Paladin doesn't suck, 2 minute rotations just ruined combat

[Mikey_R]

Not sure where you're trying to go with this whole "choice doesn't exist" bit....

Again, choice isn't opposed to there being something optimal; it's simply dependent on varying situated frames of reference. Best for the next GCD? For the next string? For sync into the next macrorotational phase? For the fight as a whole? Each requires increasing levels of knowledge that eventually approach a gamble (or, a controlled exchange of risk for reward).

For one to assuredly make the best choice overall, you'd need to know, at minimum, precisely how long the fight will last, up to the points that would threshold which choices would be optimal (far and few between in XIV, but as we've noted, it's not exactly an example of a place with frequent decision-making). For anything party-dependent or party-affecting, you'd also account for said party, up to the weight that'd affect those thresholds in turn. Which, you won't. You'll have a best guess, improving in accuracy and precision with skill and familiarity, but never a perfect one.

Except it is. Lets go through your list, next GCD? There will be an optimal one. Next String? There will be an optimal one. For the next macrorotational phase? That is something decided in your raid party so that ou can keep the 2 minute buffs lined up. The fight as a whole? Nothing you do at the start of the fight is going to affect you that far into a fight, especially not something that you can effectively make an informed decision on, which makes it a non issue.

Monk is far from I was describing above, so I'm not sure why you're mentioning it as if it were evidence that choice can only ever be illusion. Monk manages string lengths not for varying combos but simply to <minimize wasted TwS/Demo duration> while also trying to <minimize wasted CD potential effective uptime>. The combos themselves aren't the decision; the TwS/Demo string-lengths and PB timings are.

When you mention doing either AAA or ABA or BBA etc. that is exactly what Monk's rotation is. You are choosing what to use in each step. Monk's place in the combo is denoted by the form it is in. That is why I compared them as it is exactly what you described.

The actions you do is decided by the TwS/Demo timing for you. You decide to do True Strike because you have enough time left on your Twin Snakes, you Snap Punch BECAUSE you still have plenty of time on Demolish. These choices are made on a GCD basis. The same is true for Dragoon, you decide to use a specific combo based on Powersurge/Chaotic Spring timings. You have the choice of one or the other, however, based on the info, you decide to one one over the other.

This is the same with every job. You have a choice of using any Ninjutsu, however, you use Raiton for single target, unless you are under Kassatsu. You have a choice of an order for Ten Chi Jin, however, you use them in a specific order based on single target or AoE. You have the choice to use Triple Cast now, or you can save it for 10 seconds for the heavy movement phase, that That Thundercloud proc? Useful for a weave. Summoner's Summons? You have the choice in what order you put them in, but you aren't going to use Ifrit when you know a heavy movement phase is coming up etc. All of these are choices and all of them have have an optimal route to go down.

The absolute closest I can probably think of is old old Tornado Kick, when it used to take away your GL stacks. You had the choice to use it just before the boss jumps, but, when is that? and if I used it and I could get an extra GCD in, I lost damage, so really, the risk wasn't worth the reward. Now, you could increase the reward, however, that just makes that option the best and worth going for, so you really don't have a choice there either. I'm using TK as an example, however, this will be true for any risk/reward system. If the risk doesn't outweigh the reward, then there is no point and if the reward far outweighs the risk, then it just becomes part of your rotation.

You can manipulate the variables/factors in chess, too, but doesn't devolve the game into non-decisions. Again, what's key is that there isn't a singular APL / frame of reference for priority.

First, to point out a fundamental difference between FFXIV and Chess/fighting games, you go against a human in Chess/FGs, which you are deciding one decision based on a web of possibilities. FFXIV, being against an AI, there is always going to be an optimal route. Even if you mix up the boss rotation randomly fight to fight, there will still be an optimal way to deal with the encounter.

If the entirety of a piece of content follows the exact same APL at all times, then it hasn't met the constraints you just quoted.

But that is the point, there is no choice, there is just optimal play based on the encounter. Even if you randomise the boss, it will either be rotation A or rotation B.

Making a "decision" is an act of parsing a (best guess at the) best choice wherein

• the outcomes of available choices are real and varied,
• the method by which the best choice may be considered are themselves multi-leveled (one shifts between meta-APLs or the action priority lists compete/are re-weighted directly) and/or varied, and
• one has access to significant, but not the whole body of, relevant information informing the weights of those methods of determination (and therefore the choices themselves).

I.e., it's one where the "best" choice is dependent not just on immediate context, but also on a reasonably but uncertain guess as to what would best pay off in the future. A truly stagnant, unvaried APL cannot meaningfully render choice.

1. And one will be better, meaning you don't have a choice.
2. But there will still be a best way through, negating that choice.
3. Which sounds like progging a fight, since you don't have all the info, and so, you are still learning what the best way through the fight is. eventually, you will find that way, negating all the choice.

But there is nothing uncertain. Even random bosses cannot be truly random otherwise you will screw over groups. The boss can randomly TB every 15 seconds without limits? Your tank's are going to die once they run out of cooldowns, which means there is a cooldown period. The same will be true of anything a boss can do that gives you that choice of A or B. If you have to use option A, then the chance for option A comes up again, you use it, unless it is on a cooldown, then you used it earlier, so tough luck, you don't get it here. Even if you delay, it means you miss it the first time, either way, you missed that chance. You were not given a choice.

If you want to show me a specific example, you need to tell me what the choices are, what the boss is doing to warrant that choice, and how that choice is going to affect you later on in the fight. Nothing vague, I want specifics because so far, everything you have said is not a choice in the moment, but the illusion of choice.

That's called "going through the motions."

Call it what you want, doesn't change anything.

If that were true, people wouldn't absolutely despise doing FATEs or feeling like they "must" engage in non-dungeon/raids combat. But most people will comment negatively about that kind of content, because rotations are so mindless now. If you don't have mechanics to learn and react to, there's very little actually going on in XIV's combat.

Miss the bit where I mentioned you need engagement on both rotation and boss mechanics to prevent something from feeling dull? What do FATEs and the vast majority of dungeons lack?

[Shurrikhan]

You decide to do True Strike because you have enough time left on your Twin Snakes, you Snap Punch BECAUSE you still have plenty of time on Demolish. These choices are made on a GCD basis.

Except even that procedure varies with upcoming CDs unless constraining yourself to a fixed loop macrorotation (which underperforms).

There are points at which the answer to "what's best for the next GCD" diverges from that of "what's best for the next rotational string" which may further diverge for "what's best for the next burst window" which may, though rarely, be different from that of "what's best given <estimate> seconds left in the fight".

That's where it picks up a degree of decision. There's more than a single level of calculation and there is at least a space for (though it will be fight-dependent in practice) an element of gambling.

Yes, any such examples are relatively shallow and incredibly few and far between in this game at present. But they are not theoretically impossible, as you are making them out to be.

Except [choice is opposed to there being something optimal within a frame of reference, even if there are multiple frames of references that give competing answers from which no best frame can be selected with complete certainty]. Lets go through your list, next GCD? There will be an optimal one. Next String? There will be an optimal one. <snip>

That's true only when...

1. each question provides the same answer (meaning that you have only a rigid and unvariable sequence of events / in-combat complexity only in execution with all else being decided pre-combat),
   and/or
2. you have unlimited knowledge and mastery.

Just as there is a time where one is still getting execution down, that period while it's still a form of engagement, there's a period before which one has even a majority of the frames or criteria by which to determine optimal action. Moreover, there is no perfect answer for all but the most practiced of speedruns with a known group for something like "best for the given fight length" unless, again, the kit is too shallow for those different questions to be differently answered.

But hey, if you want to go the full nihilist-gamer route of "Choice can never exist so let us diversify play only by how well we can execute on a single set sequence"... all the power to you? I just want no part of that. There's too little decision-making in the game as is. As difficult as decision-making is to contextualize sufficiently to make those choices real, I'd rather allot resources towards that than be satisfied just with rigid execution.

[Gserpent]

Miss the bit where I mentioned you need engagement on both rotation and boss mechanics to prevent something from feeling dull? What do FATEs and the vast majority of dungeons lack?

Yeah, seems I did. It sounds like we're in agreement. FWIW, I've enjoyed reading the back and forth between you and Shurrikhan. Good stuff here.

[Mikey_R]

Except even that procedure varies with upcoming CDs unless constraining yourself to a fixed loop macrorotation (which underperforms)...

Yes. I have never said you cannot change what you do based on a fight, however, you have changed what you do to maximise your effectiveness. Once this is figured out, you aren't going to go back to the old way as it is inferior to what you were doing. You had the choice to change, but once you decide that is what you are going to do, you aren't going to go back and pick the other option. Once that choice has been made, it no longer becomes a choice for subsequent encounters.

Let's go about this in a different way. lets make the assumption there is someone who has perfect knowledge on their job, but they cannot predict what is about to happen. At all points in a fight you have a choice as to what to do, ranging from the best case DPS to absolutely absurd things, like /playdead. We can rank these in some sort of order based on how much it helps in a fight in that situation:

[A,B,C,.....,Y,Z]

Now, as a general question, I give you a marker. You can place this anywhere on this list of choices. Everything to the left of the marker is something that could be a valid choice, on the right, everything that is not. /playdead isn't going to help, using Throw Dagger isn't going to help when still at melee etc. Fo rthe vast majority of the time, you will have 1 choice, Chaotic Spring after Disembowel, followed by Wheeling Thrust, then Fang and Claw etc.

However, once a mechanic comes out, say, the boss becomes untargetable, then you do suddenly have choices. Your decision marker is going to shift. In this hypothetical, say there is 3 potential choices you can make, so, your marker goes after those 3 choices :

[A,B,C,|,D,......,Z]

You are going to ignore everything to the right, and are left with choice A, B or C. Bear in mind, we don't know what is going to happen after said mechanic. Mechanic happens, and it turns out, option B would be better, by that, you move the marker to reflect this:

[A,B,|,C,...,Z]

We are now down to 2 potential routes to take, now imagine this perfectly logical person finds A, you move the marker and suddenly, you only have option A to the left of the marker. What started out as a choice has now devolved into this one option.

Now, this was done on the premise that the human was perfectly logical. Obviously, this isn't true in any way, however, a similar things can be done for actual human minds. You see a mechanic for the first time, you decide how to best approach it based on where you are in your rotation and what is going to happen after, and you change what you do to follow that. Your choices become more and more limited based on your knowledge of a fight.

But hey, if you want to go the full nihilist-gamer route of "Choice can never exist so let us diversify play only by how well we can execute on a single set sequence"... all the power to you? I just want no part of that. There's too little decision-making in the game as is. As difficult as decision-making is to contextualize sufficiently to make those choices real, I'd rather allot resources towards that than be satisfied just with rigid execution.

There is nothing nihilistic about what I have said. I am merely saying there is no point where you can be given a choice that cannot just be reduced to an optimal option. I have asked several times for you to give a generic example where such a choice could matter, however, you have yet to provide one. It is even more frustrating when I am trying to use different choices based on known jobs, but you keep coming back with, 'that isn't what I meant'. Provide an example of what you mean. Show me there is a scenario that choice matters, prove me wrong. I want to be proven wrong because that is how knowledge is gained. That is how our understanding grows. Prove me wrong.

[Lustre]

You people care more than SE do it seems

[Gserpent]

Well we can afford to spend all day navel gazing, and we don't have to do any of the actual work. But I do sometimes wonder if Square-Enix needs to assign more veteran oversight to the class design teams to ensure consistency of results and things like that.

[Shurrikhan]

You people care more than SE do it seems

Not sure why that'd shock anyone? There are what, three(?) job devs who regularly fail to imagine optimal play for the jobs they've designed (or, alternately, to imagine in the practical issues that'd have the job perform below what they wanted) and tens of thousands of players each who main those jobs.

I do sometimes wonder if Square-Enix needs to assign more veteran oversight to the class design teams to ensure consistency of results and things like that.

Of all the issues with job design, though, I don't think consistency has ever been a major issue.

That would only make slight changes like PLD being moved from a spread per-minute job to a 2-minute burst (a change probably coming soon) rather than simply being tuned up to make up for its typically lost burst windows (and therefore faintly overpowered in a low-buff comp party like SGE|WHM|SAM|BLM|MCH|+2).

Otherwise, at most certain jobs seem to be treated as early testing sites for certain things, like button consolidation or longer combos, with those things then quickly being spread across all. That fixation on consistency, if anything, is why Fracture and Touch of Death, in SB and ShB respectively, were pruned with most other DoTs for being "bloat" (that also would frequently exceed the XIV's status effect cap, pathetically low for a game with 24-man content) despite their functioning very differently in that context (also situationally as soft-CDs that could be used to deal with inability to position and a means to additional SkS tiers' rotations), etc.

Which perhaps is sadder still, in that the apparent solution to those oversights are people who actually play (at varied and significant levels of optimization) the jobs they design for, which is more staff-intensive than merely adding supervision and coordination. The coordination already seems to be there, with only (apparently minor, to the devs) concerns like identity holding it back for brief spans at a time.

[Shurrikhan]

Yes. I have never said you cannot change what you do based on a fight, however, you have changed what you do to maximise your effectiveness. Once this is figured out, you aren't going to go back to the old way as it is inferior to what you were doing.

There isn't just an "old" way or "new" way, though. Those situations are fluid, as would be your estimation of which situation was current. One was better in situation A. Another was better in situation B. And you had no certainty which situation you would be in. At minimum, therefore you had to not only follow an APL but select the best from multiple APLs based on your best guess of what situation was occurring.

You're not always or simply creating a 180-step rotation and executing it precisely. You are tracking the criteria by which one decision may be favored over another and executing accordingly, each of which adds cognitive load that burdens execution in a way that makes it take that much longer to become wholly subconscious, passive, and thereby unengaged. Something you do wholly subconsciously is not engagement; it at best merely burdens other areas in the hopes that they will be more active or significant forms of engagement.

If you want execution not to be one-and-done, it and its surroundings need, in sum, enough cognitive load for that challenge to actually last a while (perhaps even with some players never getting to where that engagement goes obsolete), and the less surrounds it, the more quickly the execution itself can be focused down and solved, rigidly and finally. Once solved -- once there is a rigidly best solution -- it is done. As done as any cheesed-but-optimal way of handling a mechanic with perfect reliability. At most thereafter rotational execution can only modestly contribute to the fight's mechanics' engagement, just as the mechanics of the hundredth farm speedrun with a static on a new job can at most (for a while) modestly contribute to rotational engagement.

The more meaningful decision-making, the longer engagement lasts. And the difference, especially when starting from such a low floor as XIV's, is huge.

Now, as a general question, I give you a marker. You can place this anywhere on this list of choices. Everything to the left of the marker is something that could be a valid choice, on the right, everything that is not. /playdead isn't going to help, using Throw Dagger isn't going to help when still at melee etc. Fo rthe vast majority of the time, you will have 1 choice, Chaotic Spring after Disembowel, followed by Wheeling Thrust, then Fang and Claw etc.
<snip>
...based on known jobs...
<snip>

Which is because our actions are typically (A) bound in rigid sequences and (B) have usually zero interaction beyond universal damage amp windows. They are literally designed to minimize choice. That is their point.

My entire complaint has been that the game lacks real choices (those being found basically only --and only rarely even then-- on MNK, SAM, and BLM).

You ask me to define a "real choice." I do that.

Your response, then: You demand I point at the real choices in game when my entire point was that there are almost no real choices in this game and that such is why execution so quickly goes stale here despite the game's relatively high button count.


Me: "I'd like some seasoning. Most items have zero seasoning and the remainder, next to none. Which is probably why the meal so quickly tastes bland."

You: "Oh really?! Point at the goddamn spice, John! Point at the goddamn spice! You show me where that spice is significant or so help me god–"

[Mikey_R]

There isn't just an "old" way or "new" way, though. Those situations are fluid, as would be your estimation of which situation was current. One was better in situation A. Another was better in situation B. And you had no certainty which situation you would be in. At minimum, therefore you had to not only follow an APL but select the best from multiple APLs based on your best guess of what situation was occurring.

Except it is. You go into a fight and the boss becomes untargetable just as you apply your DoT, you aren't going to do that next time, might as well get the direct damage in. There is the old choice and the new, updated choice. Bear in mind, all boss fights happen based on the bosses rotation and not time or HP% (unless it is a major phase transition), otherwise, you end up with the old HW thing of 'Skip Soar'. Even in this example of 'skip soar', your party either meets the requirements, or it doesn't. You then have the route through the fight where you skip soar and one where you do not. You haven't made the choice.

If you want execution not to be one-and-done, it and its surroundings need, in sum, enough cognitive load for that challenge to actually last a while (perhaps even with some players never getting to where that engagement goes obsolete), and the less surrounds it, the more quickly the execution itself can be focused down and solved, rigidly and finally. Once solved -- once there is a rigidly best solution -- it is done. As done as any cheesed-but-optimal way of handling a mechanic with perfect reliability. At most thereafter rotational execution can only modestly contribute to the fight's mechanics' engagement, just as the mechanics of the hundredth farm speedrun with a static on a new job can at most (for a while) modestly contribute to rotational engagement.

This is the point, there is no choice. How far people go in terms of optimising is different for each player, but that is a choice separate from any encounter.

You ask me to define a "real choice." I do that.

Not once have you provided an example of a 'real choice'. All you have done is provide a vague statement to the fact. You have not given specifics. It is more akin to someone asking for DPS concept and asking how it will do damage and the answer is, press buttons. Great, I could have told you that, but how do you want the player to press buttons. That is what I am after, the how. How do you want to give players that meaningful choice. By the way this has been going, it also has to be a choice where you cannot just optimise away, something to make you think. Maybe I should have worded it better in asking for what I want. I want the how, not the what.

[Shurrikhan]

<snip>

So, your response here is purely a matter of semantics? Honestly, that just seems needlessly hair-splitting.

That seems the likes of... <Cold doesn't exist (only molecular average speed) and shouldn't be used even in shorthand.> / <We are all fish (or virtually nothing is).> / <There is no 'right' or 'left' side of the computer, only its faceplate or backplate, even if everyone else is plenty able to contextualize which is meant given where the item in question is located.> Etc., etc.

I think you'd have guessed full well that the term is being used as an asymptote. A theoretically perfect player will break even theoretically perfect contexts for decision-making, but you still build worthwhile decision-making for real players. Some will eventually exceed it. Others never will. We need only ensure that all players are at least thus engaged for quite a while, not forever. [/SIZE]

But fine, let us replace the word "decision-making" with "cognitive load sufficient to, for the average player participating in the present content, reach at least a point at which adding further cognitive load no longer creates an outsized (above-linear) effect on the time it takes for a player to render execution of a given sequence into a subconscious action -- typically/ideally situated in junctures where nearly no players could track all the information necessary to make a perfect estimate of the best possible outcome and must instead make a gamble that in turn compounds future criteria."

Will that do the trick?

We've already collectively clarified, repeatedly, the idea that yes, if all players were divine entities, there'd be no "choice," sure, but they're not. They're people. The game need only have enough contextual complexity (from kit, party, or content) to push them out of certainty (even with outside advice/tools) and for those choices to be near and/or interdependent enough for no single answer to be the best an absolute majority or more of the time. XIV doesn't typically have that, but that is the complaint.


The better players get, the more "decision-making" they will leave behind, replacing that instead with relatively passive tracking that informs merely execution, which will in turn eventually becomes entirely passive. My point was that as you have fewer things to shake up fixed sequences in non-fixed way (e.g., not a mechanic that affects player actions in the same way each time), the more efficiently one makes execution passive.

Whereas adding an extra step to execution may make a roughly linear difference to how quickly one learns and makes passive the execution of a given (macro)rotation, adding complicating context to it make a largely quadratic difference. That's what I mean by complexity in (or, approaching) "decision-making" being synergetic with longevity of engagement through execution.

It's also why, for a typical player, choice isn't just what's optimal. Many a Monk, for instance, may opt into an closed loop rotation for a while during early progression despite knowing that it is inferior in order to remain at a level of cognitive load they still find enjoyable. Now, that difference won't be enough to meet that ideal definition of "decision-making", but it's worth keeping in mind.

Not once have you provided an example of a 'real choice'. All you have done is provide a vague statement to the fact. You have not given specifics. That is what I am after, the how. How do you want to give players that meaningful choice. By the way this has been going, it also has to be a choice where you cannot just optimise away, something to make you think. Maybe I should have worded it better in asking for what I want. I want the how, not the what.

You asked earlier for a generic example. The model is that (the qualities that define the genera, literally a generic example). While that constitutes the "what" of a real choice, it is equally "how" a real choice is situated. It is the criteria that makes a real choice a real choice, the quality present in any and all of them. Given that we were trying to look at the bounds of what is possible, that seems the place to start, no?

Given that the "how" (the process or, as you seem to be asking for, the product), and the "what" (likewise either the process or product) differ only in contextual semantics, it's difficult to guess at what level of concreteness would pass muster for you. Again, here in XIV those examples rare --and not great even when they do form-- so I'd need to do a relatively from-scratch mock-up. They also require tremendous context (with precise balance) to actually exist, so it will take time to craft an example.

May I ask how many example moments of decision, and what degree of interrelation, would make the case for you? I'd rather not have to craft a 5-page mock-up.

I'll do for now what I can be relatively sure won't go to waste. The rest will have to wait until later.

An initial tentative attempt:

Normally, there would be three facets of context for complexity and therefore decision-making (or, the portions of cognitive load typically called "decision-making"):

1. the toolkit,
2. content-interaction, and
3. party-interaction.

For brevity, as it is the densest area, we'll mostly ignore the last for now, pretending it synergizes with, or at least does not conflict with, the toolkit and content-interaction.

To be clear, obviously not all are required to make for what I have so far called a "decision". There need only be enough complexity for there to be multiple relevant frames of reference (next weave, next GCD, next string, next burst, etc.) with competing answers to what would be best and for all but the best players to be able to definitely estimate a most productive frame of reference.

In XIV, this complexity has been typically drawn from content-interaction, since party-interaction has been made basically foolproof and automatic (even prior to simplification and consolidation, due simply to either pre-fight configuration as per set comps or the best choice being too far from the next best to be overshot in any situation). Even content-interaction, though, has diminished over the years, with fights behaving increasingly like target dummies, though not always for that great a loss to complexity; they had already rarely ever met the breaking points within a kit precisely enough to offer choice, prefering instead to offer waves of differing actions with a more obviously correct choice, as per interrupting distinctly AoE phases in an otherwise ST fight.)

Of course, all is bottlenecked by the opportunities available to a given kit (a duo of targets means nothing if you have no DoT [combo] and your AoE has no relevant advantage whatsoever until 3+ targets), so I'll try to cover that, too, specifically but generically.

Yes, party interaction and such can also reduce complexity. If you would have done something differently in order to survive but an ally could just throw onto you and otherwise wasted shield anyways, they may get some greater access to complexity in that context, but you will have less. Your would-be choice becomes theirs unless your choice is the more efficient and theirs the backup that is nonetheless better spent on you than someone else. Etc., etc.

As such, let's just assume that if the example below takes a cross-role action like protecting itself from unavoidable AoE damage or briefly gathering or tanking adds, that the party has no competing option of greater efficiency.

Some general factors contributing to complexity that may, if sufficiently balanced, contribute to decision-making. For now, axe those low-hanging fruit of universal undermechanics that would be beyond XIV.

• Stagger Systems (Damage can function also as mitigation and/or --granularly or at certain dynamic thresholds-- damage amps, offering less fixed competing tempos for optimal use.)
  
  
• Resource Managment (Usually requires granular resource effect, such as having less throughput as one has less Stamina. Requires a toolkit balanced sufficiently to allow for burst vs. sustained action. For this to result in complexity, there must be a desync (between the burst/sustained phases enticed by resource management and what the kit would oblige on its own) balanced sufficiently for the best choice to be situational. May or may not include effects for/from %HP that ["artificially"] raise the value of defensive play, especially just before one's own throughput bursts.
  
  
• Behavioral Manipulation (Mobs have, by species, a different behavioral script by which to determine target selection and their own "optimal" action, usually in accordance with base-threat-time-script-multipliers. Mobs may, as an action, change scripts or their threat coefficients may themselves be influenced by existing total threat, unit-specific threat, %HP, lead threat's %HP, etc.)
  
  
• Physics-based Effects (Being able to knock enemies up, back, down, etc., based on action weights, positioning, and mob context. Often granular. Slightly redundant with stagger systems in terms of complexity, but often complementary to overall gameplay feel.)
  
  
• Elemental Effects (Like physical effects, but using instead manipulable spectrums like Heat, Shock, Saturation, Mass. Complementary with behavioral manipulation and physics-based effects, but with some diminished returns regarding complexity.)
  
  
• Factors Affecting Uptime (Boss jumps. Movement requirements. Boss-radial AoEs of greater than melee range. Proximity based AoEs. Almost never offers any meaningful choice in itself but can allow for it if there are sufficient complementary factors.)
  
  
• Factors Affecting Target Count (Add spawns. Almost never offers any meaningful choice in itself but can allow for it if there are sufficient complementary factors.)
  
  
• Factors Affecting Relative Damage-per-Action (See Ravana Red/Green/Blue phases. Sometimes redundant, sometimes usefully competing with party interaction.)
  
  
• Randomization (of the selection, target selection, frequency, and timings of mechanics).
  
  
• "Tethers" / "Baits" (being able to bait mechanics or otherwise force their target selection -- relevant only if they spawn away from the otherwise best positioning for their ideal recipients, at an opportunity cost in movement that neither obliges nor forbids their getting said tethers).
  
  
• DPS/Healing/Mitigation checks (relevant to complexity only when competing with otherwise optimal actions, such as STing during AoE opportunities to burn down a dangerous mob or not using CDs or bankables during a certain raid bursts in order to deal with a DPS check soon after).
  
  
• Unavoidable damage that would situationally force non-free personal defensives. (Technically not in XIV anymore, but we've had them, or near enough to them, for some jobs at some points. Most relevant if they affect the flow of burst phases, but that's pretty much only been a quarter of a real thing in XIV, maybe per Stormblood DRK's TBN if its Blood spenders had carried gameplay-affecting additional effects.)

Okay, so we basically are allowed just DPS checks, boss jumps, AoEs that'd force at-cost movement (so, not a thing for decision making it itself for XIV Physical Ranged jobs), add spawns, tethers, randomization, and maybe painful-but-not-too-painful-AoEs if the kit has a way of addressing those mechanics.

Alright, that's our ceiling. Now, we need a kit that can interact with them. I'm just going to use a DPS as my example kit because those are the most numerous jobs and have the highest portion of independent complexity.

Where that kit gives a clear, straightforward fit to a given feature of a fight, we have only engagement through execution and there is little cognitive load. Where it fits less cleanly, usually via some degree of inertia (an AoE also having a higher degree of guage generation, which is more bankable, or AoE bankables being more efficient relative to their ST equivalents than AoE GCDs are to their equivalents, etc., so a next-GCD decision can affect a capacity for an upcoming mechanic), there's greater cognitive load (approaching a "decision").

• 4 enemies? No conflicting DPS check therein? No decisions, just dance.
• 6 GCDs against 2 adds, between ST dps checks wherein the boss reappears and would nearly wipe the party unless his cast-HP is burned down (and where one's 2-step AoE combo falls slightly short of their 3-step ST combo in ppgcd but has higher gauge generation)? Nearer to a decision.

  One could (A) just AoE spam so the boss arrives on an AoE opener (and can spend bankable oGCDs as AoE or ST, more flexibly), (B) downclock 1 GCD and prep the AoE finisher for the check, or (C) AoE twice and then prep an ST finisher for the boss's brief dps check. The best answer therein will depend on one's party (their gear, attention, bankables), any prior mistakes/irregularities (overspending bankables, simultneous mechanics, etc.), and the severity of the dps check (including whether it's all-or-nothing).

<Ran out of time. Releasing now in hopes for a bit more detail on what degree of concreteness you want. Will be back later.>