Awesome testing done on enmity

[Taskmage]

Enmity Testing (Part I)

• Evading or taking 0 damage from an attack causes no enmity loss
• Evading an attack through use of Utsusemi Shadow Loss causes an enmity loss
• Distances between 0'-20' at which an action is used relative to the mob does not affect the hate gained
• The act of pulling grants an extra enmity bonus to the puller

Enmity Testing (Part II)

• There is a decaying component and non-decaying component to hate in actions.
• All actions must leave at least 1 BASE UNIT of non-decaying hate
  • We arbitrarily call this amount of hate the BASE UNIT of hate, i.e. 1 unit of hate/enmity
• Dispel provides 320 units of non-decaying hate
• A hard cap on cumulative hate exists at a range between 9600-10240 units of hate

Enmity Testing (Part III)

• The Cumulative Enmity Cap is exactly 10,000 E
• Volatile Enmity can still be gained even once Cumulative Enmity has been capped
• Hate is decided by the sum of Cumulative and Volatile Enmity (Total Enmity)
• "Zoning" or "Logging Out" does not immediately reset Enmity to 0

Enmity Testing (Part IV)

• The Volatile Enmity (VE) of Provoke is 1800 VE
• The rate of VE decay is roughly -60 VE per second for Provoke
• The rate of VE decay is not related to the current total VE of a player
• The natural Volatile Enmity Decay Rate is -60 VE per second for all actions with Volatile Enmity
• Enmity Gear grants a direct percentage increase on Cumulative Enmity of actions
  • Final CE = [Base CE] x { 1 + [Enmity Enhancement] / 100 }

Enmity Testing (Part V)

• Enmity Gear does not affect the VE Hate Decay Rate in any fashion.
• Enmity Gear increases the VE of an ability in the same fashion as it affects CE - direct % increase.
• The "Enmity Enhancement" of Sentinel is the equivalent of +100 Enmity.
• "Enmity Enhancement" by special status or gear has an "upper cap" of +100 Enmity.
• Adventurer's Dirge does not affect the Volatile Enmity Decay Rate.
• Adventurer's Dirge LVL1 grants a player a straight -10 Enmity Enhancement.

Enmity Testing (Part VI)

• Volatile Enmity does cap and does so at 10,000 VE - the same cap as CE.
• The maximum Enmity any player can reach is 20,000 TE - 10,000 CE and 10,000 VE.
• If an action is performed on a player at a distance 25.0' or greater from the mob, no Enmity is gained.

Table of CE and VE values for non-damage actions (WiP)

This thread started by Ashira on the Blue Gartr forums also contains much good information and discussion not found on the LJ.

In case the livejournal has been updated more recently than this post, all entries on this topic can be found under the journal's “enmity testing!” tag.

Credit for devising, executing and publishing these tests goes to Kaeko and Ashira of Odin's TeamKANICAN. Additional contributors are credited in the journal's posts. Thanks so much guys for bringing this information to the community!

[Kuro]

Wow thanks for the link. couldn;t find this. remember seeing this data baeofre then my pld friend wanted it.

[Armando]

Damn it. Someone beat me to the punch again.

[IfritnoItazura]

There goes my favorite theory on enmity (single quantity, exponential decay). *sob*

[Taskmage]

That was the Kaeko's personal theory when he started too.

[Armando]

I hate to admit it (because I wanted to test this personally) but the testing is sound and the findings correct. I approve.

[Malacite]

very nice find

[LilithAngel]

One question that stuck in my mind (and it may be obvious or not, can't tell) is this: is a 0HP Cure I and a 30HP Cure I the same level of enmity gain or not, and if not, by how much of a factor higher is the 30 point Cure? That is to say, if Dispel is 320 units of (non-decay) hate, then is a 30 point Cure going to be the same in that it'll take 320 casts to take hate, or is it possible that a 30 point Cure is worth exactly that, 30 points? Perhaps somewhere in between?

Also, how much damage is needing to be done by the mob before it has satisfied its hate requirement and shifts its target to the next on the list? If we could find the quantity of hate lost upon taking damage, and how much is lost in correlation to how much damage is taken, then putting that info together with the previous tests presented here, we could theoretically quantify the hate value for every action a player and a monster can take in the game. Hate values for Jump, Tachi: Gekko, Random Deal, and hate loss values for Hurricane Wing, Meteor, Whirl Claws, you name it. It might also help us come a few steps closer to also understanding and better handling "hate-reset" moves as well.

I dunno, just a few idle thoughts.

[Aeolus]

Thanks for the link TM, very interesting read.

Cure for more than 0hp may well generate more than 1 unit of hate but I think from what Ive read so far the only spell they found to not be non-decaying hate would be dispel at 320 units which is also a large amount easier to work with, therefore Cure's units would start decaying after they cast and would be almost impossible to keep track off. That is if Cure is actually a decaying unit that decays to 1 base unit.

[Dymlos]

Even after five years we're still finding formulas in FFXI's system.

[Nuriko]

The tester could always use Dispel as a baseline ... fight a worm, one tosses a silence, one a Dispel, and one a cure or two (preferably just one).

[Armando]

One question that stuck in my mind (and it may be obvious or not, can't tell) is this: is a 0HP Cure I and a 30HP Cure I the same level of enmity gain or not, and if not, by how much of a factor higher is the 30 point Cure? That is to say, if Dispel is 320 units of (non-decay) hate, then is a 30 point Cure going to be the same in that it'll take 320 casts to take hate, or is it possible that a 30 point Cure is worth exactly that, 30 points? Perhaps somewhere in between?

Also, how much damage is needing to be done by the mob before it has satisfied its hate requirement and shifts its target to the next on the list?

Obviously a 30 HP Cure can't be worth a fixed amount of hate, because healing a low-level player for 30 HP generates ridiculously more hate than healing a high level player for 30 HP. That means the hate value of healing X HP while using Cure is proportional to either A) the % of HP recovered to the player; B) the level of the player being healed. One would expect the hate produced from doing damage to work this way as well. It's most likely A. B is questionable, because there's lower level NMs with disproportionate amounts of HP. If B were true, then hitting a Lv.20 mob for 50 out of 400 HP produces the same hate as hitting a Lv.20 NM for 50 out of 1000 HP, which sounds stupid and illogical. That aside, it's also more troublesome to have to come up with a table of values that says "ok, at this level, healing 1 HP creates this much hate, and at this level..." when you could simply say "ok, healing 1% of your HP creates this much hate" and not have any problems whatsoever.

My theory has been that dealing damage and healing damage produce non-decaying hate, and that most other things produce hate that decays and eventually disappears. Speaking from experience as a DD-oriented PLD tank, nothing else seems quite as efficient at generating hate as those two actions, and it would explain why an offensive PLD holds hate so much better than a turtle PLD.

[Callisto]

This excites me, after showing some of Belkin's work to my LS leader he agreed to let me and another RDM take a shot at doing some RDM/NIN main tanking. Our first test will be on Genbu once I can track down a few more Emnity+ pieces.

[tdh]

I was just thinking last night about how much help an Enmity plugin would be. As somebody who's always getting his face bashed in, perhaps knowing where I stand on the hate list would help me save somebody some MP. After more testing is done to determine how much hate is generated from damage as well as damage taken, maybe there will be something.

At least at that rate, I can see where I am on the list and then ignore in on purpose because I like making tanks cry.

[Icemage]

That's a great analysis and states something I've suspected for a very, very long time.

In fact, I talked about this exact situation as far back as January 2004
Provoke - Analyzed

Good to see my observations back then were (mostly) correct.


Icemage