Waveryn

I will say it, "Some people can not dance", therefore BL early contributes damage.
Another "Early, unless needed otherwise", I have not seen mentioned is for any fight with 6-10 minute enrage or none at all, and more than one shaman is present "Early BL" until 3.1 nerf can be used again for a much better dps than popping it late/35%.

 

Endus

Yes, that's basically what I was getting at, too, the timing is much more critical from meshing with everything else, than with any Execute phase. If there's an Enrage effect like Maexxna or even Grand Widow Faerlina (assuming you're going for the achievement), holding off until the enrage hits is advantageous because the part of the fight is shortens is the one with the highest damage on the tank, and it also makes it easier to heal through, as noted. Otherwise, like Binkenstein is saying, pop early to coordinate trinkets and ensure everyone's up and can benefit.

And yes, when I did my math earlier, I added my bonuses additively not multiplicatively. A lot of the percentile bonuses in WoW work out that way, but it may have been a poor choice. Even so, my point is still relevant; the difference is tiny to the point of insignificance, from a pure DPS point of view, so our timing of Bloodlust/Heroism should be based on other factors entirely.

Namely, before anyone dies, early enouigh you can coordinate trinkets, unless you're saving it for an enrage-type phase you need the extra speed to burn through/healing to survive.

 

Perzyx

I'd like to thank the OP and Binkenstein for illustrating the effectiveness of BL timing. Despite all the logic that spells out why it is better to use BL earlier in a fight, we would (in my old guild), use BL after 30% on most every boss. Not because we had a well thought out reason for it, but because no one stopped to think about it and followed suit on how many other guilds choose to use BL. I am sure there are many other guilds doing the same thing.

Like it was stated in your post that different classes will yield different results on when BL is used. And one of your points is that it doesn't matter how each class benefits individually, what matters is the overall raid effectiveness of bringing the target to execute range faster. However, I wanted to give a Hunter perspective and show how BL used at the end of an encounter is generally counterproductive for Hunters to further back up your theory.

Our execute talent is Kill Shot (15 sec cooldown) and can be used when our targets health is below 20%. If BL was an Attack Power buff we would want it to be used at 20% to buff our hardest hitting talent, however Bloodlust is a haste buff. For us that only affects our Auto Shot, and to a slight degree for non haste capped Hunters our Steady Shot. Because haste does not affect our CDs or GCDs, BL does NOT allow us to fit in more Kill Shots. In fact, it causes us to lose Kill Shots if used in the last 20% of target health because the raid is burning through that portion of the fight faster (less time spent in execute range) thus lowering our overall DPS for the encounter.

In addition we generally find it best to use our talents/ trinkets early in the fight so we can use it again as they come off of CD (varies by length of encounter of course). It is ideal for us to time these (Call of the Wild, Rapid Fire, Trinkets) it with a BL to further buff our Auto Shot dmg. By using BL late in the fight it is hard to time their CD's to keep them all together with the BL. By saving the abilities & trinkets till later in the fight we will not benefit from being able to use them more than once in the encounter (once again depends on the length of the fight).

So, in end, for Hunters it is much better if BL is used early in an encounter, preferably near the start.. just waiting long enough for the tank to get ahead on threat.

 

galzohar

Time-based enrages have no effect on optimal bloodlust/heroism usage. %HP-based ones do - if the "enraged" phase is the challenging part of the fight, rather than the overall fight duration, obviously use BL at (the start of) the enrage phase with all trinkets and cooldowns.


A few corrections to things I've said:

Syncing cooldowns can actually be possible every time for everyone if you know the exact fight duration and use bloodlust/heroism so that it ends right as the boss dies. In practice, if you apply proper statistics and everyone play properly, this can actually be done close enough to perfectly. However this just cancels out an argument against late bloodlust/heroism, and doesn't actually give a reason to use late bloodlust/heroism (at least not by itself).

On malygos when you factor in sparks it gets somewhat more complicated, what I said only accurately applies for no sparks (or even distribution of sparks). Basically in this case you should use it in the phase that gives the most relative increase to raid DPS. That is, if your ability to control sparks well makes up for the time spent in the vortex, use it in P1. Otherwise, use it in P2. Generally if you can pop a 2nd spark along with bloodlust/heroism immidiately after a vortex that would be an optimal time for bloodlust/heroism. You'd have to have some pretty bad spark control and heroism/bloodlust judgement for using it on P1 to not be optimal.

However the idea behind the malygos example still stands - if a fight has 2 phases that shift on a certain % of boss HP or on a death of a certain mob (or mobs), using bloodlust on the higher DPS phase is the same as using it on the lower DPS phase, since in both cases you'll be shortening the fight by the same duration.

 

◊ Kyth

Mages: 10%
Warlocks: ~12% effective
Arms Warriors: 0%

You can play the "can't your tanks do better" card if you like but tank threat doesn't scale with haste like DPS threat does, and tanks don't have "TPS cooldowns" they pop like DPS do. Especially if you're talking real fights (which you must be given the emphasis on "but all 25 might not be alive later"), threat is very relevant.

"10 seconds later" isn't even close to a real buffer, especially given how percentage threat drops: soulshatter and salv are percentage-based, and the way you keep a high enough threat ceiling is to postpone using either until as late as possible so you shed as much threat as possible. Burning soulshatter 50 seconds into a 6 minute fight will just result in problems later on.

Yes tanks can improve threat output: by putting on threat gear and specing for threat talents. This tends to fly directly in the face of the progression fights you're talking about though, where they're doing whatever they can for mitigation since these are hard encounters.


And "lern2play don't pull aggro" isn't an answer either: I'm not saying "oh noes I will have no choice but to die", I'm saying that non-arcane mages, non-demo locks, and arms warriors have significant threat issues already and rely on long-cooldown and/or percentage-based threat drops to manage. Having your big dps boost come at the end works far better with these tools.

Unfortunately I can't quantify it -- but if you have people holding back it diminishes the argument that it's the same or better DPS used earlier.

Officer of <Fusion>
http://www.stratfu.com/ - StratFu: Strategies, Guides, Movies, and Tips for Wrath Raiding

 

Dampf

Actually Arms Warriors do have a build-in threat reduction of 20% by being in Battle Stance. Berserker Stance also has this bonus.

 

Xiar

On top of what Bink has already stated...

Using it early has the added advantage of everyone still being alive.

Using it early also has the advantage of people having mana to go all out during it.

Your tank scales with haste too... so does his threat generation. Maybe not quite as well as dps does, but you really shouldn't be pulling off your tank even if you heroism at the start... if you are it's the tank's threat.

The only real argument for late heroism is classes that benefit more from it then wanting more dps on the meters (and likely at the expense of the raid).

 

Rhaegal

The first point is not true for a lot of practical reasons. At the start of a fight, tank threat is not yet fully established. I'll use warriors as an example, because I'm more familiar with the mechanics. Say you're pulling a boss and know that there's going to be a hard burn at the beginning. You reach the boss, pop Shield Block, and hit Shield Slam, a cooldown ability and one of the cornerstones of warrior threat generation. Your SS gets parried. You dodge the first couple incoming attacks. You have little to no rage, your initial high threat snap aggro ability didn't land, and DPS are prepping to pop cooldowns and burn. Many DPS classes have cooldowns that, when stacked with Bloodlust, can increase their TPS by double, sometimes more, while as a warrior tank, you don't really have personal threat cooldowns, and Bloodlust is only going on increase your Heroic Strike threat (assuming you have plenty of spare rage) by 30%, leaving the rest of your threat static. Now, obviously this isn't going to happen every time, and mana/rune-based tanks are perhaps less susceptible to spiky initial threat than rage-based tanks, but I'm just trying to illustrate the point that tank threat isn't always established at the start, and poor threat scaling with Bloodlust will only exacerbate the problem.

Obviously that said, early Bloodlust is certainly ideal for the points you and many others have listed. You just have to make sure your tanks are watching their threat abilities and know when to say "Wait, hold Bloodlust for 10 seconds while I catch back up."

The second point (which I know you're not advocating, just trying to address it for people who would) has been knocked down multiple times in this thread and Bink's blog entry on the subject. DPS classes with an execute ability will not find themselves higher on the meters with late Bloodlusts. In fact, the more unsynced cooldowns they have, the more likely they are to end up higher on the meter for the fight with an early Bloodlust. If their goal is to have nice pretty bars that show them doing lots of damage, they should want an early Bloodlust, too (or not care when it happens at all, depending on their own access to DPS cooldowns).

Stand back! I'm going to try SCIENCE!

 

Xiar

Well I think it's obvious you wait for the tank to establish threat yah.

I'm not saying pop heroism in the first few seconds...

I don't pop it until thru a full dps cycle personally.

 

galzohar

Just because you can't use it on t=0 due to threat doesn't mean you shouldn't use it as early as possible when threat is no longer an issue. Anyway, when you make arguments about threat you should really mention if you're simply threat capped (that is, average DPS X threat modifier < tank TPS), or just a burst issue with tank TPS being high enough overall, in which case you build enough of a buffer and *then* use everything. "Wasting" salv/soulshatter/etc early so BL can be used early should not be an issue if average tank TPS is higher than average DPS X threat modifier, since after the massive cooldown usage is over you should never be able to catch up to the tank. The only case where a late BL will be worth it due to threat issues, is if using it late allows a bunch of people to get an extra cooldown by using their personal cooldowns earlier. This is very unlikely to be the case, considering how fast tanks open a gap against DPSers. In a totally threat-capped (tank TPS < average DPS X threat modifier) situation you're playing a completely different game that I personally hadn't seen in a long long time on any fights (hmm, broodlord?) where you're not even trying to maximize DPS but rather damage per threat while also staying under threat cap.

 

Salus

I think this entire argument has been framed incorrectly. Really, the argument is about trinkets. I'll try to frame it from the mage perspective, as that's what i know best

In TBC, trinkets were generally activatable with 2 minute cooldowns. This meant, on Brutallus, with a fixed 6 minute kill time that everyone would get exactly 3 uses out of their trinkets. Because heroism use was generally neutral (and not raid wide) and maximizing DPS was so important to the encounter, you could have your elemental shaman, for example, save heroism until 20% (for molten fury's +20% damage) because the 2-3 fire mages in his/her group gained a compound benefit by stacking heroism (haste) with their other fixed length +spellpower effects. And since there was no harm to anyone else (everyone could save their third trinket use for heroism at 20%), the advantage of getting this extra DPS didn't cost the rest of the raid any damage. Threat was also a very big problem until the first taunt.

Now, in WotLK, most trinkets proc automatically and there are virtually no end-game trinkets that have an "on-use" effect. Most DPS trinkets have a constant bonus, or an "auto-proc" effect that occurs about once a minute. Blizzard seems to have done this intentionally to remove the compound bonus that could be gained under the old system. It's the change in trinkets, that hints to bloodlust/heroism early making the most sense.

I'm also interested if someone is willing to take up the task to determine at what point an early heroism is best used. Someone tossed out 10 seconds as a number, and that seems reasonable to me. That should allow IDS/FotFF to be stacked to full (or close) and most trinkets should still not have proc'd, but perhaps this is wrong? Assuming most DPS start their rotation 0-2 seconds after the tank engages, when will the first trinket proc (on average)? when will the last trinket proc (on average)? It seems to me, that 10 seconds in is probably much better than 30 seconds, but I'd like to see the proof. Also, raid probably doesn't matter, since almost every DPS class is utilizing similar trinkets.

 

Darian_TruBlade

Your Soulshatter example interested me, so I devised two formulas. The first formula takes the tank's non-Bloodlust/Heroism TPS (a) and the time before Bloodlust/Heroism (t) to solve for the maximum possible personal (i.e. no demon) DPS the Warlock can have without Bloodlust/Heroism (w). The second formula does the opposite, finding the minimum tank TPS required to allow the Warlock to go all out DPS (again personal) for the duration of Bloodlust/Heroism. These formulas assume that a tank only gains 10% TPS from Bloodlust/Heroism while a Warlock gains 30% DPS, and that the Warlock will Soulshatter at 125% threat.

w = a(t + 1.1*40)*1.25 / .9*(t + 1.3*40)
a = w*.9*(t + 1.3*40) / 1.25*(t + 1.1*40)

For t = 10 these formulas are simplified to the following:

w = 1.21a
a = .827w

The coefficient in the first equation is effectively an inflection point. If a Warlock's personal DPS is greater than 1.21 times their tank's threat they will have to use Soulshatter during an early Bloodlust/Heroism. Otherwise Soulshatter will be completely unnecessary. 90% threat means that 1.21a DPS is only 1.09a TPS, too little to ever pass the 130% ranged threat threshhold. You begin to lose DPS if you pass 1.227a DPS, as at that point you'll use your Soulshatter during Bloodlust and catch back up to the tank before Bloodlust ends again. However, even in this state you won't have problems down the line, just during Bloodlust.

You will only encounter threat problems later if your normal threat generation is greater than that of your tank. If that is the case there's a much bigger design issue at play here.

"A man's IQ, yearly income, sexual prowess, ingenuity, physical appearance and generally every other aspect of his character can be condensed down to four digits: his Arena rating." - Zechsy [70 Rogue - Skullcrusher (EU) - 10/23/2007]

 

◊ Wraithlin

The whole point of this post is that "conventional wisdom" about stacking molten fury and BLust was, in fact, a pile of bunk that had no basis. Apart from that it doesn't really matter when you drop BL, provided you line it up with other external cool downs. With you example of a 2m cool down on a 6m30s fight you could BL with trinkets at 20s, 2m20s, or 4m20s.

The optimum strategy on Brutallus was actually to just chain BL on your highest DPs group by rotating shaman through that group.

 

MatsT

If we disregard the part about chaining Bloodlust in just one group, the correct dps strategy on Brutallus was that the Mage group got their Bloodlust under 20% and non-mage groups over 20% to reach the 20% phase faster. It is only the raidwide Bloodlust that made it "not matter".

 

songster

Mispost, sorry.

 

galzohar

Actually, for minimum fight duration, stacking MF with BL *USED* to have a basis (in TBC), however with BL being raid wide it is no longer the case, unless the unlikely conditions mentioned in this thread are true.

 

Ravelvan

I'm pretty sure this would actually be untrue.

At least, if it were true that a single party Bloodlust was beneficial to use on classes during execute time, it follows that a single target Bloodlust would be used more beneficially on an execute class during execute time. A single target Bloodlust is essentially the same thing as a trinket (or any other cooldown that is used by an individual on himself, such as Icy Veins). This would mean that, for example, a mage must use his cooldowns during execute time to get maximum benefit from his cooldowns (which is exactly what this thread says is incorrect). If this is true, then you would also need to use Bloodlust during execute time to match up with the mage's trinket usage, which refutes the whole point in this thread (unless we are assuming that with an early Bloodlust you will use all your cooldowns on the Bloodlust and then again on the execute, but I think this assumption is unnecessary).

Edit: Woops, found the error in the math leading to the strange result, so disregard below.

I just worked what all the math would look like with one "execute" class that also receives a Bloodlust who is DPSing alongside someone else who receives neither benefit from the Bloodlust nor execute range.

The weird thing is, I expected the answer to come out the same as when you are only looking at one DPSer (that is, time to kill remains unchanged no matter when you Bloodlust), but instead the result I got suggests that using a single target Bloodlust (ie. any sort of trinket or class cooldown) early is actually better than using it late (during execute). I can't really figure out how to justify this to myself. Then again, it could just be due to an error in my math.

 

galzohar

You misread the thread. Personal buffs are best used when you do the most damage unless everyone in your raid are the same (or similar) class as you, in which case it won't matter. If executing classes in your raid scale better with haste than non-executing classes that'll also favor a late bloodlust/heroism (though the effects of this are probably not significant enough in a real raid to compensate for the more difficult cooldown coordination caused by the late bloodlust/heroism).

A simple simulation wil show that if you're just a single DPSer that can execute you can use your cooldown at any time and get the same benefit (fight shortened by X seconds based on the relative DPS increase its duration). If you add a non-executer than he's effectively "helping" you bring the boss down to execute range, making the late cooldown useage more effective (you're basically "less capable" of shortening your own execute phase making it favorable to use the cooldowns at the execute phase). If the cooldown is the same relative DPS increase for both of you, then again it doesn't matter when you use it since he'll also shorten your execute phase when you use it.

 

Watipah

Right now there are few fights that are longer then 3 mins.. so many classes won't even be able to use their cds twice... BUT they will try to... if they don't surely now they won't be able to use them twice...
Thread isn't an issue atm.... most of the Times i'm placed 1-3 of the dds in thread as enhance.. and even mages with bloodlust und 9k start dps with bloodlust won't overaggro....
Most times I'll autohit until tank thread is generated and UR is up... then call ghostwolves and use blootlust this is probably about 6-12 seconds after fight begins...
so every raiddps class will probably profit of trinkets AND their own cds stacked with heroism... regarding this thread i won't change it for a while...
But I think it would be intersting to test it on the ptr patchworkdummies if anyone can convince their guild to test it.. ?

 

◊ Wraithlin

This is not 100% correct, see the simple example on page 1 where you can create a scenario with 1 execute, 1 non execute and early BL is the correct strategy. The ideal is to use bloodlust when it generates the largest percentage increase in raid dps. With a raid wide BL that is simply a function of how your raids DPS scales with haste. When you BL a sub-set of your raid then there are 2 factors, their personal dps AND how they scale with haste. If you cannot BLust the whole raid then the first BLust should goto those who generate the largest %age change in raid dps (which in this case is the same as those gaining the largest absolute gain) at the point they gain it; normally meaning execute classes below 35%. Your second and third BLusts should be assigned similarly, but if you run out of execute classes to BLust you should just chain BLust on the highest dps class.

Especially with early Brutallus fights where 35% could last over a minute, BLusting your best group twice was a superior strategy to BLusting several groups at 20%.

See Here : [Math] Bloodlust/Heroism timing

 

MatsT

The numbers in that one are way off since you are having Joe scale with the early bloodlust but not the late bloodlust.

 

◊ Wraithlin

Oh dear, that sat about for far too long. I have fixed it and though the difference is smaller, early BL still wins out.

 

Jayde

Isn't the issue with this that the classes with execute range multipliers then have to choose if they line up their other cooldowns with Heroism -or- execute range and not both simultaniously?

One would presume that a specific static short-term bonus such as +300 spell power for 20s would be most benefical to use under the effects of -both- Heroism haste and Molten Fury, rather than one or the other.

 

galzohar

Wraithlin, I don't see how your post (#70) has anything to contradict what I was saying. I mean, I agree with what you're saying except the "This is not 100% correct" part. In the bold part I was talking about a cooldown only affecting the executer.

 

◊ Wraithlin

:HeadDesk:
:HeadDesk:

Ok, lets go over this one more time.
One might presume that you should line up cooldowns with execute in this way, but you would be wrong.
THAT IS THE ENTRIE POINT OF THIS THREAD.

How many times must I repeat this?
There is (almost) no advantage to lining up a raid-wide bloodlust with execute. Infact, if you read Binks article there is no advantage to lining up trinkets with execute either, and all for the same reason (DPS increases are modeled as a general "dps increase" without being specific as to its source). Lining up DPS buffs with execute shortens the execute phase, so while you might get OMGDPS while everything is up, you lose a significant chunk of pain old execute-dps time.

Execute is not a trinket, stop thinking of it as one.

The part in bold says that adding someone without an execute talent makes it such that it is always better to BL during execute. That is not true, it is entirely dependant upon how the raid dps scales with BL in the two reigimes (execute, and non-execute range.). This is demonstrated by the example I linked from the first page of this thread.