Battle System Suggestions - MB/SC, Forced Delay, Enmity

[Byrth]

FFXI's battle system has many mechanisms that did not scale well from 50 to 99. Lets look at them:

1) Skillchains - The skillchain patch helped out, but it made the issue with skillchains even more obvious than ever before. The skillchain window is ~4-10 seconds after the previous WS. It varies a bit by job, but 7-8 seconds is a pretty typical TP-WS cycle time for a melee DD with capped Haste. So a single DD can skillchain indefinitely and benefit immensely from the system, but multiple DDs end up counting to 3 after the previous WS, waiting, waiting, WSing less often and picking up a little skillchain damage. It's still strategically useful to stay wise to the Skillchain system as a DD, but the system itself is cumbersome and could be more practical.

2) Magic Bursts - Magic Bursting off a skillchain (0-10 seconds after the skillchain) results in a 30% magic damage boost (+5% more per skillchain level past 2) and some more damage based on your Magic Burst Bonus gear. For a 6-step skillchain with +53% MB Bonus, you would have a 2.3x multiplier to magic damage along with a huge magic accuracy bonus. So why don't people MB? Well, it turns out that any Weaponskill with a skillchain property interrupts the previous magic burst window. In a situation with multiple DDs spamming WSs every 7 seconds, even if your melee are coordinating for magic bursts, spells do not cast quickly enough to catch the errant skillchain before another melee WSs.

3) Forced (Animation) Delays - After you use a WS, you will not autoattack for the next 2 seconds and cannot initiate any other action. After you use a JA, you will not autoattack for the next 2 seconds, and can initiate another action after 1 second. After you use magic, you will not autoattack for the next 3 seconds and cannot initiate another action during that time. This animation delay makes up a huge portion of player time in the current battle system. In the example above, 2 seconds of the 7 second TP-WS cycle is spent waiting for your WS animation to finish. That's almost 30%. If you're a mage with a good fast cast set, you spend about 60% of your time waiting for the 3 second delay to be up. These forced delays are super annoying and detract from the otherwise fast pace of battle.

4) Enmity - This is really just an honorable mention that I don't particularly feel like dealing with mechanistically in this post, but the enmity system is insufficient. All standard melee DD will cap enmity quickly when fighting a boss monster and then the monster will just spin, becoming untankable. There's essentially no threat that the monster will go after a mage. Hate is perfectly stable despite how much damage the melee are taking. It's boring.



Discussion:
Issues #1 and #2 are due to the 80% Delay Reduction Cap. DDs at the delay reduction cap attack 5 times faster than DDs with no delay reduction. Look at the above and imagine a 35 second cycle time instead of a 7 second cycle time. The top two issues are solved, and action animation delays (#3) take up a substantially smaller % of a standard TP-WS cycle. However, current content is balanced assuming we will use DDs that are at the Haste cap. If you attempt Incursion without any Haste, for instance, you are going to lose.

As a counterpoint, Haste is fun. Having essential Haste buffs is kind of a cool mechanism and I still find it fun to see myself attack faster and faster. It's just too bad that we lose so much battle system depth due to it (SC/MB, people skimp on JAs due to JA delay, etc.)



Solutions:
* Dramatically reduce TP gain to 1/3 or 1/4 of its present level - Possibly just let Haste affect TP/hit. Issue #1 would be addressed.

* Allow overlapping Magic Burst windows - So if a Distortion skillchain goes off and you start casting Blizzard V, but someone cancels it into a Fusion skillchain, your MB will still land for the proper damage as long as it's within 10 seconds of the Distortion skillchain. Issue #2 is addressed.

* Reduce animation delay - 1 second for WSs and 0.5 seconds for JAs would probably be fine after the TP reduction. Magic could drop down to 2 at least, perhaps also down to 1.

* Increase Magic Burst and WS damage - Do this adjustment "to taste" to even out high performance DPS to be approximately equivalent before and after adjustments.

* Fix enmity somehow - At this point I honestly don't care how easy it is to tank as long as it's possible to tank. I find easy-tanking strategically more interesting than can't-tanking.



This would create a battle system where people actually coordinate skillchains and magic bursts. It would be a player DPS nerf initially, but players could rise to the challenge. It would also subsequently fix a host of jobs:
1) Ranger (up) - Doesn't rely on Haste, so the TP change wouldn't affect them
2) Samurai (down), Warrior, Dark Knight, Dragoon, Thief, Ninja, Dancer, Blue Mage (up) - Lower WS frequency means more emphasis on white damage
3) Dancer, Puppetmaster, Rune Fencer - JA delay reduction would help these jobs out a lot
4) Black Mage, Scholar - Reduction in spell delay would massively help these jobs as magic DDs
5) Paladin, Rune Fencer, Ninja - Fixing enmity somehow would make these jobs far more useful

[Mitruya]

3) Dancer, Puppetmaster, Rune Fencer - JA delay reduction would help these jobs out a lot

(emphasis mine)

OMG, yessssss...
I actually avoid using the mage automatons just so I don't have to keep up with maneuver spam. When I'm derping around solo, the RNG automaton can generally handle itself without a lot of interaction from me. If I'm just tagging along for something non-important (because PUP never gets invited to anything important), I'll stand back and let the WHM automaton do all the work, knowing I'd hardly get any hits in from constantly keeping maneuvers up.
And that's kinda dumb, because I'm not utilizing my job to its full potential.
This is an issue of not only JA delay but also maneuver duration.

[Ulth]

I think a better way to deal with the weaponskill spam is to change the ftp so that they do more damage per tp the higher the tp. That way you will do more damage over all by storing up tp to use instead of weaponskilling as soon as you hit 1000. Right now fudo has a ftp of 3.25 at 1000, but only 5.75 at 2000. Damage at 2000 should be over twice as much as damage at 1000, and damage at 3000 should be over triple the damage of 1000.

[dasva]

Should also add another problem with skillchaining is it doesn't seem like they put to much thought into elements. So like some job/weapons say gkt sam can literally go around easily doing 5 step skillchaining with all the skillchains being 2 or higher because they have the elements for it other job/weapons might have a hard time even doing 3 step because most their level 2/3 element wss have the same or similar elements. This combined with the increased tp gain of sam incredibly favors them. Splitting up some of the other jobs higher level skillchains so they could do similar would help balance things somewhat

Along with the magic burst dmg increase I think there should be a significant potency/duration increase for MB debuffs. Given the current state of the game I doubt even something as large as doubling potency/duration would be too powerful

Also while we are killing ja/ws delay should kill the similar waiting delay on ranged attacks that isn't effected by snapshot

[Zheta]

All of my yes. +1

[Afania]

3) Forced (Animation) Delays - After you use a WS, you will not autoattack for the next 2 seconds and cannot initiate any other action. After you use a JA, you will not autoattack for the next 2 seconds, and can initiate another action after 1 second. After you use magic, you will not autoattack for the next 3 seconds and cannot initiate another action during that time. This animation delay makes up a huge portion of player time in the current battle system. In the example above, 2 seconds of the 7 second TP-WS cycle is spent waiting for your WS animation to finish. That's almost 30%. If you're a mage with a good fast cast set, you spend about 60% of your time waiting for the 3 second delay to be up. These forced delays are super annoying and detract from the otherwise fast pace of battle.


I don't really agree with forced animation delay, I think it works as part of the battle mechanic atm, so if you remove it, it could ended up being disastrous. Just think of it as "Making a choice and using your resource" in a turn based RPG or strategy game(Sorry if it's too abstract, it's hard to explain this properly). If you make a "move" for certain character or unit, that unit won't be able to move again for a while. So it's part of the game mechanic, and all the job mechanics/content are designed based on the fact that this delay exists.

Because the entire battle system was designed based on the fact that the animation delay exists, so there's a chance that SE can broke the game balance if they remove it...if they even can.

It's quite a lot of the effort to remove the entire animation delay and redesign/rebalance the battle system, when SE can't even balance the battle system using existing battle mechanic.

It seems that the only reason why you want SE to remove it is because "it feels annoying". Personally, I'd rather feel slight more annoyed with the delay, it's much better than seeing SE spending 1 year redesigning a battle system for a 12 year old MMORPG.

[Seillan]

I don't really agree with forced animation delay, I think it works as part of the battle mechanic atm, so if you remove it, it could ended up being disastrous. Just think of it as "Making a choice and using your resource" in a turn based RPG or strategy game(Sorry if it's too abstract, it's hard to explain this properly). If you make a "move" for certain character or unit, that unit won't be able to move again for a while. So it's part of the game mechanic, and all the job mechanics/content are designed based on the fact that this delay exists.

Because the entire battle system was designed based on the fact that the animation delay exists, so there's a chance that SE can broke the game balance if they remove it...if they even can.

It's quite a lot of the effort to remove the entire animation delay and redesign/rebalance the battle system, when SE can't even balance the battle system using existing battle mechanic.

It seems that the only reason why you want SE to remove it is because "it feels annoying". Personally, I'd rather feel slight more annoyed with the delay, it's much better than seeing SE spending 1 year redesigning a battle system for a 12 year old MMORPG.

These were my thoughts on this too, but I refrained from saying anything since I haven't done any higher level combat yet. It's always seemed to me that the delay in battle was more of a feature than a game engine artifact or something. The combat in this game reminds me more of an active-style, turn-based JRPG than any other that I've played, and I personally love that. I agree that it should remain; the game just doesn't seem like it would transition well into the kind of pace that exists in a game like WoW, for example. Heck, even FFXIV has a GCD that many people find excessive (like 3 seconds I think), yet it gives the game a unique pace and a little more leeway for strategic decisions. That's not a bad thing, imo.

[Byrth]

In JRPGs there may be different delays associated with actions, but (at least for all the ones I've played) using the action itself takes no time and everyone else's ATB gauge freezes while whatever action you pick executes. That's a very different dynamic than what we have in FFXI.

This is what our global lockout timer does to gameplay at the delay cap, with all these calculations being done after the most recent update:

The job adjustments had no impact on whether or not Dancer is used, and here is why:

Every time you use a JA, you incur a forced delay (minimum of 1 second, in practice about ~1.3 seconds on average). I adjusted Motenten's sheet (https://www.dropbox.com/s/ayejryvsfpzxfi5/DPS%20Calculator%20-%20Dnc.xlsx?dl=0)to model the changes to Ternary/Striking flourish and assign their JA delay to that TP/WS cycle, as well as added JA delay for other things like Steps.

Assumptions:
* 80 delay (~1.3 seconds) per JA. -- This is a favorable assumption because it assumes every JA is followed by a WS. It represents the maximum possible efficiency given server-client transfer times and very probably underestimates the real JA delay.
* A 3 minute fight, which is essentially the most favorable possible assumption for Steps.
* The modeled player was responsible for applying the Steps.
* Berserk, Haste Samba, etc. were all applied at the beginning of the fight, and that Haste Samba would have to be applied twice.
* I used Pyrrhic Kleos and Climactic-Rudra's as my two test cases. I didn't use Evisceration because I originally set out to test Tier 3 flourishes and they're all fairly worthless with Evisc.
* No build-up time for step potency. It was assumed to be instantly 100% potent, so 10 steps would be -23% Defense from the first second of the fight.
* My own gear.
* One BRD for buffs.
* No skillchains
* Reverse Flourish is an instant 1000 TP for WS.
* Steps are 100% Accurate

Some conclusions:
1) Climactic is the only Flourish worth using before a WS. Ternary, Striking, and Building are each a net DPS loss in most situations simply due to their activation times (not counting the time it takes to get steps).

2) In a buffed situation, Feather Step starts having a negative impact on the DNC's DPS almost immediately. I found that level 1 was a net DPS gain for the DNC, but after that it went downhill. Might have been different if I modeled Evisceration.

3) Box Step is worthwhile until you cap Ratio.

4) With capped magic Haste, using a 5% DW Toetapper Mantle and not using Haste Samba was essentially equivalent to using Haste Samba (-1% DPS) and I'm not even sure that's the most efficient tradeoff. Considering Haste Samba doesn't help SAMs, DRKs (in a 3 min fight), DW jobs, or H2H jobs, this would be a way to simplify your playstyle.

5) Saber Dance is only a net +3% DPS. Given that you don't really need Haste Samba (#4), there will be times when Fan Dance is a more reasonable option.

6) Reverse Flourish is definitely worth using if you don't attribute any of the delay from FM generation to it (80 delay). If you attribute two JAs worth of delay (160) to it, it's still minorly worth using. Three JAs (240 delay) make it not worth using anymore, so you basically should never use useless steps to build FMs for Reverse Flourish, because you'll always need at least 3 JAs (Presto + Feather Step + Reverse Flourish or Step + Step + Reverse Flourish) to get a full WS worth of TP back even with Terpsichore.

Disclaimer: It is possible to precharge Presto, Striking Flourish, etc. when moving between mobs so the JA delay is effectively negated, which I do. I was looking to evaluate the changes from this update together, though, so a 3 minute fight is really the best place to do it.




Does this make the problems with Dancer a little more clear? In the ideal group-event case outlined above, you use Box Step-> WS until you cap it and mix in two Box Step -> Climactic Flourish -> Rudra's Storms combos and as many Reverse Flourish -> WSs as you can while you have free FMs. You use Feather Step at most once. You probably don't use Haste Samba or Saber Dance unless your WHM is on top of their stuff or you're doing an event with a WAR or DRG. After you max out Box Step, you just sit there spamming WSs and not using any of your Dancer specific JAs.

[Malithar]

So it's part of the game mechanic, and all the job mechanics/content are designed based on the fact that this delay exists.

I get where you're coming from, and to a degree I agree. But whether abilities and spells are designed around this or not, they most certainly are not balanced around it, else you'd never have instances where using these things is harmful to your performance. Varying degrees of use according to the situation, haste values, buffs, etc, sure, but harmful? That shouldn't be the case IMO.

[Ulth]

If you are using the job ability right before a weapon skill why are you assuming the delay is 1.3 instead of 1? Also have you tried climatic flourish <wait 1>, step <wait 1>, weaponskill <wait 3>, presto <wait 1>, step <wait 1>, reverse flourish <wait 1>, weaponskill?

[Afania]

I get where you're coming from, and to a degree I agree. But whether abilities and spells are designed around this or not, they most certainly are not balanced around it, else you'd never have instances where using these things is harmful to your performance. Varying degrees of use according to the situation, haste values, buffs, etc, sure, but harmful? That shouldn't be the case IMO.


I don't view it as "harmful to performance", it's making a choice. I use steps a lot in incursion and always cap def down in incursion. I never feel it's "harmful to my performance" because it seems that it's dmg increase in the long run.

In delve I only use 1 step or I don't use it at all, because the NM dies in less than 30 sec so it doesn't seems to worth the effort to cap it.....by the time I cap it's dead already.

If you have no penalty for using it, then you'd just cap steps in all situations, instead of making a choice between lv 5 steps or lv1 or no steps.



In JRPGs there may be different delays associated with actions, but (at least for all the ones I've played) using the action itself takes no time and everyone else's ATB gauge freezes while whatever action you pick executes. That's a very different dynamic than what we have in FFXI.


The actions in FFXI also takes no time to execute, it's just that you can't do anything after you execute it.

If I remember correctly, in FFX you make a choice first, then depending on the action you pick, you may have to wait longer for your next turn.

[Malithar]

If you have no penalty for using it, then you'd just cap steps in all situations, instead of making a choice between lv 5 steps or lv1 or no steps.

I agree, the situations you listed are good examples of making that choice, and choices are good, like when I said the potency/use could vary, but it should still be positive regardless, or at the very least not detrimental. I'm more referring to say, using Haste Samba when your, or no one else's, TP set is setup for it, incurring JA delay and TP loss, even loss of additional procs from enspells and such, for no gain, using jumps under high haste without Ryu, overwriting a main War's Warcry with your /War Warcry (unintentional or otherwise, this is rage worthy), attempting to actually keep an Automaton's Maneuvers up while DDing, etc. These types of abilities and many other's are not balanced around the game as it is currently.

Again, I'm all for choice (use x ability for y benefit on z mob, or save for a time when it's more beneficial), but outside of baked in balancing effects, CD's, etc, usage of abilities or spells shouldn't be detrimental to yourself or your group's performance unless it's a gimmick type of thing, no magic damage during this phase, no WSs unless the mobs doing this thing, etc.

[Byrth]

If you are using the job ability right before a weapon skill why are you assuming the delay is 1.3 instead of 1? Also have you tried climatic flourish <wait 1>, step <wait 1>, weaponskill <wait 3>, presto <wait 1>, step <wait 1>, reverse flourish <wait 1>, weaponskill?

IIRC <wait 1> isn't actually waiting 1 second. It's waiting 1 second and 1 packet cycle, so on average it waits about 1.3 seconds. Every macro line waits 1 packet cycle, which is why they lines don't execute instantly even without wait statements.


If I remember correctly, in FFX you make a choice first, then depending on the action you pick, you may have to wait longer for your next turn.

That's exactly what I described in the post you quoted.

[Mitruya]

I don't really agree with forced animation delay, I think it works as part of the battle mechanic atm, so if you remove it, it could ended up being disastrous. Just think of it as "Making a choice and using your resource" in a turn based RPG or strategy game(Sorry if it's too abstract, it's hard to explain this properly). If you make a "move" for certain character or unit, that unit won't be able to move again for a while. So it's part of the game mechanic, and all the job mechanics/content are designed based on the fact that this delay exists.

Because the entire battle system was designed based on the fact that the animation delay exists, so there's a chance that SE can broke the game balance if they remove it...if they even can.

It's quite a lot of the effort to remove the entire animation delay and redesign/rebalance the battle system, when SE can't even balance the battle system using existing battle mechanic.

It seems that the only reason why you want SE to remove it is because "it feels annoying". Personally, I'd rather feel slight more annoyed with the delay, it's much better than seeing SE spending 1 year redesigning a battle system for a 12 year old MMORPG.

Because these delays keep jobs like PUP and DNC and whatever else out of content because they don't win the almighty parse wars. THAT'S the real annoyance - not being invited because we can't keep up.

[Afania]

Because these delays keep jobs like PUP and DNC and whatever else out of content because they don't win the almighty parse wars. THAT'S the real annoyance - not being invited because we can't keep up.

It is not the delay that kept DNC and PUP away though. Even if they remove delay DNC and PUP are still behind SAM in terms of dmg output, just check spreadsheet and you'd know.........the correct way to fix DNC and PUP is to give them game changing buff/debuff so they get invite, instead of redesign and recode the ENTIRE battle system for them. That's too much effort and it doesn't fix things at all.

[Mitruya]

It is not the delay that kept DNC and PUP away though. Even if they remove delay DNC and PUP are still behind SAM in terms of dmg output, just check spreadsheet and you'd know.........the correct way to fix DNC and PUP is to give them game changing buff/debuff so they get invite, instead of redesign and recode the ENTIRE battle system for them. That's too much effort and it doesn't fix things at all.

Oh, I get what you're saying, and I agree. And I've said it before, it's not so much delay with PUP than too-short maneuver duration.

[Olor]

I think the delay wouldn't be so killing if durations were extended and their DD power was increased to offset the delay. But I think the delay needs to be reduced for these jobs at the very least.

[Ulth]

It is not the delay that kept DNC and PUP away though. Even if they remove delay DNC and PUP are still behind SAM in terms of dmg output, just check spreadsheet and you'd know.........the correct way to fix DNC and PUP is to give them game changing buff/debuff so they get invite, instead of redesign and recode the ENTIRE battle system for them. That's too much effort and it doesn't fix things at all.

That makes more sense, and is easier than revamping the entire combat system so I agree. Slash the costs of waltz, boost the potency of steps and sambas. Dancers have the potential to be as good a support as brd. I don't know about pups, but they could just make maneuvers last five minutes like runes. Maybe change the puppet's ai so it will cast haste II on more than just the master, other stuff like that.

[Mitruya]

That makes more sense, and is easier than revamping the entire combat system so I agree. Slash the costs of waltz, boost the potency of steps and sambas. Dancers have the potential to be as good a support as brd. I don't know about pups, but they could just make maneuvers last five minutes like runes. Maybe change the puppet's ai so it will cast haste II on more than just the master, other stuff like that.

I'd be cool with that.

[Afania]

That makes more sense, and is easier than revamping the entire combat system so I agree. Slash the costs of waltz, boost the potency of steps and sambas. Dancers have the potential to be as good a support as brd. I don't know about pups, but they could just make maneuvers last five minutes like runes. Maybe change the puppet's ai so it will cast haste II on more than just the master, other stuff like that.

Honestly, just haste2 for everyone and better pet survive-ability is enough to warrant PUP a spot even if PUP's output stays the same.