Enmity testing - Still Going!

[arshesney]

Intuitively, death has to be a full reset. We know now that zoning and logging don't. I believe shut down is a full reset though.

I would be surprised if shutdown were different from logout except in that it means you're logged out for longer. It looks like you only tested logout followed by immediate logon; it might be interesting to test staying logged out for, say, a minute.

My belief is the mob has to "look" for you for hate to be reset. I have no test results to prove it, but we have had many times where mages logged for hate and when they logged back in they got smacked. On the other hand we have had mages log out and log back with seemingly full hate reset.

Ex1. Tank is kiting, mage logs out, tank goes down, mob then looks for the mage, doesnt see it, hate for that mage gets reset, and mob goes to the next person on its hate list.

Ex2. Tank is kiting, mage logs out, mob stays focused on the tank the entire period of the logout, doesnt look for the mage. Mage logs back in, tank goes down, mob then looks for the mage and sees it, attacks mage.

Like I said, this isnt the end all because I have no proof, but these 2 examples have happened to us so this is the logical explanation I have come up with. Of course it seems as though death is full hate reset, maybe a mob can tell when he has defeated someone and hate is completely reset, on the other hand if they die from a link, even though they were not defeated by the mob in question, hate would still be reset - I have no explanation of that one.

I've had similiar results in Campaign, and saw that death works same way as logout, mobs has to look for you for hate to reset.

First time happened it was a "wtf" situation, but I've been able to repeat it every time.
In Jugner as WHM: I slept a mob (of course not close to any player/npc), flashed another then ran other side of the tower and died. Made sure to die and RR out of the original slept mob range (over 30') and waited for mob to awake.
Well, every single time (provided other players did not interfere) that mob hunted me down regardless of my death.

[Edrick]

I've had similiar results in Campaign, and saw that death works same way as logout, mobs has to look for you for hate to reset.

First time happened it was a "wtf" situation, but I've been able to repeat it every time.
In Jugner as WHM: I slept a mob (of course not close to any player/npc), flashed another then ran other side of the tower and died. Made sure to die and RR out of the original slept mob range (over 30') and waited for mob to awake.
Well, every single time (provided other players did not interfere) that mob hunted me down regardless of my death.

This seems to imply that the hate list for mobs is part of the mob itself. It probably keeps anyone's name who had any interaction with it or with anyone on the list ordered by total enmity, and can really only check the top name on the list. I'd bet if you did the same test and waited to RR after the mob woke up, then it wouldn't go after you anymore if no one touched it in the meantime.

I'm guessing that when a player's name on the list gets to the top, it constantly checks "Is the player dead/logged out/in a different zone?" and if the answer is yes to any of those, then it wipes that name off the list and goes to the next one.

[Valisk]

I've had similiar results in Campaign, and saw that death works same way as logout, mobs has to look for you for hate to reset.

First time happened it was a "wtf" situation, but I've been able to repeat it every time.
In Jugner as WHM: I slept a mob (of course not close to any player/npc), flashed another then ran other side of the tower and died. Made sure to die and RR out of the original slept mob range (over 30') and waited for mob to awake.
Well, every single time (provided other players did not interfere) that mob hunted me down regardless of my death.

I'd bet if you did the same test and waited to RR after the mob woke up, then it wouldn't go after you anymore if no one touched it in the meantime.

I think that's the point he proved indirectly.

[Spider-Dan]

Intuitively, death has to be a full reset. We know now that zoning and logging don't. I believe shut down is a full reset though.

I would be surprised if shutdown were different from logout except in that it means you're logged out for longer. It looks like you only tested logout followed by immediate logon; it might be interesting to test staying logged out for, say, a minute.

My belief is the mob has to "look" for you for hate to be reset.

I approve this message.

Taking another example, if you sleep a skoffin and then "zone" through to MJSP, unless the mob wakes up while you are in the staging point, it will still remember you if you "zone" out. If logging is considered the same as zoning, then I would imagine that the mob remembers your hate level until you're at the top of the queue, at which point you are simply removed from the queue if you're not available to attack.

edit: Taking the dying example from above, it looks like death might work the same way, too. If you are killed by a mob, then it removes you from its hate queue, but if you are killed by something else, you aren't removed from other mobs' queues until they actually come looking for you.

The question I'd like to answer is: does VE still deteriorate while you're zoned?

[LeifofKujata]

I approve this message.

Taking another example, if you sleep a skoffin and then "zone" through to MJSP, unless the mob wakes up while you are in the staging point, it will still remember you if you "zone" out. If logging is considered the same as zoning, then I would imagine that the mob remembers your hate level until you're at the top of the queue, at which point you are simply removed from the queue if you're not available to attack.

edit: Taking the dying example from above, it looks like death might work the same way, too. If you are killed by a mob, then it removes you from its hate queue, but if you are killed by something else, you aren't removed from other mobs' queues until they actually come looking for you.

The question I'd like to answer is: does VE still deteriorate while you're zoned?

Well in this example you haven't acutually zoned out, so you never leave the mobs hate list until it wakes up(assuming they cant "look" for you asleep). Passing through into the SP isn't zoning, even though you get a CS and the mob can no longer attack you. I do agree with the rest though, seems a mob has to be able to "look" for you to either lose hate or maintain.

[Spider-Dan]

Well in this example you haven't acutually zoned out, so you never leave the mobs hate list until it wakes up(assuming they cant "look" for you asleep). Passing through into the SP isn't zoning, even though you get a CS and the mob can no longer attack you. I do agree with the rest though, seems a mob has to be able to "look" for you to either lose hate or maintain.

That's why "zone" was in quotes.

In any case, entering the SP seems to have the same effect as zoning if the mob isn't asleep.

[fuergrissa]

Just as an example of an oddity it appears that KV doesn't work under the same principles as normal mobs. CE seems to decay very quickly without being hit being the major oddity. Another weird scenario yesterday KV popped and was claimed by PLD/NIN, he kited using flash and some tools for hate, When I arrived on NIN he dropped to change job while I kited, About 15 min later the PLD had zoned to east alteppa, Warp to metelworks,, Bastok Market, Mog House, Change job, and then come back on same path. Whole time I was voking and tooling the mob as I was kiting, the RDM were spamming Gravity and Curing etc. as it was kited. After the original PLD (now on WAR) zones back into west alteppa my game crashed and I was D/c'ed. KV completely ignored the mages and started trucking off to the North after the original PLD and drw him in when it got about half way to him. Even after logging back in and voking/tooling over and over KV would still turn intermittenly and go after him. and if anyone cares, earth crystal ftw >.>

[sleekmotorwurkz]

Just as an example of an oddity it appears that KV doesn't work under the same principles as normal mobs. CE seems to decay very quickly without being hit being the major oddity. Another weird scenario yesterday KV popped and was claimed by PLD/NIN, he kited using flash and some tools for hate, When I arrived on NIN he dropped to change job while I kited, About 15 min later the PLD had zoned to east alteppa, Warp to metelworks,, Bastok Market, Mog House, Change job, and then come back on same path. Whole time I was voking and tooling the mob as I was kiting, the RDM were spamming Gravity and Curing etc. as it was kited. After the original PLD (now on WAR) zones back into west alteppa my game crashed and I was D/c'ed. KV completely ignored the mages and started trucking off to the North after the original PLD and drw him in when it got about half way to him. Even after logging back in and voking/tooling over and over KV would still turn intermittenly and go after him. and if anyone cares, earth crystal ftw >.>

That is a good example of what I was talking about, Since you were at the top of the hate list when the PLD zoned, kv did not look for him but he probably still had a lot of CE built up. Id assume that the rdms still had less then him. When you dc'ed The formor pld (now on war) was at the top of the hate list. KV looked for him and saw he was in the zone and smacked him, until that point he was focused on you and didnt even look for the pld so never even realized he left to begin with. But as far as the CE/VE degrading, im assuming most of the enmity testing that ashira and kaeko did were on primarily the same mobs. It is possible that different mobs have a different hate loss model, possibly just notorious monsters, but who knows.

[Gunslinger]

These last few posts make a lot of sense. Is it possible that the VE/CE decays differently when the mob is "looking" at you (only if it has "checked" you on its list)? That may explain differences in decay and why you keep hate on a mob after zoning, etc. Just a thought.

Ps. Sorry if this has been brought up- my understanding of all the testing was that decay happened once you established enmity, not on mob focus (thus this would be a bit different).

[LeifofKujata]

I've kited KV on different occasions, and found during earth weather, its hate loss was extremely fast. Conversely during no weather hate loss was like a normal mob. Its kinda interesting that the pld(war) had retained stored hate even after being away from the zone for many minutes.

[Gergall]

I see in the enmity blog that Sirvente information was removed due to a potential flaw in the testing process. Is this believed to be bugged or is it just totally up in the air right now?

[Ashira]

I see in the enmity blog that Sirvente information was removed due to a potential flaw in the testing process. Is this believed to be bugged or is it just totally up in the air right now?

We were unable to reproduce results of our first test, and haven't had a chance to figure out what precisely caused the results we got in our first test.

[shaddix]

i'll paypal you each $1 to do some more tests

[VZX]

Maybe I missed it...
how the DoT spells enmity works?
Is there enmity gain per 'tick' ?

[Kaeko]

Maybe I missed it...
how the DoT spells enmity works?
Is there enmity gain per 'tick' ?

Never formally tested this, but just anecdotally base on using Poison in some of the tests, I'd say no, it's just the original cast that gains Enmity.

[Kaeko]

I see in the enmity blog that Sirvente information was removed due to a potential flaw in the testing process. Is this believed to be bugged or is it just totally up in the air right now?

We were unable to reproduce results of our first test, and haven't had a chance to figure out what precisely caused the results we got in our first test.

We probably messed up, even though both of us we pretty sure we did it right when we got that original result. Alas, you have to scrap anything you cannot reproduce. I think right now that it probably reduces the amount of Enmity you lose when you get hit, but that's just a guess.

i'll paypal you each $1 to do some more tests

Been really busy at school, I've only logged one once in the last maybe 2 weeks and that was to chat.

[Gergall]

I see in the enmity blog that Sirvente information was removed due to a potential flaw in the testing process. Is this believed to be bugged or is it just totally up in the air right now?

We were unable to reproduce results of our first test, and haven't had a chance to figure out what precisely caused the results we got in our first test.

We probably messed up, even though both of us we pretty sure we did it right when we got that original result. Alas, you have to scrap anything you cannot reproduce. I think right now that it probably reduces the amount of Enmity you lose when you get hit, but that's just a guess.

I doubt this was the issue but I'll toss it out there. If a Bard gives someone a song while using an instrument for a duration bonus (song A), then a song without a duration bonus (song B), and then plays any other song (song C), song C will overwrite B, not A. This is because B has the shorter remaining duration, even though A was played first. Bards know this but I bet many other players don't.

Do I think that this is what happened to you? Not really, as I can't imagine that you were mixing in other songs with Sirvente during your test.

EDIT: And on the topic of unlikely explanations, maybe SE fixed it.

[Kaeko]

http://masurao.cocolog-nifty.com/blog/

If anyone feels like reading Japanese this guy has continued parts of our original research.

* HP, maximum nor current, does not affect enmity gain

Also lots of detailed graphs on PLD vs. RDM tanking. I personally think it's a bit flawed b/c it doesn't account for the fact you can't spam 100% of the time because the mob is usually trying to hit you, but it's interesting nonetheless (maybe works for JoL). I personally don't find simulating how different tanks will compare against each other very useful since on hard stuff you have to worry about survival as well as capping hate. This is one of the main reasons I feel a PLD still outperforms others on really really hard mobs despite not looking quite as good based on graphs and simulations.

He also is not comfortable with the fact an "English speaker" figured this stuff out and encourages people to verify and continue the research.

Interesting nonetheless.

[Wraith]

He also is not comfortable with the fact an "English speaker" figured this stuff out and encourages people to verify and continue the research.

JP racism ftw

[Not Kuno]

He also is not comfortable with the fact an "English speaker" figured this stuff out and encourages people to verify and continue the research.

http://www.savvy-baby-gear.com/wordp...aby-crying.jpg