Madlax

I agree with one of the OP, not doing CoE in that raid setup definitely doesn't work in your favour or towards a good starting point.
As for the topic itself: My guild prefers giving Bloodlusts depending on how big the Threat margin is towards the tank at times. Having a 2300 DPS quad warlock group running only a handful of TPS behind the tanks and Soulshatter used getting Bloodlusts doesn´t help the process either. Mages(that would be 2 at max) get their Bloodlusts at 20% usually - that´s about the only time they can keep up/push ahead of the other high DPS classes though.

 

Sethrod

another question,

I've seen quite a few top Warlocks wearing Quagmiran's eye again. Could anyone figure out at what spelldamage/haste/crit levels it will actually be superior to the Icon?

 

LockApologist

You haven't proven this. Run the analysis that was done a few pages back for locks stacking skull + lust for mages. Since both locks and mages simply cast one spell, their mechanics are similar. Stacking more cooldowns may enhance the effect, but what you listed above does not prove that.

I went through our WWS and found that EVERYONE dropped when I ticked this. The only people that gained had >100% DPS time. people with 100% dps time stayed equal. Give me a link where Mages increase, but locks drop where the mages don't show up with >100% dps time.

There is a thread in this forum where a lot of people argue this point. I have yet to see any math out of it to show if which classes gain the most. There's a lot of anecdotal stuff and some 'I feels' but no math anywhere.

 

Jakiri

For buffs that work for a certain amount of time rather than a certain amout of casts it's barely worth saying that increasing the number of spells cast while the effect lasts increases the number of spells cast under the effect. If you're stacking haste effects there may be issues to do with lag and the mechanics of casting a large number of spells, like running out of mana, but with haste it is generally better to have 200% of the effect 50% of the time rather than 100% of the effect 100% of the time.

With things like Destruction potions the mechanics of chain casting are identical, and the increased damage done merely depends upon the number of spells cast, as he suggests.

If you want to compare mages and warlocks then this isn't sufficient, however. One of the benefits of being a warlock is that an anticipated heroism can be met with a full mana pool and pots off cooldown; our mages have to drink mana potions to keep their DPS up through Brut.

All things being equal, the stacking of effects would suggest mages get more use from heroism, but all things are not equal. I don't know enough about mage mechanics to create a definitive answer, unfortunately.

 

Lambi

Isn't the main reason of heroisms being better for mages that they can use cooldowns such as icy veins at the same time coupled with perhaps molten fury and a destro pot?

 

Coll

Hy I've been thinking of this taletn since teo weeks and I1m interested what would you recommend for me.. Talent Calculator - World of Warcraft I use mainly incienerate while immolate and corruption are up and no shadowbolt at all... Usually i keep my imp summoned because the raid wants it so i tought this could be ideal for me.. bytheway I want to be the best dps so any other ideas and game playing tactics? thanks bye

 

Thanahtos

Using proper grammar might help. And that talent build is terrible. If you need to keep your imp out, spec affliction.

 

Pintofbrew

LockApologist: Perhaps you'd be better guided to seeking the proof you're looking for in mage threads rather than the excellent warlock pve thread. It would take a fool in the face of clear mathematics to cry fallacy when the fact remains that a warlock pressing one button using no cooldowns makes much, much more dps than a mage doing the same. Thus, it means a competitive mage gains more DPS from his cooldowns than a lock does. It is rather obvious that adding another CD to this amplifies the effect rather than flatly increases it.

At the risk of sounding obvious, perhaps you should try and disseminate some mage wws data provided by a competent mage on a correctly-done fight and notice the massive spike in the "all CDs going off" time-frame, or perhaps run your own version of Rawr and add/subtract the use of consumables/BL and see for yourself the changes, rather than going "yea, you have 7 different effects going off at once and we have 2 so what's your point?" and replying to "there has been proof BL is better on a mage" with "I haven't seen it".

 

Krazen

At the Sunwell level (full BT gear, maybe a couple sunwell pieces), I have the Icon losing to Quag's eye by 6 dps. It's a moot point half the time, though, as the Darkmoon Card is superior to Qaug's, and the Skull/Hexhead obviously are too.

It's a real shame the Eye of Mag couldn't be an epic version of Quag's.

 

Arelenda

While I do agree to your points, I think you're exaggerating. As far as I can tell mages have Icy Veins and Molten Fury. Everything else both locks and mages can pop, so 7 vs 2 is not very accurate.

I still stand by the earlier conclusion, though. I doubt I can outdamage a well played mage getting 20% haste for 20s and doing 20% extra damage during a 40s Bloodlust/Heroism.

 

Cohren

Before I run the numbers myself, I just wanted to check and see if anyone had bothered to run the numbers to see if the "most optimal" gear set only involved going for the best in-slot items and ignored the hit cap. I know at first glance there seems to be no way this can be true but there are some major sacrifices made to get 202 hit so I'm wondering how close it will be and if its really close if it would be worthwhile going that way instead of taking sub-optimal items and gems.

 

Flamingcloud

I think it is safe to assume most of us are under the impression that warlocks are clearly favored for heroisms 100-20% regardless of icy veins, and sometimes are favored in <20% range. I definitely think a heroism at 80% with the mage popping all of his stuff(including icy) would not be as good as a lock heroism at that point.

As to quag's eye apparently there is a point in which you gain an extra shadowbolt out of it making it jump to 37haste or something like that.. There is a graph a few pages back.. It is usually better than icon on a fight like brutallus (while icon would be better on felmyst due to like 80 second dps chunks reducing the cooldown)


Just use all the best items and socket the rest of the hit (should be like 20hit short alliance side with dranei)

 

Jakiri

Going by the numbers I have on Leulier, going best in slot (ignoring hit) has identical results from going best in slot in all but a couple of slots.

You end up at the hit cap, there or thereabouts anyway, filling in the gaps with veiled pyrestones and with a Draenei in group.

 

Morwen

It is actually not necessary to run the numbers: when you are not at the hit cap (which may be 202 or less depending on group comp) then it is optimal to socket Great Lionseye in yellow slots and Veiled Pyrestone in red slots, and there is no conceivable best-of-slot gearset for which such a socketing scheme will not get you to the hit cap considering the number of sockets available.

Or in other words 'best-of-slot' takes the hit cap into account because hit is so far and away the best stat. If there is a gearset that is not hit capped and purports to be optimal, then the same gearset socketed for hit will be better.

 

LockApologist

I am not asking for anyone to do the math, but an assertation that 'It's obvious that stacking cooldowns are better than using them seperately' is not completely valid when there was just the comparison of skull + lust stacked vs seperate showing practically no difference. Especially considering you probably had the trinket/pot off C/D for some time waiting on lust.

Since this forum prides itself on theorycraft and facts, saying it's 'It is rather obvious that adding another CD to this amplifies the effect rather than flatly increases it.' is rather ironic. I'm not saying it doesn't. I haven't done the math either, and, since I don't have a 70 mage, I probably won't ever do it. That doesn't prevent me from pointing out that it's a hypothesis.

This was my point.

I have no allusions that during MF range I can out dps a lusted fire mage. That would be silly. But saying that Skull + lust + IV is greater increase stacked over seperate when that was shown to have no noticeable increase for just skull + lust doesn't make it so. Stacking haste on top of a +dmg trinket actually makes more sense than stacking skull.

 

Roywyn

For warlocks, that isn't even an issue. With the GCD scaling, you can safely life tap through bloodlust/skull and it shouldn't affect your overall DPS at all.


I'd actually love see the numbers between "life tap during +damage trinket" and "shadow bolt during +damage trinket", I have no idea which one is better by how much.

Edit: Let's take a hypothetical "+200 damage for 15 seconds" trinket proc, no base haste.
Spamming life tap would affect 10 life taps (15s/1.5s), adding 10*200*0.8 = 1600 mana.
Spamming shadow bolt would affect 6 spells (15s/2.5s), adding 6*200*1.06 = 1272 base damage.
Multipliers add 1272*1.05*1.13*1.1*1.15*0.99*1.15*(1+0.4*1.09) = 2742 damage (Misery/imp.CoS/SPriestShadowVuln/SucSacc/1%resist/75%imp.ISB/40%crit+CSD).

Over 2 minutes, the choice is 1600 mana/120s = 67 mp5 and 2742 damage/120s = 23 DPS.
According to Leulier in HS/BT gear, 67 mp5 adds 21 DPS.

Seems like life tapping during damage trinkes is only 10% worse than pure shadow bolts during damage trinkets.
Roughly, since some of the steps are just napkin math.

Edit II: I aimed a bit high with ISB uptime (2-3% damage) and crit chance (3-5%), but forgot about 4T6 (6%). So, correct the trinket value down by about 0-2%, it's all an estimate after all.

Also - yes, I am aware that no one life taps 10 times in a row. I just wanted to take a look at the two extreme cases of how to use procs. Makes it easier than comparing 1 life tap to 0.6 shadow bolts and then trying to find out whether that 1 DPS increase in Leulier's is 0.5 DPS or 1.9 DPS.
It just makes the comparison easier to do.

The result is that Shadow Bolt gets up to 10% more out of damage procs than Life Tap. Emphasis that it's very close.

// End of Edit.


For a random heroism, warlocks should be better. For a 2min+3min cooldown timed heroism, they should be pretty close, same with an execute timed heroism without CDs (rare occasion, maybe in a long execute phase). For an execute phase timed with cooldowns, mages should be better.


About the mana pot issue - a mage should get through this fight with 1 pot + 1 gem at most (and a few evo ticks) at the 2/3 minute marks, leaving you free to DPot/FCap on the pull and at the 5 minute mark.


About the joke someone made ("Minmaxers give warlocks first dibs on gear upgrades") - I wouldn't mind that since Minmaxers would in the same thought give mages first dibs on clicky trinkets

The difference between stacked and staggered skull + lust is getting 30% more from your skull. Or 6 extra seconds of 175 haste, or an extra 0.7 seconds cast time.
What is "practically no difference"? 0.7s cast time may sound negligible. Buffing the skull by 30% doesn't sound negligible any more.

The Blue Bar and you - the complete Fire Mage 2.4 mana compendium: http://elitistjerks.com/658230-post3191.html

DPS spec and class comparison in Naxxramas gear: http://code.google.com/p/simulationc...i/SampleOutput

And [Timbal's Focusing Crystal] doesn't proc on AM.
Neither does [The Egg of Mortal Essence] since 3.1.

 

Jakiri

Rough and ready numbers for Life Tap vs Incinerate here.

Using the HSH for life tapping will produce an extra ~170 mana per cast, which for me is another 9%. Assuming an infinitely long fight and infinite health, this will save me ~0.13s per life tap cast. Assuming you can go from empty to full mana in 6 lifetaps, this means you have spent 8.58 seconds of the cooldown to save you 0.78 seconds. At ~2200 DPS this is equivilent to 1700 damage. In this time, you can do an extra 4 casts of Incinerate instead. At (say) 26% crit this corresponds to 211*2.5/3.5*1.26*1.1*1.15*1.05 = 1009 damage.

Over an infinite fight, life tap looks superior. I don't know if this translates to real fights, though - although saving ~1/2 life tap cast per cooldown is a convenient number, as is the need for just fewer than 6 lifetaps per 2 minute cooldown. Something worth testing (and someone tell me if I've screwed up the numbers or made a bad assumption somewhere). Obviously in a fight like Brut you need to time your lifetapping so suddenly losing 10k HP won't kill you, rather than the continuous loss over a 2 minute period. It's also inferior shadowpriest synergy as most of the heals you'll be getting in that time period will be wasted.

[edit]

Yeah, I forgot a lot of modifiers. Whoops! Adding T6 4/8, Sacrifice, Emberstorm, Shadow and Flame and Resisting, we get ~1607. By nearly killing yourself you gain maybe 0.8 dps.

The whole thing is moot because I still can't think of a boss in which it would be worthwhile to do this; you need to stand still for the whole fight, raid damage as a whole must be quite low because you need healers to top you off and any damage done to you must be predictable. So, Patchwerk then?

I just spent a couple of minutes working out the theoretical improvement between, over the course of a minute, running the skull and heroism simultaneously vs having them run consecutively. The result came out as an improvement of between 2 and 7%, depending on how granular you make the test spell you're dealing with. Now I admit this was rather a rough and ready approach and it's definitely possible to model it better than the quick jottings I managed. However, the numbers serve for the example.

2% improved DPS is essentially unmeasureable over the timeframes we're talking about. It is, however, an improvement; and an improvement that scales with the use of other cooldowns and buffs.

Haste is very simple to model, because its improvement is independent of gear. Modelling the full effect of stacked cooldowns vs sequential (or when-available) use is, to be honest, something of an arduous task that I don't personally have the time to do right now, it being nearly midnight. I'll probably give it a go tomorrow.

 

Flamingcloud

Warlocks do not have 40% crit Roywyn.. 35% is probably the high end... Though you did miss the +6% shadowbolt bonus from t6... Jakiri missed alot of modifers. Most fights have moments or threat capped places where it is best to lifetap.. Not to mention most fights it isn't safe to lifetap 6 times in a row.

The numbers for stacked haste cooldowns vs non-stacked were ran far more accurately than "quick jottings" already.


"About the joke someone made ("Minmaxers give warlocks first dibs on gear upgrades") - I wouldn't mind that since Minmaxers would in the same thought give mages first dibs on clicky trinkets "

That is on +dmg trinkets not on +haste trinkets.. So skull is still safely warlock priority in this world... M'uru trnket.. well it affects the entire group so I can use it and affect you.. so thats lock too.. You can have hex shrunken head prio. =P

 

Arelenda

I have 375 crit rating, giving me around 30% crit chance raid buffed. Ret pal +3, Elemental Shaman +3, Moonkin +5, 30-35 is probably a good estimate of the average. I'm fairly high on crit and low on haste myself.

He did leave out 6% T6 and ISB (can estimate at 50% uptime, if you like, for another 10% bonus). But the difference is marginal.


You're actually giving me the same numbers as what I calculated a while ago. Life tapping during trinkets or Heroism/Bloodlust isn't that bad a deal post 2.4. I try to start a cast right before heroism wears off, since that'll get full bonus anyway, and Life Tap can help there.

About the skull vs heroism/icy veins... Can anyone explain how these stack again? Is there an exception that I'm not aware of? There seems to be some misunderstanding about the whole thing.

 

Flamingcloud

Thats higher than me by a tiny bit(around 28-29% raid buffed).. but moonkin and elemental shamans are both rare and I think a poor choice for a baseline. Roywyn actually put isb at 75% which is extremely high.. even with 1 spriest and 3 destro locks it wouldn't be that high.

Also.. It is possible to under heroism have instant spells below 1 second and thus not benefiting fully from the haste, but they were purposing lifetapping during hex not during hero.

 

LockApologist

This is true, but if you don't have to LT at all > hasted LTs.


See PVE Raiding Compendium

Stacking gains .1 SBs over not stacking. Which is not enough to even get an extra started in the difference between stacking and chaining. You're better off stacking lust w/ + dmg trinket, then follow with skull.

 

Roywyn

It's the latter. Heroism and Icy Veins are multiplied.

If they were additive, stacking wouldn't do anything.
But they are multiplied, hence stacking adds a bit.


Edit: An example
Assume no passive haste, Skull (11%) and Bloodlust (20%) over 60 seconds.

- Nothing used => 24 casts (60/2.5)
- Skull used => 24.88 casts (40/2.5+1.11*20/2.5). Skull adds 0.88 casts.
- Bloodlust used => 28.8 casts (1.3*40/2.5+20/2.5). Bloodlust adds 4.8 casts.

- Skull + Bloodlust staggered => 29.68 casts (1.3*40/2.5+1.11*20/2.5). It just both added up.

- Skull + Bloodlust stacked => 29.944 casts (1.3*1.11*20/2.5+1.3*20/2.5+20/2.5). 1.144 more than just Bloodlust, 5.064 more than just Skull.

You can interpret this in two ways:
A) Bloodlust increases the value of Skull from 0.88 to 1.144, which is a 30% increase (surprise!).
B) Skull increases the value of Bloodlust from 4.8 to 5.064, which is a 5.5% increase.


Yes, all these numbers are very small. But a single additional cast for ~6k damage is a 16 DPS increase in a 6 minute fight. The small things really add up.

The Blue Bar and you - the complete Fire Mage 2.4 mana compendium: http://elitistjerks.com/658230-post3191.html

DPS spec and class comparison in Naxxramas gear: http://code.google.com/p/simulationc...i/SampleOutput

And [Timbal's Focusing Crystal] doesn't proc on AM.
Neither does [The Egg of Mortal Essence] since 3.1.

 

Flamingcloud

Torq's calcs here are the updated ones with the right formula http://elitistjerks.com/729900-post2730.html

Doing the math for icy veins myself however..

10% passive haste
2.5 sec spell..

40 sec of heroism
2,5/(1.1*1.3) 22.8cast
20 sec on icy veins
2.5/(1.1*1.2) 10.56cast
total 33.36


20sec of both
2.5/(1.1*1.2*1.3) 13.73 cast
20sec of heroism
2.5(1.1*1.3) 11.44cast
20 sec of nothing
2.5/1.1 8.8casts

33.97casts

So you gain 0.61 casts by stacking at 10% haste... So clearly icy(a +20% increase) stacks better than guldan does (a +rating increase) though its hardly a massive increase.

 

LockApologist

So, as frost you gain .6 bolts, which may be enough to get another bolt started before the haste drops. But, FB is 3.0.

That works out to:

Not stacked
19.047 casts for BL
8.81 casts for IV.

27.857 casts total

Stacked:
11.44 both
9.53 Lust
7.35 of none
28.32

Difference = .46 casts, but likely 1 extra FB. I didn't double check my math, but meh.

 

Pintofbrew

You may add [Flame Cap] to the mix, or the gem-effect of [Serpent-Coil Braid] (it does stack with other trinkets, provided it's "used" second in the chain). Also consider Combustion as a mechanic to grossly exaggerate the caster's crit rate during the "all-CDs going" time-frame.

You may also add Rolling Ignites to the combination, provided a very specific amount of haste (I believe 138 will provide Ignite Rolling under IV). This has been theorized to provide a rough 4-10% damage increase by granting extra ignite damage due to bad mechanics application on Blizz's part, but mandates a fireball cast time of 2.07-2.11. Not exactly a Cooldown per-se, it is however a bug which appears due to cooldowns and thus further compacts the effect.

All other cooldowns are shared with out shadowbolting brothers I believe.