Curse

Servant of Kali · Jul 20, 2009, 08:07 PM // 20:07

I've read it's related to health.

But can someone be more specific? Does it matter total health, or does it matter current health? As in, if party members gets hit down to 200hp by AoE even if he wasn't first target, does that mean monsters will switch to him?

Is there any other algorithm besides hp?

I'm asking because I'm curious whether it's good to have a "bait" party members, the one with intentionally lower hp but buffed in various ways.

Arkantos · Jul 20, 2009, 08:27 PM // 20:27

From wiki:

Quote:

Gaining aggro or holding aggro means increasing an ally's threat level, therefore making that ally more likely to be attacked by the AI. Known ways of creating aggro are:

* Being within a foe's danger zone.
o Holding a wand or staff as opposed to martial weapons such as a sword, bow, or scythe creates additional aggro. Only applies to allies within the foe's danger zone.
o Having lower armor and/or health than team mates. Accordingly, having higher health/armor will have the opposite effect. Only applies to allies within the foe's danger zone.
* Attacking a foe or affecting it with a skill.
* Healing or enchanting an ally which already holds aggro.

12 chars

Braxton619 · Jul 20, 2009, 08:51 PM // 20:51

It's usually the lowest HP or the monks. They tend to also go for the ones that are enchanting you.

Icy The Mage · Jul 20, 2009, 08:52 PM // 20:52

Yep, when you're 600/smiting that's why your hero will very rarely get attacked (if they're equipped with max armor that is).

Another trick if you want aggro to focus on you rather than someone else (especially if you're using Shadow Form) is unequipping your chest piece. Whether it's the hawt digital pixels or lack of armor-level monsters are attracted to I will never know, but it seems to work fairly well.

Servant of Kali · Jul 20, 2009, 09:04 PM // 21:04

They look at inherent armor or buffed-armor too? As in, would Dervish with Conviction get less attention than without it?

Arkantos · Jul 20, 2009, 09:12 PM // 21:12

Quote:

Originally Posted by Servant of Kali

They look at inherent armor or buffed-armor too? As in, would Dervish with Conviction get less attention than without it?

Correct. That's why when playing a character with SY, you're generally going to be getting attacked more because you have 100 less armor than everyone else.

Bristlebane · Jul 20, 2009, 09:32 PM // 21:32

Yeah that should be easy enough to test out as well. Get a ranger with Barrage + Save Yourselves. Usually rangers rarely gets attacked first, I noticed a huge difference in working on a ranger survivor in comparison to other professions.. I assume that inherent +30 AL vs elemental damage plays a huge part.

Back to topic.. a Barrage/SY ranger would end up with ~70-100 armor less than rest of your team and should most likely attract a whole lot more aggro. While at it, could add in dmg conversion like Conflagration/Greater Conflagration to utilize that +30 inherent.. which is still 70 less than rest of your party.

Owik Gall · Jul 20, 2009, 09:33 PM // 21:33

Quote:

Originally Posted by Arkantos

Correct. That's why when playing a character with SY, you're generally going to be getting attacked more because you have 100 less armor than everyone else.

I would think that the monks and others in the back line can wield martial weapons to help lose targeting aggro in addition to the SY! spamming. That guess is based on the example from the wiki you've posted earlier. So, would this be a good idea, you think?

gameshoes3003 · Jul 20, 2009, 11:30 PM // 23:30

So a martial weapon helps eh?
I've always wondered why monsters would attack the other monk...
I usually have an axe with the +5 energy mod out while most of the time I see other monks with staffs.

qvtkc · Jul 21, 2009, 08:31 AM // 08:31

I read that health degen also helps to attract aggro, and that regen helps you get rid of it. One more reason to dislike Mending wammos

Chthon · Jul 22, 2009, 02:48 AM // 02:48

AFAIK, the monster decision-making process goes something like this:

1. Make a list of all potential targets.
2. Assign a "score" to each potential target.
3. Attack the target with the highest score.
4. Repeat.

Some details:

The list of potential targets includes:
* Anything within the monster's aggro bubble.
* Anything within the aggro bubble of something within the monster's aggro bubble.
* ANything that attacked or cast upon the monster.
* Anything within the aggro bubble of something that attacked or cast upon the monster.

The score is a weighted composite of a number of factors. Listed in what I believe is descending order of weight:
* Currently being hit by the monster
* Proximity to the monster
* Life/Armor (probably resembles BADITTKY calculation)
* Fleeing/Kiting (totally out of monster's range)
* Being the one to initially aggro the monster
* Degen/Regen

There also seems to be something that causes a "go for the monk" effect. Some people theorize this is based on weapon type, while others claim there is an inherent ordering of which classes to attack. I've never seen a well-designed experiment to determine which is correct. Whatever this factor is, it seems to have a similar weight to life/armor.

A small note on "currently being hit by the monster": Rather than being a linear variable, the AI seems to use a "satisfice" threshold formula. If the monster is doing "enough" damage to its current target, this factor gets a huge weight; if it's not, this factor gets a rather weak weight.

The decision process seems to repeat every second or so. However, the "currently being hit by the monster" factor is usually so strong that monsters only change targets when something stops them from hitting so that their current target loses its "currently being hit by the monster" status -- usually triggering monster flee or KD-ing the monster.