の近くを行進しようとしている人形だけでなく、すべての人形を見る必要があります。) さらに悪いことに、アリーナは円形で、そのマーキングは人形の「レーン」に対応していないため、人形の将来の進路を視覚的に追跡することがより困難になっています。また、人形は比較的密集しているため、プレイヤーは安全エリアを見つけるために正確な位置を必要とし、中央視野からそれを見つけなければなりません。

3 番目に、開始位置だけでは、人形の全体の経路や、人形がその経路を横断するタイミングを知るのに十分ではありません。人形の広場内の異なるグループは、異なる時間に動き始めますが、人形の動きを観察する以外にはまったくマークされていません。(順序が静的であったとしても、プレーヤーはそれを見つけるために、仕事や回避を気にせずに人形を長時間観察する必要があります。東のグループが常に最初に移動する静的な順序だと思います。) そして、人形が Team Spirit のラインアップから行進した後、アリーナのほぼ端 (おそらく隠し広場の端) に到達すると、パーティー メンバーの位置を (一見ランダムに) 狙い、その方向に 2 回目に進みます。 (戦闘のあるセクションでは、人形はターゲットを選んで行進するというこのプロセスを何度も繰り返します。ボスは他のメカニクスをキャストするのに忙しく、別のチームスピリットやスピリットチャージを提示する余裕がないためです。) その結果、理論上でさえ、各人形が向きを変えるだけでなく、最後の行進でどのプレイヤーキャラクターをターゲットにし、どの方向を選んだかを示すのに十分な距離を移動するまで、プレイヤーは人形の進路とタイミングを知るのに十分な情報を持っていません。 人形の向きから人形の目的を判断しようとすることはできますが、各人形の正確な向きを見るには中心視野が必要であり、グループ内のすべての人形を一度に中心視野に入れることはできません。 (繰り返しますが、人形の頭は床とほとんどコントラストがなく、人形のドレスもそれほど良くはありません。)

4 番目に、プレイヤーが対処しなければならないのは人形だけではありません。最初のスピリットチャージでさえ、人形が行進を終える前に別のメカニクスが解決し始めます。 当然、プレイヤーは複数のメカニクスを同時に処理するため、メカニクスの取り扱いのあらゆる面でより大きな要求を課せられます。それに加えて、重複するメカニズムの性質上、人形に対処する最も快適な方法が妨げられます。

人形のみの場合の推奨戦略は、できるだけ多くの人形が 2 回目の行進のターゲットを選択する間、アリーナの中央で静止し、その後、2 つの人形の進路の間に中央から移動することです。(アリーナの中央は、プレイヤー キャラクターと人形がターゲットにする可能性のあるポイントとの間の最小距離を最大化するため、人形がターゲットを選択してからキャラクターを移動させるかメカニズムが失敗するまでの、プレイヤーにとって最長の保証されたリード タイムを提供します。) しかし、この戦略は重複するメカニズムのすべてに失敗します。技術的には、Looming Nightmare の追跡 AOE により、それらの AOE のターゲットになっていないキャラクターのパーティーの半分がこの戦略を採用できますが、それを使用するすべてのプレイヤーを確実にカバーするには十分ではありません。 一部のプレイヤーだけを不快なタスクに押し込むのは、やはりゲーム デザインとしては良くありません。

5 番目に、これは反応的なメカニズムです。プレイヤーは厳密には他のプレイヤーの選択に反応する必要はなく、人形の進路に反応するだけで済みますが、実際にどのような情報を持っているか、いつそれを受け取ったかを考えると、人形の進路に反応する時間はほとんどありません。

人形の最初の行進で 1 つの場所の安全性を確認するためだけでも、プレイヤーは隠された正方形の 4 辺すべてをチェックして、通り過ぎる人形の正確な位置を確認する必要があります。間隔が狭いため、少なくとも 8 体の人形が必要です。プレイヤーの視覚的な進路追跡能力の問題はさておき、人形の位置を中央視野に収めるだけで 8 回の目の動きが必要です。そのため、人形が動き出す前にこれを行うには、人形ごとに停止しながら 1 秒あたり少なくとも 2 回の目の動きが必要です。この速度は、すでに深刻な問題であった Tender Valley の Greatest Labyrinth よりもさらに速く、負担が大きいため、プレイヤーがそれを実行できないという現実的な仮定しか立てられません。

代わりに、プレイヤーは人形が動き始める前に、隠された正方形の一部の辺のみを確認します。 (たとえば、これを行う最も快適な方法は、隠された正方形の角を選択し、それを見ながらキャラクターをその方向に移動させ、近くの人形が自分のキャラクターのそばを通り過ぎるように位置の正確な選択を洗練することです。 即時の脅威となるすべての人形は視覚的に互いに近く、潜在的な安全地点に近いため、プレイヤーは人形間のスペースを見ることができます。 プレイヤーは、キャラクターを選択した安全地点に配置するために他の場所を見る必要がなく、角は隣接していない辺の人形が付近に到達するまでの時間が最も長くなります。) しかし、これは、人形がすでに画面上で移動している間に、プレイヤーが正方形の他の辺を確認しなければならないことを意味します。

人形は視覚的にアリーナを約 7 秒で横切ります。これはかなり長い時間のように思えます。しかし、プレイヤーがキャラクターの移動を完了するのにかかる時間はそれよりも短くなります。これは、人形の実際のサーバー位置がクライアントの表示よりも前にあることと、キャラクターをアリーナの端に配置すること自体がメカニズムの失敗であるためです (人形に捕まるよりも罰則の少ないメカニズムの失敗です。壁は通常の即時 KO ではなく DOT を与えるからですが、それでも失敗です)。キャラクターの移動には時間がかかり、残りの人形の進路に反応する意思決定の側面も同様です。キャラクターの移動は、サーバーに到達して処理されるまでに時間がかかります。近接 DPS をプレイしている場合 (またはパーティーに近接 DPS がいるタンクがいる場合)、キャラクターをコーナーに移動させたくないでしょう。コーナーを選んだとしても、最初にスタートする人形の隣のコーナーではなく、最後にスタートする人形の隣のコーナーを選んだ可能性があります。覚えておいてください。隠れたパターンがあっても、初めてのプレイヤーはそれを知りません。 (そして最後の人形は最初の人形の約 2.5 秒後に動き出すので、それ自体が隠されたパターンを知っているテスターに​​かなりの余分な時間を稼ぐことになります。) プレイヤーは、自分のキャラクターの近くにある最後の人形が動くのを待っている間、自分のキャラクターから遠く離れた最初の人形を見ることができます... しかし、初めてのときは、自分のキャラクターの近くにある人形がいつ動き出すかわかりません。そして、後の人形や重複するメカニズムのために自分のキャラクターを安全に動かすことができるタイミングを知るために、その情報が必要です。 (同様に、人形のヒットボックスが視覚モデルよりも大きいかどうかも知っておく必要があります。) そのため、代わりに自分のキャラクターや近くの人形を見る可能性が高く、タイミングを十分に理解してそうするのをやめるまで、時間的なプレッシャーがさらに悪化します。

最悪のケース、つまり仕事で近接距離内に留まるように訓練されている初めてのプレイヤーを想定しないようにしましょう。代わりに、プレイヤーが選択した最初のキャラクター位置を隅に置くと、近くからスタートする人形を回避できると確信し、最終的にその最初の位置の近くにある遠い側から人形を探しにスムーズに移動することを考えてみましょう。

これらの人形は、アリーナの端に近いものの、完全にはそこに達していないため、人形の動きが始まってから 6 秒で選択した最初のキャラクター位置に到達するとします。クライアントとサーバーの往復に 0.5 秒、およびプレイヤーのクライアントとサーバーの内部遅延をすべて差し引きます。最終的に決定した移動距離を取得するには、移動キー、ジョイスティック、マウス ボタンを 0.25 秒間押し続ける必要があり、さらに 0.25 秒を生の生理的、非意思決定の反応時間に残すとします。つまり、必要な視覚情報をすべて取得し、移動する場所を決定するのに 5 秒かかります。

先ほどスキップせざるを得なかった 2 つの側から、ほぼアリーナの反対側にある 2 つのグループで、確認する人形が 4 つ残っています。人形の進路を正確に決定するには、人形の位置を正確に特定する必要があります。そうしないと、後で進路予測を更新するためにもう一度見直す必要が生じます。人形はすでに動いているため、人形は当然どこを見ればよいか正確にはわからないため、周辺視野からそれを把握する必要があります。そうそう、この間にもう 1 つの重複するメカニクスが始まります。このメカニクスの詳細は、戦闘タイムラインを完全に知っていたとしても決定に影響します。ほとんどのプレイヤーはエキスパート ルーレット ダンジョンのタイムラインに関する完全な知識を得ることはなく、それを理解したプレイヤーでも以前の経験が重要であるため、タイムラインの知識は持っていないと見なす必要があります。したがって、重複するメカニクスには少なくとも 2 回の目の動きを割り当てます。1 回はキャスト バーに、メカニクスがいつ開始するかを大まかに把握し、もう 1 回はキャスト名または目を引くアリーナ テルに目を動かします。つまり、プレイヤーが決定を下す前に 6 回の目の動きが必要になります。

プレイヤーが、他の人形のグループとキャストバーへの 2 つの大きな動きを含め、1 秒間に 1.5 回の目の動きをすることができ、各オブジェクトで十分な時間停止して必要な情報を引き出すことができると仮定します。(Greatest Labyrinth の分析で、この速度は Ultimate の要件をはるかに上回っていることを思い出してください。) この場合、プレイヤーは 4 秒で 6 回の目の動きを完了でき、キャラクターをどこに移動するかを決定する時間は 1 秒になります。これはメカニカルなオーバーラップとしては狭く、プレイヤーが以前に学習した一般的な回避戦術に大きく依存しますが、管理可能です。実際、このケースは、このメカニクスの最初の部分を私が処理する方法です。

しかし、これは最悪のケースではありません。

プレイヤーが初めて完了するためにここにいて、近くの人形が通り過ぎるのを待ってキャラクターの周りを見回している場合、意思決定の 1 秒全体が簡単に消費される可能性があります。すると、プレイヤーはスピードのために信頼性の低い方法 (動いている人形の間のスペースを探したり、周辺視野で人形を追跡するなど) を試すか、不十分な情報に基づいて故意に作業するか、進行中のメカニクスの一部に気付かずに捕まるかのいずれかを余儀なくされます。プレイヤーがキャラクターをコーナーの近くに置いて開始することを選択せず​​、代わりに近接戦闘のアップタイムのためにアリーナの中央で作業しようとすると、意思決定の時間が 1 秒以上消費され、初めてプレイするプレイヤーと同じ選択肢しか残されません。プレイヤーがオーバーラップ メカニクスが表示されたときに反射的にそれを見ると、人形をチェックする計画全体が台無しになる可能性があります。

これは最初の行進の場合だけです。後の行進ではさらに悪くなります。

オーバーラップ メカニクスは、人形の餌のためにキャラクターをアリーナの中央に保持すると失敗するため、組み合わせを成功させるには、その期間中にキャラクターをアリーナの端に少なくとも適度に近づけておく必要があります。人形の方向転換点はアリーナの端に沿ってほぼ均等に配置されているため、プレイヤーがターゲットとするプレイヤー キャラクターを選択するときに、一部の人形がプレイヤー キャラクターの近くにあることが保証されます。また、プレイヤー キャラクターが中心に近づくほど、最も近い人形をチェックする時間が稼げるため、最も遠い人形をチェックする時間は短くなります。

ただし、後の行進では、プレイヤー キャラクターがアリーナの中心からどのくらい離れているか、それに応じてプレイヤーが目の動きをどのように分配するかについて、多くのケースを考慮する必要はありません。成功するには、16 体の人形すべてをチェックする必要があり、1 秒あたり 1.5 回の目の動きで必要な停止を行いながら各人形を個別に見るには、10 秒と 3 分の 2 秒かかります。これは、すべての人形がアリーナを横切るのに十分な時間です。プレイヤーが選択した順序に関係なく、すべての人形の詳細なパス情報を取得することは現実的ではありません。

そのため、プレイヤーは周辺視野の動きの感度を使って人形の広範囲な動きを追跡せざるを得なくなり、関心のある特定の場所、通常は自分のキャラクターの足に近づくと思われる人形だけを直接見ることになります。(これは、プレイヤーが計画できる強いパターンを提供しない弾幕ゲーム/レベルでは自然な戦略です。) つまり、プレイヤーは人形が方向転換点から十分に離れて周辺視野で選択した場所への動きを感知できるまで移動した後、または見ている場所に十分に近づいて中心視野ではない視野で具体的に確認できるまで移動した後でのみ、見る人形を選択します。前者の場合は十分に反応的です。これは、人形がプレイヤーの選択した場所に向かってすでにかなりの距離を移動していることを意味しますが、アリーナ内の人形の密度により、プレイヤーは周辺視野のみに基づいて確実に安全地点を選択することはできません。後者の場合を待つか、選択した場所から目を離して人形を直接見る必要があります。

プレイヤーがキャラクターの足元を見続け、人形が視界の中心から外れたところに入るのを待つと、その基準でキャラクターをどこに移動させるかを決める時間が足りません。各人形は実際のサーバー位置より後ろにあることを思い出してください。最初の行進で説明した回避成功の要素はすべて適用されますが、時間は大幅に短くなり、他の人形は周辺追跡に格下げされます。プレイヤーがその前に人形を見ると、その人形に関して決定を下すのに十分な時間ができますが、キャラクターの足元から目を離し、その間に別の人形を完全に見逃す可能性があります。

ああ、アリーナの一部はボス自体によって視覚的にブロックされます。そして、プレイヤーは同時に重なり合うメカニクスを処理する必要があり、これはさらなる視覚的および意思決定タスクです。

人形は本当に反応的です。プレイヤーが必要な視覚情報を入手し、反応時間を圧迫せずに決定を下すのに十分な時間はありません。たとえその情報を入手できたとしてもです。デザインを確定する前に、ましてやリリースする前に、メカニクスの各部分にかかる時間を分解して、簡単に修正できるデザイン段階でこれらのことを発見できるようにします。既存のメカニクスと非常に似ているように見えてもまったく同じではないものでも、分解するようにしてください。さまざまなレベルのコントラストと視覚的明瞭さを備えたテストアートライブラリを用意して、大きな変更が必要な場所を示す早期の内部プレイテストを行うことができます。

6番目に、人形に当たったプレイヤーキャラクターの1人が頭を落とすと、近くの他のプレイヤーキャラクターにも当たる可能性があります。そのため、プレイヤーは確実に成功するために人形と重なり合うメカニズムを追跡する必要があるだけでなく、近くのプレイヤー キャラクターや他のプレイヤーのメカニズムの失敗によって生成された落下する頭も追跡する必要があります。これは視覚的な負担をさらに増やすだけでなく、人形を追跡するという視覚的な負担により、落下する頭はアリーナの床にぼやけて見えなくなります。そのため、プレイヤーは、自分のキャラクターが当たっていない人形から来る、見ていなかった落下する頭によって不当にペナルティを受けます。

The Copied Factory では、ピンクのオーブは触れると近くのすべてのプレイヤー キャラクターに爆発しますが、少なくともプレイヤーはそれらが来るのを見て、すべてのオーブから遠く離れることができます。(これは、近くの各オーブの近くにいるすべてのプレイヤー キャラクターを追跡するよりも視覚的に簡単です。) そして、その場合でも、アップタイムの理由でうまく回避し、他の誰かがオーブを爆発させることでペナルティを受けるのは依然として面倒です。

7 番目に、プレイヤーのキャラクターが人形に当たったときのペナルティは特に注目を集め、不快で、プレイヤーの戦闘への参加を妨げます。最初の拘束により、プレイヤーはキャラクターがさらにダメージを受け、ベノギン状態になるのをただ見守るしかなくなり、その後、アクション外とヘビーの制限により、10 秒間キャラクターの動きが遅くなる以外は何もできず、さらに人形の頭と戦闘のセリフで恥をかかされます (そして、キャラクターがすべての人形を避けたとしても、プレイヤーは後者の 2 つのペナルティを受ける可能性があることを覚えておいてください!)。脆弱性アップはより長く続くかもしれませんが、ヒーラーにとってもペナルティの厄介な部分ではありません。

このメカニズムが確実に実行可能であれば、シャープ ペナルティで問題ありません。一撃 KO も同様に混乱を招き、比較的簡単なメカニズムの通常の任務で問題なく使用されています。 (ハイエンドの任務での使用に加えて、任務の難易度が上がるにつれて、中程度の簡単なメカニズムからほぼすべてのメカニズムに徐々に上昇します。) しかし、このメカニズムは確実に実行できるものではないため、このような厳しいペナルティは、その煩わしさを増します。ペナルティがこれほど厳しいものであれば、The Copied Factory のオーブに対してもかなりの苦情が寄せられていたでしょう。

8 番目に、このペナルティの性質は、パーティーにとってさらなる機械的な故障や困難に簡単に連鎖します。バインドにより、プレイヤーのキャラクターが重複するメカニズムにヒットし、KO されていなくてもさらにダメージを受け、脆弱性アップがスタックします。Benoggined のヘビーにより、さらなるヒットを回避することがより困難になります。また、キャラクターが Benoggined されている間、プレイヤーは軽減、回復、および/またはレイズでパーティー メンバーを助けることができず、苦戦している仲間のプレイヤーを助ける機会を失います。

(はい、ベノギネドは累積しませんが、ペナルティの連鎖的な性質を実際に緩和するわけではありません。一撃でKOすると、後の人形もプレイヤーキャラクターの真上を走り抜けることになります。)

9番目に、ペナルティの性質上、戦闘が長引くことになります。ベノギネドはプレイヤーキャラクターがボスにダメージを与えることを直接阻止し、連鎖的な失敗は間接的にパーティーのダメージ出力を減らす可能性があります。その結果、プレイヤーがこのメカニズムに苦労すればするほど、平均して戦闘にかかる時間が長くなり、イライラに対処する時間が長くなります。さらに、これによりパーティーメンバーは戦闘を長引かせたことにイライラし、プレイヤーの身体的限界についてはどうすることもできない社会的摩擦を引き起こします。

このメカニクスを救済するには、これらすべての問題を解決する必要があり、人形とボス全体の完全な再設計が必要になります。人形の行進は、人形の王としてのこのボスのコアメカニクスだからです。このメカニクスの設計意図についてはほとんどわからないため、代わりに、うまく機能している既存の移動オブジェクトのメカニクスと比較し、なぜそうなのかについて説明します。

The Grand Cosmos では、5 つのほうきがゆっくりと移動するため、プレイヤーの画面は実際の位置と比較的正確になります。ほうきは大きく、適度なコントラストがあり、アリーナの床に明るいレーン マーキングを提供するタイルと組み合わされているため、プレイヤーは周辺視野を使用して各ほうきの位置を特定し、レーンに合わせることができます。 (アリーナはちょうど 20 タイルの深さで、5 つのレーンに分かれています。また、ほうきも交互に出現するため、プレイヤーはそれに気づけば、片側のほうきの数を数えることができます。実際、ほうきは戦闘中ずっと敵対リストに残り、その永続性を強調しています。) ほうきの動きが始まる前に、プレイヤーはほうきと土の水たまり (その上を通過するほうきは停止して掃き取る) の両方を確認するのに十分な時間を与えられており、これがほうきの進路とタイミングを知るためにプレイヤーが必要とするすべての情報です。プレイヤーは必要に応じて反応できますが、計画を立てる機会もあります。不意を突かれた場合、それは計画不足から生じる正当な不意打ちであり、次回はそれを変更することができます。 (実際、ほうきの動きに合わせてアップタイム戦略を練ることは、スキルを表現する楽しい機会となります。ほうきが動くたびにアリーナ全体が覆われるため、プレイヤーは成功するためにはキャラクターを動かさなければなりませんが、いつ動かすかを計画することができます。) 重複するメカニズムは、不合理な視覚的要求を生じさせません。また、ほうきが失敗した場合のペナルティは注目を集めるスイープですが、短時間で、プレイヤーは戦闘に完全に参加し続けることができます。これは、エキスパート ルーレットの繰り返しに耐え、MSQ で要求されるダンジョンに十分なほど穏やかな楽しいメカニズムです。

The Jade Stoa のすべてのイテレーション (通常、エクストリーム、そして現在の Unreal) で、Byakko は Unrelenting Anguish を使用し、アリーナの真ん中に立ち、自分から動くオーブを放出します。床にはレーンはなく、プレイヤーを支援する恒久的なほうきもありません。オーブは中程度の速度で動きます。しかし、彼のオーブは固定された螺旋パターンで、アームの間隔が広く、ノーマルの最初のフェーズでは 3 本ではなく 2 本のアームを使用して、メカニクスを穏やかに導入しています。床に対するオーブの高コントラストと組み合わせると、プレイヤーは必要に応じて広いスペースで周辺視野だけで完全に回避できます。これが近接攻撃のアップタイム戦略です。キャストのアップタイムについては、プレイヤーは代わりにキャラクターをアリーナの端の近くに維持し、迫り来る螺旋アームだけを直接見て、キャラクターをスライドキャストする隙間を特定する時間を稼ぐことができます。両方の戦略は楽しいものであり、重複するメカニクスと組み合わせると視覚的に合理的です。そして、繰り返しますが、失敗した場合のペナルティは、ダメージと脆弱性アップスタックだけです。

これらの成功した例の両方で、プレイヤーは移動するオブジェクトの位置を見つけてその経路を知るだけでよく、移動するオブジェクトの向きは使用されないことに注意してください。向きや経路の情報を得るためにオブジェクトを注意深く見る必要があるため、視覚的な要求が大幅に増加します。おそらくこれが、弾幕シューティングゲームが、人形とは明らかに異なる、形だけで向きがはっきりしている高コントラストの弾丸を好む理由であり、プレイヤーが周辺視野で大まかな形と大まかな動きを素早く組み合わせられるようにするためだろう。

邪悪な企み

アリーナの中央からエクサフレアの標準的な「花」を放ち、すべてのカーディナルがすべてのインターカーディナルより先にスタートするか、その逆になります。他の場所では優れたメカニズムでしたが、ここでは人形と組み合わされており、プレイヤーにとってもう 1 つの厄介な存在になっています。

迫りくる悪夢

もう 1 つの標準的なメカニズム、追跡 AOE。

これらは人形と組み合わせると特に厄介ですが、その理由は 3 つあります。まず、動き続ける必要があるため、人形を視覚的に確認することが難しくなります。プレイヤーは 1 つの場所ではなく、パスの周囲を確認する必要があります。2 番目に、AOE がパーティーの邪魔にならないようにこれを行う自然な方法は、アリーナの端で作業し、プレイヤーのキャラクターを人形の方向転換点のすぐ近くに配置することです。3 番目に、プレイヤーのキャラクターが人形に当たった場合、追跡 AOE がバインド中に追いついて、繰り返し攻撃します。

叫び

もう 1 つの標準的なメカニクス。床全体を 2 つの円錐の 3 つのグループで覆う円錐 AOE シーケンスです。これは人形と重なるメカニクスの中で最も単純なものですが、人形の存在によって再び台無しになっています。これにより、1 回の Spirited Charge から多数の人形の行進のシーケンスが開始されます。

Falling Nightmare (繰り返し)

これは、2 ラウンドで、各ラウンドでランダムに配置される追加のヘッドがいくつかある、以前とほぼ同じメカニクスです。ヘッドのコントラストが良ければ、このメカニクスは問題ありませんが、キャスト プランニングには適していません。

ただし、戦闘のダイアログ「キャラクターを壊して申し訳ありませんが、抵抗をやめる時が来たのではないでしょうか。私の忍耐にも限界があります。」が付属しています。メカニクスが楽しいものであれば、このダイアログは面白いでしょう。この時点で、プレイヤーの忍耐は人形によって繰り返し試されているため、イライラします。

全体的な印象 (ボス)

人形の使用は反応的な大失敗で、他のすべてのメカニクスは人形王の伝承のコンセプトを表現していません。ボスが他の標準的なメカニクスの複雑さを通じてそのコンセプトを表現するのは問題ありませんが、全体の組み合わせが合理的である必要があります。ここでは、人形だけでは合理的ではありません。

最初のボスと2番目のボスの間のモブ

2つのパック

ここではメカニカルに興味深いものはありませんが、ダブルプルできます。

あと 2 つのパック

これらのパックの間には、パーティーが通り過ぎると倒れる観覧車があります:

https://i.imgur.com/M3gIIVV.png

具体的には、プレイヤー キャラクターが観覧車に十分近づいた直後に、プレイヤー キャラクターがスプリントをアクティブにしている場合は、そのプレイヤー キャラクターが走り抜けるのに十分な時間よりもわずかに長い時間で、観覧車が倒れます。

初めてのプレイヤーにとって、これはちょっとした「落とし穴」です。なぜなら、これからやってくるモブに集中している間は、道の脇を見る理由がないため、最後に短時間点滅するテレグラフまで観覧車の存在に気付かない可能性が高いからです。

そして、プレイヤーがこの AOE について知ったら、選択を迫られます: 観覧車より先に行けるように自分のキャラクターをパーティーの先頭に留めておくか、観覧車が倒れるまでわざと待つか、わざと攻撃を受けるか。これらの選択肢はどれも良くありません。観覧車より先に進もうとするには、パーティの相対的な位置を判断する必要があり、クライアントとサーバー間の往復によってプレイヤーの画面が不正確になることが、この判断に大きく関係しています。待機時間自体と、パーティメンバーが先に突撃してくるところまでキャラクターが到達するまでの遅延の両方から、待機はプレイヤーを戦闘から引き離します。そして、わざと攻撃を受けるのは気分が悪いだけです。

ゴミの引きに影響を与える環境機能は楽しく興味深いものですが、この実装はそうではありません。

代わりに、回転する観覧車がこのセクションの最後のパックの後ろにあり、パーティがモブと戦っている間にサウンド効果とともにゆっくりと不安定になると仮定します。これら 2 つはどちらも、プレイヤーが観覧車の AOE よりかなり前に観覧車を見るように促します。そして、AOE がエリア全体をカバーするのではなく、観覧車が減速して倒れると回転テレグラフが減速して、スポークの間に安全地帯を残します。 (これも物理的に正確です。ジャイロスコープは回転が速いほど安定します。) こうすることで、プレイヤーはモブの基本的な AOE と環境機能の AOE の組み合わせを体験できます。これはそれほど難しくはありませんが、引き寄せを面白くするのに役立ちます。

2 番目のボス - ジャックインザポット

https://i.imgur.com/xwIbuD8.png

厄介なティーカップ

アリーナの周りにティーカップを設置し、ゴーストがそこに入ることで、大きな至近距離の AOE を作成します。ゴーストのターゲットは、それぞれが選択したティーカップにつながり、つながっている間、ティーカップの紫色の煙に白い煙を追加するため、わかりやすくなっています。

(最初のキャストにはティーカップの外側のリングとゴーストが 1 つだけあります。その後のキャストにはティーポットのリングが 2 つとゴーストが 2 つあります。すべてのリングには均等に間隔を空けたティーカップが 4 つあります。)

これは、アクティブ グローイングに頼る必要がない、優れたセットアップ キャストの例です。アリーナの床が多色であるにもかかわらず、テザーは効果的で、白い煙が紫色の煙と対照的です。

ティー アワール

ティーカップがアリーナの周りを回転し始め、ティーカップが止まるとすぐに AOE が発生します。(ティーカップのリングが 2 つある場合は、リングが互いに入れ替わることも含まれます。) ゴーストはティーカップを満たす紫色の煙の中に隠れているため、これは視覚追跡メカニズムであり、プレイヤーはティーカップの動きを追跡する必要があります。複数のゴーストに対処する場合は、プレイヤーが視覚的に追跡することを選択したティーカップ/ポイントに対する結合された AOE の形状を記憶するメモリ コンポーネントもあります。

純粋な角度運動では、すべてのティーカップが同じ速度で回転し、すべてが同じ回数入れ替わります。これらの事実は、1 つのティーカップまたは 1 つの固定位置を中央視野に保持しながら、周辺視野でティーカップを数回観察することで明らかになります。これは、初心者プレイヤーにとって自然な戦術です。次に、これらの事実を知ることで、プレイヤーはこれらの視覚方法のいずれかが効果的な解決策を生み出すことを確信できます。プレイヤーはアリーナの周りでティーカップ 1 個を追いかけて、それに関連する複合 AOE 形状を覚えたり、アリーナ内の 1 点を見て、そこを通過するティーカップの数やリングの交換を数えたりすることができます。この柔軟性はありがたいです。

このメカニズムは問題なく実行できますが、ティーカップが回転し始めるとアリーナ上空に煙が脈打つのは好きではありません。面白くも面白くもありません。実際、このレビューのビデオを見るまで、あまり気付きませんでした。むしろ、神経学的テストの一部のように感じられ、プレイヤーの視覚処理にストレスを加えるための手段にすぎません。(アリーナの変形がプレイヤーにとってそれほど不快でないように、同様のテクニックを使用するのには理由があります。) 神経学的テストは、一般的にゲームプレイをうまく行いません。

また、ティーカップの比較的均一な回転が、一部のプレイヤーに乗り物酔いを引き起こす可能性があるのではないかとも思います。私はこれで乗り物酔いすることはありませんが、他のプレイヤーがそうなっても驚きません。対照的に、Eden's Promise: Litany (E10N) とその Savage バージョン (E10S) のジャンプする犬は、アニメーションと個々のジャンプから、アリーナの床ではなく犬が動いていることが非常に明確です。

この 2 つの点を除けば、このメカニズムは問題ありません。これは、このボスによって頻繁に繰り返されるため、良いことです。

Toiling Teapots

ティーポットが静止したカップに溢れるほど注いでお茶を残していくというテーマの、拡大する円形の AOE 水たまりのシーケンス。比較的遅いため、水たまりは機能します。最初の AOE に回避すると、水たまりの DOT を受けることになります。

これは、ボスの伝承コンセプトを表現するためのメカニズムの選択でもあります。

全体的な印象 (ボス)

特別なことはありませんが、大丈夫です。マスキングの煙のパルスを取り除いてください。

2番目と3番目のボスの間のMob

2つのパック

ここではメカニカルに興味深いものはありませんが、ダブルプルが可能です。

さらに2つのパック、大きな人形付き

これは興味深いダブルプルです。最後のパックを形成する2つの大きな人形は、通常の基本的なAOEに加えて、軽くずらしたレイドワイドを実行します。これにより、プレイヤーは緩和ツール、AOEヒーリングツール、さらには「レイドワイドを半分にするために最初に人形を1体殺す」などの戦略を検討するように穏やかに促されます。

通常のダンジョンではこれより厳しくすることはないでしょう。かなり弱いヒーラーは、パーティーメンバーの怒りを買うことなく、最低限のアイテムレベルで任務を完了できなければならないからです。

3番目のボス - トロイメライ

https://i.imgur.com/nDcRCOr.png

苦い後悔

アリーナの中央または側面にヒットし、ハデスのスタイルで鮮明な光るアニメーションが流れます。それ自体はシンプルですが、後でシーケンスで戻ってきます。エッジは床の線とうまく一致し、キャストの最適化に役立ちます。

ポルターガイスト

ポルターガイスト テザー セットアップ用の短い純粋な警告キャスト。プレイヤーに何かに備えるよう合図するのに適しています。最初のメモリアル マーチの前にのみ表示されます。

メモリアル マーチ (壁 + ポルターガイスト)

ゴーストの形で通り抜けることができる壁をアリーナに配置します。ポルターガイストは、プレイヤー キャラクターごとに 1 体ずつ、アリーナの端に近い位置を選び、2 つ目の壁が視覚的に下がり始めると現れます。明確なテレグラフのある狭い AOE は、プレイヤー キャラクターをアリーナの 4 分の 1 に確実に誘導し、必要なゴースト フォームを活用できるようにするために効果的に機能します。

ポルターガイストが現れると、近接テザー ブラストをキャストし始めます。プレイヤーが戦闘のダイアログに気付いていなくても、壁のせいでキャラクターを移動させてテザーを通常の方法で伸ばすことができないことにすぐに気付くでしょう。これは、ゴースト/肉のプラットフォームと (正しい) 青いテキスト「床の一部が光り始めます!」のアクティベーションに最適なタイミングです。プレイヤーは、このことに気付いた後にプラットフォームの 1 つに行くと、壁を通り抜けて通り過ぎる時間があまりないため、最初は正しく理解できないかもしれませんが、情報に向かって押し進められ、成功する方法を見つけます。

(このテキストが赤ではなく青なのは良いアイデアです。危険の警告ではなく、プレイヤー キャラクターへの支援について語っているからです。もちろん、プレイヤーが色の違いに気づくことを期待することはできませんが、ここでは気づく必要がないので、ちょうど良いです。)

これは、1 人のプレイヤーが正しく実行すると、自動的にパーティー メンバーに解決策を示すようなメカニズムでもあり、これにより、初めてプレイするプレイヤーは、キャラクターを通常通り逃げるときに必要な情報がどこにあるのかわからなくても、理解しやすくなります。このような状況では、完全なパズルではありませんが、通常の任務で少し直感に反するメカニズムを使用できます。

テザーは十分に伸びていない場合、それほど大きなダメージを与えず、中程度のダメージ増加と脆弱性アップ スタックを使用して失敗を示すのは良いことです。これにより、初めてプレイするプレイヤーは、メカニズムを初めて学習するときに、1 回か 2 回の失敗の可能性に落胆することがなくなり、後の重複と組み合わせが全体としてあまり罰せられなくなります。

ゴーストダスター / フレッシュバスター

視覚的にはこれらは拡散 AOE ですが、キャスト名と赤いテキストは拡散とは関係がないことを示しています。(テキストはそれぞれ「トロイメライは霊に対する防壁を準備します!」と「トロイメライは生者に対する防壁を準備します!」です。これらは毎回表示されます。)

キャスト終了時にプレイヤー キャラクターが反対の形態になっている場合、プレイヤーはダメージをまったく受けず、「回避」のフライング テキスト、頭上の緑のチェックマーク、プレイヤーへの「ディン」サウンド エフェクトが完備されます。キャスト終了時にプレイヤー キャラクターが同じ形態になっている場合、プレイヤーは即座に KO され、「9999999」(実際のダメージ) のフライング テキスト、頭上の赤い X、プレイヤーへの別の「ディン」サウンド エフェクトが完備されます。これは、他のほとんどのメカニズムと比較して、両方の可能な結果が比較的極端であるため、プレイヤーがメカニズムを学習するのに役立つフィードバック ツールの適切な使用法です。

(これらの最初のものは、ゴースト形態が必ずしも好ましいわけではないことをプレイヤーに教えるための、常に Ghostduster です。最終的にはランダムになります。)

キャストの長さは適度に長く、プレイヤーは最初の位置が悪くてもキャラクターを変身プラットフォームに移動できます。これは、失敗に対する厳しいペナルティを考えると重要であり、正しく行われます。

このメカニズムについて私が変更したいことの 1 つは、スプレッド マーカーの代わりに、汎用の「何かにターゲットを絞った」マーカーを使用することです。スプレッドはこのメカニズムにはまったく関係ありません。パーティー全体がスタックしてもダメージを受けません。汎用のターゲット マーカーは、その繰り返しの形と独自のサウンド効果を考えると、十分に明白です。

メモリアル マーチ (AOE シーケンス)

ボスはアリーナの前方に 2 組のゴーストを出現させます。どちらもテザーで繋がれており、列はセット間で交互に配置されます。これらは、ボスが Bitter Regret をキャストしている間、同じ順序で解決するライン AOE を作成します。このキャスト中にテザーを表示し、プレイヤーにゴーストを見る時間を与えることで、このシーケンスが機能するようになります。これにより、解決するゴースト AOE ではなく、Bitter Regret アニメーション テルを見ることができるからです。

メモリアル マーチ (組み合わせ)

壁とポルターガイストを他のものと組み合わせます。近接テザーと同時に 2 つの象限を攻撃する前方のゴースト セット、Bitter Regret、Ghostduster / Fleshbuster などがあります。プレイヤーがゴースト フォームの使い方を理解すれば、これらすべてを管理でき、さまざまな最適化の機会が提供されます。 (たとえば、ゴーストが 2 つの象限に当たると、プレイヤーは自分のキャラクターを近接範囲内に維持し、ゴーストの前にキャラクターを配置してコンビネーションに成功することができます。こうすることで、ポルターガイストがゴーストの近くに出現し、キャラクターを両方から逃げさせることができます。) これは非常に歓迎すべきことです。

各要素を個別に最初に導入しなければペースは速かったでしょうが、そのように導入されているのでうまく機能しています。

ゴーストクラッシャー

壁が上がっている間にライン スタックすると、パーティはスタックするために壁に突っ込むように促されます。これはうまく機能し、良いことです。奇妙な名前ですが、ゴースト フォームのキャラクターにペナルティを与えませんが、キャストの長さが十分ではないため、煩わしくならずにペナルティを与えることはできません。

全体的な印象 (ボス)

ストームブラッドや Shadowbringers のダンジョンに簡単に登場したであろう楽しいボスです。ゴースト/肉体の変身を巡って戦略を立てる機会は、エキスパート ルーレットで非常に歓迎されています。

全体的な印象 (ダンジョン)

人形がこのダンジョンを台無しにしています。これは「たった 1 つの悪いメカニクスで任務が台無しになる」という典型的な例です。ここには実際に悪いメカニクスが 1 つしかありませんが、その 1 つのメカニクスがあまりにもひどいため、若くて健常なプレイヤーでさえ簡単にイライラしてしまいます。

----------------

あなたの「実験的」ダンジョンはどちらも、実行不可能な視覚的要求に悩まされています。ハイデリンの愛のために、メカニクスの時間的プレッシャーを考慮した視覚的明瞭性とアクセシビリティの標準を採用して従ってください。