Provoke - Analyzed

[Extreme]

‚΂©ŠO?l?B?B?B‚È‚¢Ž„‚ÌŽžŠÔ‚̉¿’l‚ðŽ?‚Á‚Ä‚_‚È‚½‚Ì

[DannyForgette]

ok ill try to explain extreme's theory in numbers
x>/=0(provoker's hate)
y>/=0(puller's hate)
z>/=0(provoke's effect)

lets say that provoke adds 5 hate and subtracts 3 from the party
z=x(0)-3(0) y(0)+7(7)
the puller provokes gets 5 hate
z=x(0)+7(7) y(7)-3(4)
the tank pulls it off the puller

in this example their where two party members

now whe're in a party of six
provoke's value is the same as listed previously

hate is .1*(damage)or.1(# of hp healed)

a=WHM's hate
b=BLM's hate
c=RDM's hate
d=MNK's hate
e=THF's hate
f=WAR's hate
z=provoke's effect

z=a(0)-3(0) b(0)-3(0) c(0)-3(0) d(0)-3(0) e(0)-3(0) f(0)+7(7)

This is what happens when the warrior pull's the monster with provoke

z=a(0)+3(3) b(0) c(0) d(0) e(0) f(7)-3(5)

This is what happens when the WHM heals the WAR for 30 hp, as of now the monster has only hit the WAR for 30 damage. the monster sontinues to attack the warrior.

z=a(3) b(0)+3(3) c(0)+1(1) d(0)+1+1(2) e(0)+.7(.7) f(5)+2-1(6)

Everyone attacks as normal, WHM doesnt do any heals. BLM uses water and hits for 30 damage. RDM hits for 10 with his sword. MNK hits for 20(10 each). THF hits for 7 since he is using a dagger.
WAR hits for 20 with his Greataxe. Monster hits for 10. WAR mantains aggro.

z=a(3)+1(4)-3(1) b(3)-3(0) c(1)-3(0) d(2)+7(9) e(.7)+2(2.7)-3(0)
f(6)-3(3)

WHM heals WAR for 10. THF criticals for 20 damage. MNK provokes
since he has WAR sub. this is about 15 seconds after the WAR's initial provoke.

This is a somewhat basic example of extreme's theory. but in-game i think provoke is a percentage. i just used arithmetic for simplicity. if provoke was a +/- number, then provoke would become useless at later levels.

just my 2 cents.

[laicram]

Great info. Extreme, your attitude's bumming me out =(

[Spider-Dan]

Danny, I understand Extreme's theory; I just think it's totally wrong.

First of all, instead of saying (using your example) Provoke is WAR+5, party-3, why not just say Provoke is WAR+8? Since aggro is a relative calculation (who ever has the most hate, has aggro), +5-3 and +8 mean exactly the same thing.

The only time they don't is when dealing with zero-hate situations (assuming hate can't go negative, which does not necessarily entail running away, BTW; just a no-attack state). Extreme's theory has already been shown to be provably false in zero-hate situations as it is (there are numerous start-of-battle tests you can run that blow this theory apart), so I don't see any value whatsoever in it.

[Aeynrieus]

I've got a quick question. I've never experienced this, so I'd be interested to know the answer. Let's say Joe, John and Jane are running through Ghelsba Outpost and Joe gets aggro'ed by an Orc. The whole party stops in response to this. Joe gets hit for 5 points of damage. Barring any action on the part of John and Jane whatsoever, what happens? Does the Orc continue to pound senselessly on Joe? Or does the Orc immediately redirect aggro to someone else? Or, thirdly, does he pound on Joe a little while only to "get bored" and eventually go after John or Jane?

I'm currently writing an essay on hate and how it works and the answer to this problem would help to clarify a few things in the essay.

[Icemage]

Extreme's theory does not impress me in the slightest.

Again, personal experience talking here:

Counterexample:

While I was in Qufim a couple months ago I had the unusual experience of having the following party at level 29 in Delkfut's Tower:

WAR/MNK
WAR/MNK
WAR/THF
MNK/WAR
THF/MNK
WHM/RDM (me)

Battle is going badly, so I cast Blink, Divine Seal, then Curaga in battle with Blink active(didn't have Stoneskin yet at this point). Aggro (naturally) goes immediately to me. I recast Blink.

WAR/MNK #1 Provokes. Nothing happens.
WAR/MNK #2 Provokes. Nothing happens.
WAR/THF Provokes. Nothing happens.
MNK/WAR Provokes. Nothing happens.

I finish recasting Blink.

3 combat rounds later I am getting hit. Monster (some sort of Gigas) continues to hit me.

WAR/MNK #1 Provokes. Nothing happens.
WAR/MNK #2 Provokes. Nothing happens.
MNK/WAR Provokes. Nothing happens.
WAR/THF Provokes. Nothing happens.

I eventually die after 19 consecutive attacks(4 Blinks, 2 Blocks, 1 Parry, 3 miss, and 9 hits if I recall correctly) and 8 Provokes (I counted to make sure, because it WAS so strange). If Extreme is correct, and Provoke indeed has any significant subtractive effect, that would have been 8 times that my enmity would have been decreased. During this period I generated the following aggro:

Blink
DS + Curaga
Blink
Regen(self)
Cure II(self)

I'll let the audience figure out how much credence I give to Extreme's theory as a result. While its possible this occurred because of a game glitch, I doubt it.

---

Getting back on the topic of just what Provoke does, it occurs to me that there IS a way to empirically test the scope of the value it assigns.

Test situation:

PLD/WAR partied with WHM/xxx.

Let PLD take some damage (maybe gilfarm or something)

PLD Provokes some target enemy that is really weak and does not attack.

WHM immediately casts moderate Cure spell (maybe Cure II) on PLD.

Count how long it takes for the monster to turn towards the WHM.

Repeat process with different +/- CHR items to see if CHR affects it.

I'll see if I can find a bored PLD to test this with me sometime in the near future - I think it should provide some very interesting results no matter what the findings.


Icemage

[Icemage]

Let's say Joe, John and Jane are running through Ghelsba Outpost and Joe gets aggro'ed by an Orc. The whole party stops in response to this. Joe gets hit for 5 points of damage. Barring any action on the part of John and Jane whatsoever, what happens? Does the Orc continue to pound senselessly on Joe? Or does the Orc immediately redirect aggro to someone else? Or, thirdly, does he pound on Joe a little while only to "get bored" and eventually go after John or Jane?

Easily tested, but I can give you the correct answer from extensive experience.

Barring outside interference, once a monster has targetted a player for attack(assuming that it has accumulated 0 enmity on any players), that monster will continue to attack the player until one of several things happens:

(1) A more "desirable" target appears. This can happen if there are any players resting within range of the monster, as this is a higher priority target for monsters than standing/moving players (assuming that no one has any enmity). This can also happen with other more unusual cases (player with yellow or red HP wandering near an undead monster that was chasing another player, etc.)

(2) The monster completely gets out of follow range of the player. This can happen if the monster stops to cast a spell (Ghost type enemies, Elementals are notorious for this), gets caught behind terrain, or the player is moving faster than the monster (THF using Flee, Ninja wearing AF boots at night). The monster can indeed "lose interest" in this case. NOTE: If you're being chased by a monster, it's usually a good idea to hug walls and run past pillars/rocks/other obstructions, as this will sometimes slow the monster AI down long enough for you to get out of range.

(3) Someone draws the monster's attention with a hate-producing action such as an attack, spell, or ability.


Icemage

[kazthegun]

Originally posted by Extreme
‚΂©ŠO?l?B?B?B‚È‚¢Ž„‚ÌŽžŠÔ‚̉¿’l‚ðŽ?‚Á‚Ä?E‚È‚½‚Ì

nice argument: "stupid foreigners, not worth my time" ?

i agree with icemage's theory, if i'm correct in believing that it's something like this: enmity is a measurable quantity that is added to and subtracted from by various actions, and decreases over time. however, i'm not so sure if there are two types of enmity, the immediate and the residual; it's not that it doesn't make sense, but i just haven't seen anything in-game, personally, that shows there are two types. i've seen a monster's attention bounce repeatedly between tanks, mages, and myself, and i think it's simply a matter of each PC's enmity going up or down depending on their actions. tanks that don't provoke every 30 seconds seem to have a harder time of maintaining hate due to enmity reaching very low levels before receiving provoke's enmity boost.

i think the damage <-> enmity relationship is based upon a percentage of the PCs' and NPCs' hit point totals; that is, if my sneak attack does 20% damage, i'll have more enmity than if my sneak attack does 10%, regardless of the specific numbers. if a clipper hits me for 10% of my hit points, i believe that it reduces its enmity towards me by more than if it had done 5% damage to me. i have no knowledge of monsters' actual hit point totals, so i have no way of correlating this other than my personal opinions, derived from my experiences in valkurm, buburimu, and qufim.

[imac2much]

Originally posted by Icemage
I'll see if I can find a bored PLD to test this with me sometime in the near future - I think it should provide some very interesting results no matter what the findings.


Icemage

<-- Count me in Icemage

As for Extreme's attitude, it's clear that he's elitist and nothing we say will change that. But to counter his (irrelevant) arguments about other import/JP players... I have not met ONE import/JP who is as elitist as him. Then again, it could be that elitists don't want to talk to me :sweat: But even so, I (and several of my friends) get tons of invites from JP parties, so it's clearly not the 'norm' to carry an elitist attitude. Also, they have several times mentioned I was a good PLD, since aggro almost never leaves from me; they have said the same to my PLD friend as well. They have never gotten "disgusted" that I don't understand provoke or something. I don't think everyone shares the same attitude as him.

Thanks for trying to be helpful though, in your own way I guess. But in an analytical debate such as this, saying "this is common sense" and "everyone says so" just doesn't cut it.... especially if people are providing arguments from experience countering your claims. It is equally likely you misunderstood this so-called "common knowledge" that you read from japanese websites. And the way you originally worded it, you DID make it sound as if Provoke only sets your hate to a certain amount, not add hate. If people misunderstood you, don't blame others, it your own fault for incorrect wording.

[DannyForgette]

I have a way to test if provoke lowers someone else's hate. your going to need a THF/WAR and a _ /WAR or WAR/_ doesnit matter.
the THF/WAR does perfect dodge, provokes an enemy(not one that can use magic) and then the _/WAR or WAR/_ provokes. If extreme's theory is right the _/WAR or the WAR/_ will get the aggro of the monster. I dont have a theif at lvl 10 so i cannot do this test.

[Spider-Dan]

Originally posted by DannyForgette
I have a way to test if provoke lowers someone else's hate. your going to need a THF/WAR and a _ /WAR or WAR/_ doesnit matter.
the THF/WAR does perfect dodge, provokes an enemy(not one that can use magic) and then the _/WAR or WAR/_ provokes. If extreme's theory is right the _/WAR or the WAR/_ will get the aggro of the monster. I dont have a theif at lvl 10 so i cannot do this test.

This still isn't a good test. I assume this is based upon the theory that missed hits increase a mob's hate for the target? Even if that's true, if Icemage is correct about decreasing hate over time, there's no way to know to what degree the misses outpace the degenerating hate, if at all.

There's also the issue that unless you have two THF/WARs with identical stats, you don't know whether the target selection is because of timing, or simply that one person has a stronger provoke than the other. If two party members provoke a mob at the same time, it has to choose one of them, so even if everything is equal, there's no way to tell why the mob chose who it did.

[Shiia]

I will try to help aswell. Wether Extream's idea or Icemage's idea is correct matters very little to me. I also want to know the effect of provoke does. Fortunately I have 21 thf, 27 paladin 18 war to experiment with. I have a whm/blm (40/31) friend who I play with. Guess we can run some tests.

[Aeynrieus]

Originally posted by Icemage


Easily tested, but I can give you the correct answer from extensive experience.

Barring outside interference, once a monster has targetted a player for attack(assuming that it has accumulated 0 enmity on any players), that monster will continue to attack the player until one of several things happens:

(1) A more "desirable" target appears. This can happen if there are any players resting within range of the monster, as this is a higher priority target for monsters than standing/moving players (assuming that no one has any enmity). This can also happen with other more unusual cases (player with yellow or red HP wandering near an undead monster that was chasing another player, etc.)

(2) The monster completely gets out of follow range of the player. This can happen if the monster stops to cast a spell (Ghost type enemies, Elementals are notorious for this), gets caught behind terrain, or the player is moving faster than the monster (THF using Flee, Ninja wearing AF boots at night). The monster can indeed "lose interest" in this case. NOTE: If you're being chased by a monster, it's usually a good idea to hug walls and run past pillars/rocks/other obstructions, as this will sometimes slow the monster AI down long enough for you to get out of range.

(3) Someone draws the monster's attention with a hate-producing action such as an attack, spell, or ability.


Icemage

That's precisely what I thought. Thanks.

[ZenXIII]

I always thought of hate as a stack...and the person with the biggest stack is the one getting the aggro.
Monster attacks you, it pops some off of your stack. You attack, it pushes some back on.
I *could* see extreme's theory (i use that loosely, as he doesn't even claim it as his own...which is probably why he can't defend it very well without calling everyone a n00b) working. We all have a hate stack and when a voke is used it pops a bit off of everyones and pushes a decent amount onto the main tank's.
But, I believe Spider said it first, why don't we just say it adds "8" because that is exactly the same as saying it adds 3 to the tank and subtracts 5 from everyone else?
I've been sitting here trying to figure out a way to test this to empirically find out whether voke drops hate on all but one ... but i don't think there is ANY way to prove this that would provide different results than simply a big push onto the voker's stack.
If your hate can't be reduced to 0 through vokes ... there is no way to prove it one way or the other...logically though, why would you waste the CPU cycles to subtract from all stacks, when you can simply add to one. It is more efficient, and (as far as i can tell) does nothing different.
I'd be very interested in hearing what the actual difference is ... or maybe i'm just being thick. I'll go reread.

[Rekiem]

This is the way I think it works:

Monster has a hate list. He basicaly rank everyone with a number, the one who is #1 is the one that he will attack. Whenever someone attack, heal or use an ability, he gets more points and the more points you have, the better your "rank" will be, if youre #1 you will get attacked. This is the way aggro works.

Now Provoke:

When someone use Provoke, it gives them "hate point" so they can score higher on the monster hate list and the tank can end up #1 on that list so he gets attacked. Provoke is based on a stats, which one, I don't know but CHR makes the most sense but it could also be STR.. which would be logical since its the Warrior highest stats. But one thing for sure is that Provoke is based on a stat because otherwise, mage wouldnt be able to cast curaga and other powerfull magic and not get aggro when at low level, those spells get you attacked almost 100% of the time.

This is the most logical and simple explanation.