Heroism usage.

[Protico]

EJ Thread (http://elitistjerks.com/f79/t49476-math_bloodlust_heroism_timing/) - Binkel's Blog (http://elitistjerks.com/blogs/binkenstein/234-bloodlust_execute_talents.html).

Linking for people who were asking. Basically what they are saying is that heroism below 35% might increase DPS for certain classes, it also shortens the duration of already increased DPS. Popping it early will make sure more people are up and more cooldowns are available to use during the trinket, because additive and multiplicative abilities benefit most from heroism.

Obviously on some fights it makes more sense to pop the heroism at specific times, that is an entirely different discussion.

[crivens]

is it just warriors and mages who benefit from heroism at the end?

Is threat ever a problem for our top dps early in the fight?

I would think that whatever % we do decide to use heroism, it is best to be consistent that way the dps knows its coming and can save their cooldowns for maximum synergy.

[Syrio]

That's why we've been using 35%, mainly so that people know when to use it and when to expect it.

[Protico]

Affliction locks also get a benefit below 35%, but that doesn't necessarily mean that using it below 35% benefits them more. Haste, especially for cooldown based classes, has significantly diminishing returns. For certain classes, enhancement shaman being a notable one, we have certain short cooldowns which benefit a lot from heroism, and if we blow heroism at 95%, and use our cooldowns then, there is a good chance we will have them up again during the fight. Then again if we know heroism is coming at a predefined time later in the fight we are inclined to save them because we might not get the most use out of them if we blow them early and miss the heroism opportunity. That is the whole point of those posts, there are more considerations than how hard warriors will be able to execute.

[Lemmiwinks]

need pro nigma strats with 12 warriors pray for 20% then watch the boss fall over

[meer]

The issue about using bloodlust early or late has little to do with getting maximum dps out of x raid. So, even though using it early in a theoretical setting (assuming X boss, has Y life, and raid dps is is Z, class comp is AA, and a lot of other small crap to factor in) might be a little bit better it wont' matter.

We use bloodlust depending on the encounter. Some encounters favor using it early(twins), some late(muru), some between certain phases(kael thas). So in the end, it really doesn't matter. It all depends on the situation and encounter regardless of maximum raid dps output.

[Tatsuo]

I think this debate is when to use it on bosses that will take at least 2 minutes to kill but the strat never changes. Thad, patch, and so on.

[Protico]

The debate is about bosses which don't favor a particular heroism point because of strat, and the consensus is early is better.

[Lemmiwinks]

this just in: heroism is situational

[Ibalis]

Next time tell me heroism is coming off the bat so i don't blow a 5min cd, kthx

[Protico]

Next time tell me heroism is coming off the bat so i don't blow a 5min cd, kthx

Hey, next time Heroism will be off the bat.

[west]

bump for isarac!

[Ghanos]

bump for isarac!

Thanks for posting this. I did some of my own analysis and basically agree with the EJ thread conclusions. However, they don't explain clearly what the trade-offs are of early or late hero. This is because it's an algebraic analysis and not a statistical model analysis.

I tried to Simcraft it this weekend, but couldn't figure out how to override the heroism timing. By the way, Simcraft which is a Google collaboration project of tons of geeks has heroism set to Boss Time To Die <= 60 seconds. Since there are so many eyes looking at that, I doubt it's wrong.

The EJ points are basically that the late heroism benefit is small, and if not all DPS is alive then it's a DPS loss.

From a min/max standpoint, though, my Festergut (where I requested a late hero) model estimated about a 1.5% raid DPS benefit to a 35% hero versus hero off the bat. This assumes a multiplicative heroism benefit, i.e. hero+CDs improves raid DPS by 40%.

If you model heroism as additive raid DPS, such as heroism adds 25,000 raid DPS then it doesn't matter when you hero.

Is the 1% to 2% benefit worth it? I think so for farm content. As a top guild, we expect a lot of min/maxing like professions and what not. Looking for those 1% to 2% improvement opportunities are the challenges that I think make the game interesting.

To make it easier to manage a variable hero, we might need to have all DPS download a time to die addon. Instead of basing hero off a % (which would vary based on TTD), we could also shoot for time to die <= 60 secs.

[farmergregor]

An extremely early hero (ike 95%) lets me line up all my cooldowns/trinket procs.

[Florun]

An extremely early hero (ike 95%) lets me line up all my cooldowns/trinket procs.

Yeah gargoyle + full trinket procs + hero = like 20k extra damage atleast

[meer]

lust is only good for dps and dps alone. using it for certain phases or for heals is rubbish. so when we determine when to use lust we use it when it gives us absolute maximum dps gain and not maximum gain for encounter it self.

i'll just quote myself cuz i rok


We use bloodlust depending on the encounter. Some encounters favor using it early(twins), some late(muru), some between certain phases(kael thas). So in the end, it really doesn't matter. It all depends on the situation and encounter regardless of maximum raid dps output.

[Florun]

hey how am i supposed to parse without heroism for max dps

[Corrosion]

you will hero when all MY procs are up so i can summon gargolye and you will like it meer!

[Zahlan]

100% hero for max shadowfriend power

[Brades]

I was looking through the world of logs top dps parses for Saurfang and I noticed the guilds that are using heroism at 95% are killing saurfang somewhere between 2:15 and 2:45. Most guilds with over a 3 minute kill are waiting to use heroism until 30% or so. Not that it really matters either way, but is interesting to see that there is a noticable increase in total dps this way. Probably due to everyones initial CDs lining up with heroism.

Because the fight takes us just over 3 minutes it's an odd fight for CD management, mainly berserk. If I berserk right when we start I may be able to get a 2nd berserk in, but its not optimal to just start off the bat with a berserk as you really want to maximize the number of shreds you get during one. It usually takes me about 10 seconds into the fight to have all my bleeds up and thats when I berserk. If we are going to get the kill down under 3 minutes than I should save my berserk for heroism.