Enmity Testing

[Kerberoz]

The kaiden HQ mega-boss actually seems to attack recently raised people... possible?

[arus2001]

The kaiden HQ mega-boss actually seems to attack recently raised people... possible?

Well, if the measure of "weakest" might go by HP, being Soul Sick would naturally have your HP diminished. Was it happening while they were still weakened or some point after? Maybe built up hate doesn't wipe upon death, but instead only when the mob's completely idle.

[Syntex]

The buff you get when you were recently raised might be a priority target.

[Kerberoz]

Well, if the measure of "weakest" might go by HP, being Soul Sick would naturally have your HP diminished. Was it happening while they were still weakened or some point after? Maybe built up hate doesn't wipe upon death, but instead only when the mob's completely idle.

He kept running to people 1-2 seconds after raise, oneshotting them, and running back to the tank, lol. Must be either low HP or the raise effect.

[Saxon]

He kept running to people 1-2 seconds after raise, oneshotting them, and running back to the tank, lol. Must be either low HP or the raise effect.

or a shitty tank

[arus2001]

or a shitty tank

I'd be more inclined to believe it's a specific AI choice. Fighting Lepharist Mages, I've noticed they'll try to Soul Freeze a Chanter buddy no matter how much hate I have. Grinding on KHQ Krall some have this Successive Strike/Rapid Succession combo of moves where the first is either a partial hate drop on the tank or the game telling the mob to focus on #2 (usually me) and whack them with RS for 2.5k+.

[Saxon]

I'd be more inclined to believe it's a specific AI choice. Fighting Lepharist Mages, I've noticed they'll try to Soul Freeze a Chanter buddy no matter how much hate I have. Grinding on KHQ Krall some have this Successive Strike/Rapid Succession combo of moves where the first is either a partial hate drop on the tank or the game telling the mob to focus on #2 (usually me) and whack them with RS for 2.5k+.

Yeah but you know, you have no hard evidence to prove any of this. Maybe you just didn't have as much hate as you thought you did.

[arus2001]

Yeah but you know, you have no hard evidence to prove any of this. Maybe you just didn't have as much hate as you thought you did.

Every single time we fought them? Doubt it.

Edit: More specific in the duo working on Lepharists.

Apprentice Mages don't have a whole lot of HP, but as with most magical mobs, they can hit hard. Magic Missile could do 1500+, but ideally I'll have them knocked down mid-cast. Normal melee from them can do anywhere from 300-600.

At 40, I'd initiate the fight with Cleave and chain into that, possibly getting a third whack. Then I'd use Provoke III. Severe Weakening Blow, Weakening Severe Blow, then my 3-hit chain to knock them down and land Crippling Cut then use my 2 AoEs. By this point they should be at 3/8 HP or less, which is about the point they begin with Soul Freeze. I also had two Enmity traits at this level (3 at 41). Relative to the Chanter's contribution, they wouldn't have engage until I started on my 3-hit, wouldn't have cured me because I had more than enough HP still, and perhaps did 2/8 of that remaining HP in damage since I'd still be waiting on CDs.

[RKenshin]

Honestly, the way I've seen things lately it should be fairly obvious when the second hate table comes into play. The dialogue of the mob always changes to something particular to suggest it. Or really any dialogue at all does.

In MM, the raiders always yell "It's your presence that annoys me the most!" or something similar as they run to a random group member and charge up a spell/attack. IIRC it's usually a sorc.

[Kaeko]

Disclaimer: I don't play Aion and this is a long post which may have glaring errors in thought due to the fact I don't play Aion.

************************

First ranting on actual testing:

My advice for testing something like this is to actually just be patient and play through the game (at least to high levels) and observe what you see. Once you reach high levels and actually have access to "all" of your potential abilities it's easier to get an overall feel for how things work. Having a high level toon also makes it easier to screw around with easy mobs.

The problem with the way my enmity testing is written is it's written like a science journal almost, which is frankly a backwards lie (yes real journals do this). They basically write the paper as if they knew all along this is what they were looking for when really they spent years dicking around usually and finally found something after observing countless failures. People write them this way because it's easier for someone that starts from scratch to understand what you did, but it's not how you would actually test something. It takes a ton of time observing.

So basically I would just note basic observations (e.g. this move "X" seems to generate a lot of hate...) and avoid quantifying anything. Something like the "hate-listing" system mentioned is incredibly interesting and it probably took a lot of observing to note that. I'd also try to keep a look out for things that can do:

(1) somehow mitigate damage you take to 0 or only misses (as you probably have to null the enemy's attack damage to get an initial control down).
(2) you think generate large amounts of threat/enmity/hate without hurting a mob (I believe Aion has a provoke equivalent yes?)

************************

Other thoughts:

In general just by trolling a few threads, it seems the system is far more complex and mob-specific than FFXI's (which as complex as it looked years back, is very universal to nearly all mobs in 11, making it very easy to test and apply). Lots of mob specific and unique targeting AI makes testing very hard and the results less relevant. For instance, almost everything known about FFXI's enmity system is based on tests done lvl-1 mandies, the weakest mob in the game. We apply some observations we saw on the weakest mob in the game to all our endgame mobs - that's practically a gift that you probably can't pull in this game.

Also, on a more meta-game related note, there is usually a very obvious tanking strategy or class in each MMO (in FFXI's case, PLD/WAR), but often times it's not the best once people filter though enough mechanics. I'd try to just observe and mess around with unconventional tank ideas. For instance, the fact that your main tank class in Aion appears to not be heavily based on cures (forgive me if I'm wrong, but Templar seems to have a lot of provoke and counter block moves for hate) leads me to at least suggest a Cleric based tank or some variation.

Another thing I've noticed reading about Templar is that there are quite a few enmity generators based on block reactions or other things that require the enemy to be focused on the Templar to maximize your hate generation. A good system imo since it places emphasis on hate control by everyone else (someone stealing hate doesn't just get himself killed, but hurts the Templar hate generation and can lead to wipe).

But this system would appear to have some obvious downfalls - (1) you're always a couple hate steals from potential wipe if your generation requires you to have hate, and (2) difficulty using multiple templars since only 1 can have hate. Whoever has it will have an easier time keeping it basically. Interesting system that really deviates from how FFXI worked.

[Syntex]

Join us and test it, you know you want to :D

I enjoyed your FFXI data about enmity and it was really helpful but I must ask in the time before you can boost stoneskin to a 5+ minute duration buff how did you manage to pull off 300+ something cure's before SS dropped and reapplied messing up the test or did you already calculate the hate generated by a stoneskin cast and factor it out?

Reason is stoneskin lasts 4 minutes and if we are to find the values of spells hate level we need it to last longer but reapplying it could skew sample results.

[Kaeko]

Join us and test it, you know you want to :D

I enjoyed your FFXI data about enmity and it was really helpful but I must ask in the time before you can boost stoneskin to a 5+ minute duration buff how did you manage to pull off 300+ something cure's before SS dropped and reapplied messing up the test or did you already calculate the hate generated by a stoneskin cast and factor it out?

Reason is stoneskin lasts 4 minutes and if we are to find the values of spells hate level we need it to last longer but reapplying it could skew sample results.

LVL-1 mobs usually can only hit for more than 0 damage on a really big fluke critical hit. It happens really rarely. We just skin or phalanx to avoid those because it's really annoying to redo your tests. Also, even though we didn't know the formula for losing CE when taking damage then, I knew just from observation that if a mandy hit me for 1 it was -2 CE and -1 CE if it hit Ashira. So really if it did hit us, no big deal I can back calculate.

When I tried testing a lot of things were simple observation at the time. I feel like you can't just dive into testing unless you have a fairly strong intuition about the system, so the beginnings are almost like an art more than a science.

Ultimately your goal is to quantify 1 ability or move. If you can do this, everything else just falls in place usually. You want to find 1 thing that basically represents your basic "unit" in which you will measure all other things. For instance, the absolute most technical term for CE is "the amount of non-decaying enmity generated by curing a player for 0 HP". That was what our unit was based on and even though it's universally assumed that this is the lowest unit, we didn't know this at the time. If you can make a short list of candidates for this basic "unit" that might help you.

After you've determined a basic unit, you want to quantify 1 move. In our case, it was Dispel (which turns out was a very poor choice due to how high CE it was). But before you can do this, you need to remove all factors which can change the values of your base unit or your move to be quantified. The "Hate-Listing" phenomenon is an example of something that can screw you up, so before you begin doing a formal test, you would have to do something like kill the mob so many times you're almost certain you're capped on this list or something. Other things that can screw you up potentially are level differences, skill levels (if these even exist), mob AI (if the mob isn't always on the guy on the top of the list you're in trouble), and decaying hate (no promises there is even a non-decaying hate component). These are all things observation can help you flush out.

Finally, regarding your stoneskin question, you can put it near the top of your list of things to quantify to help you test, but remember that there can be a ton of components to the system. For instance, in FFXI, we had to at least be aware of both the CE and VE components in order to safely use Stoneskin in the middle of our tests. It was also helpful that FFXI skin gives the same amount of enmity every time - this is not something you can assume in AION. Is "stoneskin" self-cast in AION? If you can just get a 3rd party to help you during the test.

[Abandon]

Yes, stoneskin is self-cast.

[Hitoshura]

This shield dropped, it is the first piece of Enmity gear I have seen around those kinds of levels.

Hirmilden's Shield - - Aion

Unfortunately I passed on it, I have a very similar shield +7'd, would be a waste but damn I did want to test that Enmity.