Bring back Aggro, resource management, bring back consequences!

[Shurrikhan]

If you make running out impossible, why have the system? If you want to make a system that gives you more damage the higher your TP is (from a certain value) then just doing your normal rotation would neccessarily need to force you to be negative, which would then require the implementation of an action that cost less TP to make up the deficit.

No. If you build around a particular rate of consumption (which is GCD scaled anyways, in this case), then that rate of consumption does not force you to go negative. You can weave in more burst skills than you typically would, such that, much as the case with enmity, you end the fight with minimum excess TP if the value of those higher-TP-cost skills over the normal rotation is greater than the value of the %TP potency modifier, or likely just anywhere just short of or barely dipping into the penalized range if not, unless the fight's situations give cause to burst further.

You can, of course, build short of that particular rate of consumption, as to oblige the use of Support skills after X minutes, but that's an externality. All a ubiquitously-tied resource system fundamentally does is allow you to trade long-term gains that are greater on paper for short-term gains that may be greater, or at least vital/necessary, in certain contexts.

In AoE, as I stated, your actions cost between 2 and 3 times as much TP

This, too, is externality. You can tune them to whatever you damn well please; it's merely a matter of balancing higher tuning against lesser available resource. For instance, if potency is decreased by 1% for each %TP below 75, to a minimum 50% potency at 25% TP, but TP costs decrease over that same span to a minimum of a quarter, the AoE penalty of using (stronger) AoEs would begin to right itself after the first 4-5 GCDs of burst, as their (and other) potencies decrease; similarly, your TP would quickly right itself to over 80% with time not spent on using those burstier skills.

The issue here is that the games UI does not support such a system [as managing enmity].

XIV UI has been sufficient for that purpose. That's why we have the bars. The Enmity percentile bars weren't hard to use.

With low enmity values, like at the start of a fight, you have a good idea of how far ahead of the pack you are. As the fight progresses and enmmity increases, the same space on the enmity bar is eqivalent to a much higher value.

Yes, because that's how percentiles work. Again, it wasn't difficult to understand or work with. ...Do you want bars that traverse your whole screen? Unlike HP, precisely because of Enmity modifiers, there is no maximum to enmity that wouldn't make the UI worthless early in a fight if were linearly representative rather than percentile.

As for Enmity, the important thing to note is that it is a Binary system, you either have it or you do not. If the old system was so bad, what could replace it?

Short version: You include more than a single mob script (what we sometimes call "AI") and make further factors available to those scripts. At present, there is only one in the whole game (attack top threat regardless of other conditions), which at most adds the feature of also attacking 2nd threat with a second, dedicated periodic attack.

Longer version: You differentiate, under the game's hood, between latent and active enmity, mob script and mob mode (a different behavioral set procedurally called upon by conditions set in the script, such as being frenzied, wary, cocky, etc.), and generate undermechanics for affecting enmity that interact with mob behavior.

[Mikey_R]

No. If you build around a particular rate of consumption (which is GCD scaled anyways, in this case), then that rate of consumption does not force you to go negative. You can weave in more burst skills than you typically would, such that, much as the case with enmity, you end the fight with minimum excess TP if the value of those higher-TP-cost skills over the normal rotation is greater than the value of the %TP potency modifier, or likely just anywhere just short of or barely dipping into the penalized range if not, unless the fight's situations give cause to burst further.

You can, of course, build short of that particular rate of consumption, as to oblige the use of Support skills after X minutes, but that's an externality. All a ubiquitously-tied resource system fundamentally does is allow you to trade long-term gains that are greater on paper for short-term gains that may be greater, or at least vital/necessary, in certain contexts.

So basically, take what I said and flip it on its head, so your main combo builds it up and you spend it on burst damage, whilst keeping above a certain threshold. This is just another gauge and personally, I would say it was better suited as a main mechanic for a job rather than something all TP users should have. Infact, we even had a job similar in the form of original SB Machinist. You wanted you build up heat to get above 50 Heat to get access to the heated shots, then used the stronger Heat Blast to bring it back down before you capped at 100 (Overheating was a DPS loss on the original MCH), so you had to strike a balance between 50 and 95 for maximum damage.

As for the enmity comments, I don't think you quite understood where I was going. To give simplistic examples, tanks have 2 attacks, one that does enmity, one that does damage, and DPS just damage consistently. This means you want to use the enmity combo to keep above the DPS, but not so much that you could potentially lose out on a damage hit. This means you need to keep your enmity higher than the DPS by some amount. This amount would remain consistent throughout the fight as enmity gain is essentially linear.

For this, I am only going to say, you get 1000 enmity per second. This enmity is tracked by a bar which has a certain resolution, to make life easier, lets just give it 100 units of resolution. After 1 second, each unit represents 10 enmity. After 10 seconds, you have 10,000 enmity, each unit now represents 100 enmity. We have already lost a whole order of magnitude to the resolution, but lets continue. 5 mins, which is 300 seconds, or 300,000 enmity, each block is now worth 3000 enmity. 10 minute fight, each one is worth 6000. The size of the bar itself has not changed, but how much each unit represents has. If you know you need to stay, say, 500 enmity above the dps to be safe, after 1 second, the 500 enmity is equivalent to 50 units. After 10 seconds, the dps would be at 9500 to your 10,000, each block represents 100, so your have 5 units to show that you are still just above the dps. After 5 minutes, you at 300,000 enmity, staying 500 above the dps, 299,500 enmity. Each block is 3000, so the DPS have filled 99/100 units. The DPS could be 500 behind you, they could be 3000 behind you, which would allow more damage on your end, you just do not know. After 10 minutes, the DPS is still at 99/100 units, but the range they could be in has increased from 500 behind you to 6000 behind you, you just do not know. The longer the fight goes on for, the less accurate you would be, and so the more likely you are to use your enmity combo, just to be safe and bear in mind, this is a very very static example simplified to get the point across. In reality, it is going to be much more dynamic with different jobs catching you at different rates, it is going to depend on how lucky the got with direct/crit hits, how much self healing have they done etc. It would make the whole enmity thing a complete guessing game the later in a fight you go.

Short version: You include more than a single mob script (what we sometimes call "AI") and make further factors available to those scripts. At present, there is only one in the whole game (attack top threat regardless of other conditions), which at most adds the feature of also attacking 2nd threat with a second, dedicated periodic attack.

Longer version: You differentiate, under the game's hood, between latent and active enmity, mob script and mob mode (a different behavioral set procedurally called upon by conditions set in the script, such as being frenzied, wary, cocky, etc.), and generate undermechanics for affecting enmity that interact with mob behavior.

And I fail to see how this would make things fun. Ok, a mob decides it wants to smack the healer, fine, I am now forced to use the enmity combo to get it back, or, more likely, just provoke it, if provoke is down, shield lob/tomahawk etc. to get it back. This isn't going to make tanking more engaging, it is just going to be annoying. If it is just, the mobs hits the healer/dps for a hit, well, we already have that, so that is nothing new. I suspect there is something else you had in mind, however, I fail to see what you are trying to get at with your vague explanation.

If you want to make things more interesting, you need to make use of the tanks full kit, things that are currently woefully underused are Low Blow and Interject. Apart from the odd scenario here and there, you have no reason to think about them. Provide some interesting things that go with them, the big thing I have always said was changing a moves properties rather than just outright stopping it. As a tank, you should be there to control the enemies, both in positioning for easier DPS uptime and mitigating damage to help the healers keep you and the party alive.

[Sqwall]

Agro management was terrible. With all the things I have to manage now as a main tank....I should NOT have to worry about a BLM that is about to overtake me because they are spitting nuclear hot death.

I'm so glad this isn't WoW agro style management. In FFXIV you either have it....or you don't.

The tank busters in this savage tier ALONE have given me a moment of pause to think about agro management. The tank busters in P1S are fun and make you actually work with your OT.

Agro management sucked....and it was outdated. I'm glad it's moving towards new and interesting tank mechanics that require the tanks to basically be in tune with each other. At least I am with my static OT.

Plus this allows the DPS to balls out on without having to hold back.

[Shurrikhan]

So basically, take what I said and flip it on its head, so your main combo builds it up and you spend it on burst damage, whilst keeping above a certain threshold. This is just another gauge and personally, I would say it was better suited as a main mechanic for a job rather than something all TP users should have.

Sure, you can flip what you said on your head all you like... but it still won't have anything more to do with my position or what I described to you.

A resource which starts at maximum and is regenerated over time is the exact opposite of a resource which starts at 0 at is generated by actions. How did you quote the prior and arrive at the latter?

Infact, we even had a job similar in the form of original SB Machinist. You wanted you build up heat to get above 50 Heat to get access to the heated shots, then used the stronger Heat Blast to bring it back down before you capped at 100 (Overheating was a DPS loss on the original MCH), so you had to strike a balance between 50 and 95 for maximum damage.

I am aware of how StB MCH played. (Though, no, Overheating was not always a dps loss. Its potential damage gain was frustratingly minimal for how much effort was involved in order to exploit it at all, but those opportunities did exist.)

It's still the exact opposite of what I'm talking about here. You're describing a from-zero resource built by actions and which must on occasion make use of a weaker action in order not to be penalized, as opposed to a from-max resource regenerated over time and which allows some n actions per minute and, further, per fight (as per the time the time it'd take to drop initially to the effective minimum resource).
_____________

As for the enmity comments, I don't think you quite understood where I was going. To give simplistic examples, tanks have 2 attacks, one that does enmity, one that does damage, and DPS just damage consistently. This means you want to use the enmity combo to keep above the DPS, but not so much that you could potentially lose out on a damage hit.

No, I understood completely. I'm asking, if the default UI is so useless as to make enmity management impossible, what "better" UI solution you'd propose. I had little to no trouble minimizing excess enmity with the default UI.

I'd be fine with seeing a "+<value>k" descriptor added beside the lead enmity, or a + or - value added for ones, to better show that exact difference, but that's something that took a modder an scant afternoon to add to the game. If the devs wanted to actually bring back aggro management into the game's modern climate, they could provide something equivalent.

Again, as per my very first post here, I have no interest in readding enmity management if it remains a mere matter of remaining the lead entry on a damage-times-modifier table by a margin that ought to be minimized by the fight's end. But to say assume absolutely that the effective removal of enmity management, as per the past two expansions, would be reversed without any chance of accordant UI improvements or surrounding changes by which to situate it to the current game... is not sensible.

And I fail to see how this would make things fun. Ok, a mob decides it wants to smack the healer, fine, I am now forced to use the enmity combo to get it back, or, more likely, just provoke it, if provoke is down, shield lob/tomahawk etc. to get it back. This isn't going to make tanking more engaging, it is just going to be annoying.

You're strawmanning again. I've not asked for the return of Enmity skills.

I have, multiple times in this thread alone, pointed out that I think they're bloated -- inefficient ways of handling things that should instead be covered by more involved undermechanics, rather than GCD choices limited to a single role's kits.

I merely don't think that Enmity skills competing with damage skills makes them non-mechanics any more than buffing GCDs like Twin Snakes or Heavy Thrust would compete with the damage skills that benefit them. Minimizing excess margin, in either case, is the mechanic -- however shallow it may be in the case of Enmity-enhanced skills.

If you want to make things more interesting, you need to make use of the tanks full kit, things that are currently woefully underused are Low Blow and Interject.

Mob AI and stuns/interrupts are not mutually exclusive.

What's next? Mobs can't have more than a single AI script because we couldn't then have active mitigation? We can't have involved rotations because then we couldn't make use of Interject? None of these things conflict with each other.

[Zarkovitch]

with how the game works Idk about this one chief.

[Daeriion_Aeradiir]

What's next? Mobs can't have more than a single AI script because we couldn't then have active mitigation? We can't have involved rotations because then we couldn't make use of Interject? None of these things conflict with each other.

Considering how they handle multi-targeting attacks, it likely actually is impossible within their framework.

In its real case, any attack a boss launches that targets more than 1 player simultaneously, is actually the boss + a bunch of invisible entity copies of the boss that copy its buffs/debuffs hiding under the map all targeting different players, since FF14's enemy framework can only have a single valid target for an enemy for any one attack. So for your idea, the boss wouldn't be able to track more than one target as far as latent and active aggro unless it was physically 2 bosses on the field and one of them ran a different script from the main boss.

You can actually see an instance of this in EX2 where the boss aggro name physically changes to whoever they're targeting for the shared raid AOE hit on the sword attack, since its launched by them directly and not one of the copies that are responsible for other attacks in the fight, like the 2x water stack.

Some other fun examples include:

-Garuda, Ifrit & titan all cast a jail on a different player during uwu, even though Garuda & ifrit are off the field and hidden.
-E12s's boss has a second copy that targets the #2 in hate specifically for both hits, while the main boss only ever targets #1 for both hits.
-E8s boss has the same deal with ahk morn as E12s's boss does.
-Every instance of Limit cut is actually 4 instances of Cruise Chasers with different targets set for their mechanic
-Photon in TEA is casted by two to eight hidden Cruise Chasers in the Brute/Cruise phase at a set time. This results in an interesting situation where if you kill BJ/CC too quickly, the Photon meant to drop the tank's HP to 1 in the later half of the phase will instead happen during the lead-up to timestop: https://www.twitch.tv/ffxivmomo/clip...4L5HyFi7gz3rvy

TL : DR, boss AI has a ton of jank hidden under the hood, likely making a lot of complex interactions effectively impossible.

[Shurrikhan]

TL : DR, boss AI has a ton of jank hidden under the hood, likely making a lot of complex interactions effectively impossible.

What I'm asking for, though, doesn't require multiple simultaneous targets any more than is presently the case. It merely requires a greater variety in who is targeted and what tanks can do about it such that being merely at the top of the Enmity table isn't the be-all and end-all of target selection prior to separate "raid mechanics" (or "double-tank busters" alone), and that space gives tanks more to do.

Mobs should at least have an illusion of some small degree of intelligence, and thereby allow for more to be involved in tanking than merely having their stance on and sometimes leading about a cylinder as mechanics fall or, at best, emanate from said cylinder.


__________

Now, to get the extent of gameplay mentioned before for if I had as much resource as I wanted (i.e., under that pipe-dream heading of "ideally"), that boss/mob/NPC script would also have to be susceptible to change given particular conditions -- able to be manipulated. That's a very separate story from how many enemies a single attack can target, however.

[Mikey_R]

1. A resource which starts at maximum and is regenerated over time is the exact opposite of a resource which starts at 0 at is generated by actions. How did you quote the prior and arrive at the latter?
_____________

2. No, I understood completely. I'm asking, if the default UI is so useless as to make enmity management impossible, what "better" UI solution you'd propose. I had little to no trouble minimizing excess enmity with the default UI.

3. You're strawmanning again. I've not asked for the return of Enmity skills.

4. Mob AI and stuns/interrupts are not mutually exclusive.

What's next? Mobs can't have more than a single AI script because we couldn't then have active mitigation? We can't have involved rotations because then we couldn't make use of Interject? None of these things conflict with each other.

1. Whether it start at max or 0, whether it regens over time or not, whether it is gained via GCDs or not, at the end of the day, the mechanic is, keep it between 2 values with things that manipulate the value. If you are forced out of the ideal range, you get back in as quick as possible. Trying to adopt this into the existing jobs is not going to work, it has enough complexity to warrant a whole job based around it.

2. Yes, that is my point. There is no way to do it without numerical values, but I guarantee SE will not put numbers onto enmity that we can see by default, which then means you cannot have a mechanic that requires you to stay a bit ahead of the DPS.

3. Going back and re-reading past posts, it seems I did misunderstand/miss a detail. Neither of us want the old enmity system back, you just want something different to replace it, which then goes into the next points.

4. I never said they were, however, I fail to see what adding different mob scripts would actually do? You say they could be frenzied, wary, cocky, etc. but how does that translate into what I have to do differently as a tank to deal with it. What different scenarios could be made? I have already gone with a randomly attacking healer/dps, easy to deal with and if done too much, would get annoying. You could have a mob run away, that would get tiring fast having to chase it down. If you want a boss to have different 'moods' that is just different phases of a fight that might come in a random order (which could be a good thing to have more of). I suspect you have some ideas, I just want to know what they are.

[Shurrikhan]

1. Whether it start at max or 0, whether it regens over time or not, whether it is gained via GCDs or not, at the end of the day, the mechanic is, keep it between 2 values with things that manipulate the value. If you are forced out of the ideal range, you get back in as quick as possible. Trying to adopt this into the existing jobs is not going to work, it has enough complexity to warrant a whole job based around it.

I disagree, but only very mildly. I feel that as long as MP is to be a universal component, we may as well make use of it (though such does mean getting rid of Lucid Dream, scaling MP ticks to GCD speed, and so on). It'd also help balance out the rDPS value of having access to rez / greatly diminish any need for a rez tax, as getting someone back up means that much less burst potential over time available for the rezzer. It's not where I'd choose to use developer resources, unless they were damn near boundless, but any necessarily universal UI component (for the space for an MP bar must always be made, even if the bar need not always be shown) that is not also made a universal mechanic will at the very least be an eyesore.

2. Yes, that is my point. There is no way to do it without numerical values, but I guarantee SE will not put numbers onto enmity that we can see by default, which then means you cannot have a mechanic that requires you to stay a bit ahead of the DPS.

You know what, I probably am overestimating the devs here, given the amount of rather important QoL changes that, despite being desired since ARR, have seen no adjustment outside of modder communities. I concede this point. Sorry for being a bother.

3. Going back and re-reading past posts, it seems I did misunderstand/miss a detail. Neither of us want the old enmity system back, you just want something different to replace it, which then goes into the next points.

Any my apologies for thinking you were purposely strawmanning. Looking back on it, that detail hadn't been repeated for a couple posts.

4. I never said they were, however, I fail to see what adding different mob scripts would actually do? You say they could be frenzied, wary, cocky, etc. but how does that translate into what I have to do differently as a tank to deal with it. What different scenarios could be made? I have already gone with a randomly attacking healer/dps, easy to deal with and if done too much, would get annoying. You could have a mob run away, that would get tiring fast having to chase it down. If you want a boss to have different 'moods' that is just different phases of a fight that might come in a random order (which could be a good thing to have more of). I suspect you have some ideas, I just want to know what they are.

I don't have much time at the moment --I had a long convo with Shougun, I think, and a couple others about this back in the day, but I couldn't seem to locate the old threads. I'll try to give the short of it, but I'll need to situate it within some context first.

Namely, we wouldn't be able to drag-and-drop this upon our current tanks, given current undermechanics. For more shit happening that isn't controlled by tanks' merely existing (and being dps with occasional mitigation skills to press according to schedule) to be fun... tanks would need to be able to skillfully deal with that.

So how do we make tanks able to deal with shit, especially without relying on enmity tools? Well, primarily by making everyone able to perform a degree of tanking at a moment's notice, without need for enmity, which is best done by changing the way attacks work. Consider, for instance, if nearly all attacks were among four types:

• Split-damage/soakables,
• Smart-damage soakables (damage is distributed according to remaining eHP)
• Degradable cleaves (damage is decreased cumulatively by mitigation, and perhaps a further amount per hit, as it moves through its victims), or
• First-intercepted skillshots (it hits whoever's first in its path or emanation)

At that point, you don't actually need to be a Tank to 'tank' (in the sense of defending others); instead, being a Tank would simply give some obvious synergies, such as having a larger hidden/secondary hitbox that'd be supersede other interceptees' if and only if another is hit or reducing the total damage dealt by any soak in which it contributes by its mitigation (probably OP, and should instead work akin to smart-damage soakables, but w/e).

Further, you perhaps give more some Tanks (Paladin, at the very least) to-ally gap-closers, a bit more abundant ability to suppress enemy damage directly (rather than only via diverting attacks to you and then mitigating them), more integral and frequent use of job-thematic ability by which to protect or replenish the party (think Dark Missionary revisions, like Passage of Arms being way more available, Warrior getting a more versatile War Cry mechanic in place of just a dull raid barrier, etc.).

Okay, now we're finally on a trajectory that, at least from the Tank perspective, could deal with more varied mob behaviors.

Now, why would I even want that, though, in general? Primarily, because I see Tanks as a role whose purpose is to increase the ratio of party damage done to party damage taken through leveraging (at least momentarily) above-average eHP and/or threat to manipulate enemies, and I want to do more with that. Secondly, because I want there to be content that isn't a set DDR experience. I want variety not just in fights, but within a given fight, at least across certain content types. That means there need to be factors by which to generate alternate events. I'm not looking for "smarter AI" (which just leads to us getting the absolute poorest showing out of our on-paper capacities as they kick our asses with a given set of stats and abilities finally well utilized), but simply at least a pretense of our fighting an enemy, rather than striking dummy, and the variance that ought to contain.

The (bareness of the) current enmity system basically forbids that, which is why I want to revise it. Let's put it this way: no matter what changes we might imagine have, in the enemy's mind, occurred over the fight, the enmity calculations will remain exactly the same. Even if, after a whole few lines of monologuing, two ASTs instantly heal the whole party back from 1 HP to full in just over a GCD's time, nothing would change. Even if the healers, repeatedly, leave all but the tanks at just enough HP to survive the next raid-wide, nothing would change. There are no thresholds under which a fight can differ based on player behavior. Now, while I'd be as pissed as anyone if the boss left the nearly dead party to slowly regen the first time around only to, without any warning, decide simply to finish them off the next time, but that's where you utilize those kind of clues and make use of your tanks not just as people who go 123 while aiming the enemy out, but actually have to thwart the enemy's offenses. If there are times when the boss will go for the most wounded target, or attempt to just maximize damage across all your squishies, that introduces new considerations to tanking -- especially regarding positioning and use of defensive CDs.

Now, the far end of that pipedream even involves actually turning the enemy into something like a player, with a given body of characteristics, skills and resources and certain thresholds (unknown to it) that may unlock further buffs, skills, and resources such that you'd want to have a Paladin especially earn its hatred for a time until it attempts to unleash a party-wipe skill focused solely on that PLD (who Hallows it) while your Warrior soaks lighter cleaves for a flanking group, sharing lifesteal buffs based on proximity before taking the higher part of the mob's attention when it realizes it may die --putting it into a panic so it burns defensive resources when you're actually overextended anyways-- etc., etc. But, for the start it's just a matter of giving tanks more to do and shaking up fights a bit, giving us some further reason for our actions beyond just "heal everyone to at least Y amount by time X:XX" or "hold at least one point of Enmity above any non-tank".

[Mikey_R]

But, for the start it's just a matter of giving tanks more to do and shaking up fights a bit, giving us some further reason for our actions beyond just "heal everyone to at least Y amount by time X:XX" or "hold at least one point of Enmity above any non-tank".

I think this one point here is the crux of everything. Tanks just need more to do. We now all have tools to mitigate damage on other people, but as far as I am aware, they are mainly used on the other tank. They have used mechanics where the tank has to be in front to take the most damage before (Thordan EX comes to mind) but it really is rare and I don't think it has come again since then (from memory).

One thing I will say though with the example of the AoE, only healing enough to survive and then the boss throwing out another, the problem with things like this is it could be seen as a punishment mechanic. As a healer. you knew the party would survive, they do, but the boss decides to throw out another anyway to kill everyone, whereas if you had healed more before the first, you wouldn't have the second AoE. It could be seen as a punishment as you have to spend more resources before/after to ensure you stay above the HP threshold rather than knowing you would survive and allowing your regens to do their job. This is where you have to be careful with mechanics as you do not want to feel like you are being punished by taking more damage for no reason.