corsair

Yeah, sorry, I think this line of thought is flawed. Just take a 25 second time frame for example.

350 mp5s from felhunter sac is 1750 mana generated. And you have 10 shadow bolt casts, which cost 4200 mana, so a net cost of 2550 mana and 15% of your dmg.

0 mp5s from pet sac is 0 mana generated. And you have 9 shadow bolt casts, which cost 3780 mana and a life tap , which generates ~2k mana (really 1.5 life taps so 3k mana but it's unneccessary to be that precise in this example). So a net loss of 1780 mana. This cost you 10% of your other damage (only 9 of 10 SB casts), but you gained 15% from the succy sac. So it's a net gain in dps, and you lose less mana.

You can generalize this to be more analysis friendly, but I learn best with examples, so hopefully this one is clear enough that you can understand why saccing a succubus is always better if your healers can support you. If you trace down the spreadsheet formulas, it also shows the math for why succy is better.

Personally, I like the felhunter on Kaz, but he's the only one, and I'm sure that I'm wrong to "feel" this way. I just don't like chain life tapping when there are still people that explode.

 

galzohar

Spread out more not to mention you can tap way before the explosion when the raid is topped off so when you do get the mark you have more than enough mana and don't need to tap when people are actually exploding.

Regarding fel concentration, I was replying to specs that were linked that included it over the defining talent of shadow embrace as well as other talents that actually increase DPS like nightfall. Of course if it comes as your "first 15 affliction" talent there's no problem... For example if your gear happens to have +14% hit on it already something like Talent Calculator - World of Warcraft would be great. However it is not worth intentionally gearing for hit so you can spec out of suppression. As affliction your dps shouldn't be aggro pulling enough to warrant more hit on soulshatter at the cost of DPS, and 1 hit rating is slightly inferior to 1 spell damage for affliction, not to mention 5 hit is even more inferior to 6 spell damage. I still dont' see how you can have 5/5 and still get amp curse which is probably going to be more useful more often for CoD on an off-dot target, not to mention you're losing soul siphon - while it doesn't stop interrupts it greatly increases the amount of drains and is actually useful on the fights where you would maybe need it and actually not take any pushback (say, a bad kaz'rogal with oom healers and everyone blowing up buying that bit of extra time to finish him off). With 14% hit on gear I'd probably still take Talent Calculator - World of Warcraft.


haste diminishing returns
While haste is quite close to a linear stat like all other stats and therefore its relative value diminishes the more you have of it, just like any other stat does, the fact it doesn't increase your efficiency does have a cost. Look at the relative value of haste compared to other stats on lieuler's with 350 mp5 and 0 mp5 and see that it goes down by a very small amount when you have 350 mp5. Also if you look at the actual (#) DPS increase from adding a point of haste rating when you have a little compared to when you have a lot you will notice the actual dps gain is a tiny bit smaller with higher base haste. This will not be true for spell damage or crit. This effect is extremely small though when you actually look at the numbers - so while haste does have a little extra diminishing returns over what every stat has, it has an extremely small effect.

 

Krazen

Sorry. Key words were 'will use', ie in my 2.4 spec.

I know the jury is still out on this, but I'm personally convinced that fire will be a superior dps spec in 2.4.

During movement, there are 3-4 options:

1. Drums
2. Lifetap
3. Conflag
4. Do nothing


Sometime 1 and 2 are not possible or desirable. Conflag should be a net dps gain in this (situational) situation. Keep in mind almost every boss fight requires some kind of movement. Or it resets aggro and you want to get a last bit of damage in.

Either way, 2.4 fire Destro has a free point; you might as well take it.

 

• frmorrison

Well, you will be able to find out tomorrow if this is the case for you.


From my personal testing of the spreadsheet, Fire was higher personal dps, however imp SB was up less so there would be less raid dps.

DK - Ashbane Failure is the condiment that gives success its flavor.

 

galzohar

You also need to move close to the end of immo or else you will easily get rid of more immo dps than you gained by conflagging, or at least close enough for it to not be worth the mana. Conflag is so rarely useful that once immolate is break-even on DPS (taking ISB into account) there's absolutely no point and there's very very little point to get it regardless. Heck 1/5 cataclysm will provide more DPS on some cases, maybe even enough to outweight conflag's talent point even for the immolate user. Not to mention if you're at the/a point where immolate isn't worth casting it makes conflag even more of a waste. Not to mention speccing into fire talents means you have exactly 1 floating point in destruction (while if you don't take fire talents you can take pretty much whatever you see fit in destruction while keeping your max dps - be it tanking talents, nether protection, 2/3 soul leech etc).

 

Seir

Hi, my guilds been on Mother for about 2 weeks now and we've been trying out switching from SB to Incinerate during the ice and shadow aura's; however, because we've been on her only a short time I don't feel i have enough data to determine if its more effective then just spamming shadowbolts at her, so I was wondering if any other locks with more experience on mother has determined this yet?

 

Morwen

Balanced by the fact that ISB will be 'worth' less in a raid with fewer shadow users. It is pretty close for any sort of reasonable number I plug in. For a quick raid-tailored estimate, just take:

S = total current raidwide shadow dps
U = current ISB uptime
P = your personal dps
U' = ISB uptime after you switch to fire
P' = your personal dps after switching to fire

In order for fire to be worthwhile, it must be the case that

P' > S - (S-P)(1+0.2U')/(1+0.2U)

Or to put it in terms of 'how much does fire need to beat shadow by?'

P' - P > (S-P)(1 - (1+0.2U')/(1+0.2U))

A couple toy examples would be:
If there's currently 5000 shadow dps excluding you, U = 60%, and U' = 50%, then fire needs to beat shadow by 89 dps to be worth switching.
If there's currently 8000 shadow dps excluding you, U = 70%, and U' = 66%, then fire needs to beat shadow by 56 dps to be worth switching.

 

Trickykid

My non-crit incinerates were hitting for less than my non-crit SBs during "less shadow" and "more fire" auras so I keep to SBs now when ISB was not up. That's a lower bound too since incinerates would need to hit for ~10% more (depending on Shadow-users) than SB without ISB to be worth switching to.

 

Krazen

Immolate is probably worth casting during shadow/frost auras, but incinerate probably isnt. Don't be dumb about this, but try to squeeze your lifetaps in during shadow auras.

 

Seir

Yeah, I didn't have a doubt about Immolate, and I personally tend to tap between aura changes (so if a holy aura pops i can devote the entire aura uptime to it) as well as during the Shadow Phase.

It just seems in theory it would get you greater damage but I'm tending to lose damage to the switching between the two do to mostly human error and do to the randomness of the auras its hard to tell.

I figure if no one has already figured it out for sure then I'll just keep track of my damage on shahraz, and over time draw a conclusion off of the relative frequency data.

 

weijonny

since the 2.4 spell haste big buff, is it a good idea to go full spell haste SB?
then it would be point less to cast immolate, just like casting any other dots?
just wondering, if the haste can be buffed to 2sec or less per SB, then dps would be around the SB dmg you are doing.
gears depended.

 

Cronjob

I guess I'm still now following or maybe I've been missing out. But in my normal rotation I never cast Immolate except on the initial blast then I go into a pure bolt spam. Occassionaly I will toss another one up after an aggro reset but for the most part I just bolt spam.

Now if what you are saying is that in 2.4 switching to fire you will be using Immolate + conflag + Incinerate then thats an interesting spell rotation although I'm not sure I'm completely sold on fire spec yet.. Soulfrost ftl

 

Cohren

There was no ISB change so I don't know why Fire vs. Shadow is still being discussed. Unless your guild brings like 5+ Warlocks, they(You) all should be shadow spec'd to help the raid. Raiding has never been about the individual but rather the synergy between the different classes. There are some steep DPS fights in Sunwell and your raid is going to have to squeeze out every last drop of damage so unless the gain in DPS of a Fire lock out gain's the loss of DPS from lost ISB uptime I see no reason to spec fire for progression.

 

Eph

Wouldn't it be difficult to keep Immolate up in an Incinerate rotation while gaining haste gear? Or would it be best to Conflag or Life Tap perhaps at the very end of the Immolate? It seems like haste's value would be more difficult to determine with this uneven rotation. Or is everyone assuming another raid member's Immolate will be up and there would be no downtime?

 

Liminality

that is completely not true...


The italicised part is incredibly false...


All things kept constant, the move from 0-1% is not as big of a dps increase as 50-51%. In fact, it increases exponentially. When you move from 98% to 99%, the DPS increase is astronomical compared to even the 50-51%. I am not a math major or anything, so I wouldn't know exactly how to prove it. The increase in the number of attacks in a given time frame increases exponentially as downtime is reduced. Does that make sense?

I did play FFXI for a long time where there is a combo for Dark Knight that involves stacking haste that follows the same principle, so I am sure its right.

This is also assuming no global cooldowns and whatnot to fuck everything up


At least that's how I believe haste works, correct me if I am wrong...



As for Crit. I think the increase in DPS is linear. From 1% to 2%, the increase is static. From 98% to 99% the incrase is the same. However, more crit is more ISB uptime, which is a hidden factor that will incrase DPS (raid and individual). So I guess, ISB uptime would actually skew crit from a linear incrase in DPS to a line that curves upwards like haste, but not nearly as efficient at higher levels.

overly wordy, i know, i apologize

 

Krazen

Not really. It takes a bit more brainpower than rolling the mousewheel while watching TV, but it shouldn't be considered difficult.

I posted some ideas on that a few pages back:

PVE Raiding Compendium


But the cycle is ultimately quite simple:

1. If Immolate isn't up, cast it.
2. If Immolate has time > (Your incin cast time), cast incinerate.
3. If Immolate has time < (Your immo cast time), cast immolate.
4. If Immolate has time between 2 and 3 (usually between 1.4 and 2.1 seconds or so for the average t6 player), lifetap, pushing you into 3. If you don't want/need to lifetap, just cast an unbuffed incinerate.
5. If you move, Lifetap (this is the same as Shadow, really). If you don't want/need to Lifetap, Conflag, and start at 1.

 

Trickykid

The DPS increase from each stat generally remains constant across levels of investment. Adding 1% haste if you have zero will add 1% of your unhasted DPS. Adding 1% if you have 10% will add 1% of your unhasted DPS. The point the previous posters were trying to make was that relative to other stats every individual stat will decrease the more you invest. In my above example, if you have 10% haste, 1% more will give you 1.11/1.1 => 0.91% increased total DPS, but an extra 1% crit gives a full 1%.

The same happens for each of the stats. As one increases, the relative value added by the other stats increases. None of them have exponential "take offs" like you say.

 

NanoHaxial

Incinerate gains a bonus damage regardless of who put up the Immolate, doesn't it? That means as long as a you had a couple Locks then it should always be up, and the only issue you'd face is that you'd have less than 100% Immo uptime yourself.

 

Krazen

I believe it does, but that shouldn't change a whole lot, since you want to be casting Immolate for its superior DPCT. If you get a SPriest, you might not have to Lifetap, but most of us aren't that lucky and have to LT at some point anyway.

 

Villeraz

Given enough other warlocks casting immolate, it should always be up, and will probably not need to be refreshed before it expires. There may be a possibility of small gaps where it isn't up, but this probability decreases rapidly to negligible past the 2nd warlock.

Someone could try to model "immolate uptime" .

A simple model would be to work with maximum overall % downtimes, which would correspond to the cast of an incinerate (factoring in haste) minus the cast time of an immolate (factoring in haste), divided by the duration of immolate (which may change with Tier 4 bonus) and the downtime calculated.. This would represent the worst-case scenario of casting an incinerate at the last possible moment before you would have the opportunity to cast a fresh immolate.

In a simple case with no haste (other than emberstorm), this is (2.25-1.5)/(15+2.25-1.5), which is ~4.76%. With two warlocks, the chance of both warlocks being in "downtime" is then ~0.23%.

You could increase the complexity of the model by factoring in the fact that incinerate is not being chain-cast, so it will not overlap entirely with the overall % downtime, although I think my model shows that with even two warlocks, the uptime on immolate is fairly good. However, it does not factor in affliction warlocks (who place immolate lower in priority than their other dots), the few occasions where a curse takes priority over immolate, and obviously, shadow destruction locks who do not cast immolate. Also, it assumes that the warlock will be able to watch timers precisely enough to know when their immolate can be renewed without overlap.

 

Nicarras

I've tested this on our last 3 kills.

Even with Immo up, Incin still hits for less than my SBs do in my shadow dest spec even when shes in Shadow phase.

Silly, yes...but I just spam SBs the entire fight. I might LT more during the SR phase though.

It's a Melee fight for the most part...you are just there to provide support DPS and not kill them.

 

weijonny

after 2.4 the haste buff would denfinatley uneven, aff lock might have a hard time to change their dots route.
if i can get all haste gears from the game, including ZA, 2.4, and sunwell etc.. then all i would do is spam SB, immolate+instant corruption from t5 bonus, then if free time, incernate.
phew phew phew phew! SB dps 2000+

CANT WAIT!!! 12:30 grrrrr

 

Nicarras

At my gear level, adding Immolate to my rotation (meaning between SBs when its down) adds a whopping 2dps in the Warlock spreadsheet. And after I get my gear from Sunwell that I want (basically finish out my BC gear) I actually lose about 60 dps casting Immo.

I don't even know why we keep talking about if Immo is worth it. It is really simple.

If Aff
Then Cast Immolate
Else if Dest
Then Dont Cast Immolate

 

Trickykid

I think the above posters are referring to fire builds. With improved immolate, emberstorm and sacrificed Imp, immolate has a higher DPCT than incinerate.

 

Goedel

Discussion about 0/21/40 raid specs in 2.4 built for sacrificing an imp and using fire damage got me to thinking about builds designed to keep the imp around. However, it has been difficult to analyze this possibility quantitatively.

Please note that I have followed the discussion here closely, and I am well aware of the phenomenon of people urging a less effective approach that has been clearly discredited, on the basis of anecdotal evidence or inappropriate comparisons (such as out-performing guildmates with less skill or other poor specs or gearing). There is a little discussion of this possibility months ago in another thread but no solid quantitative answers that I have yet found.

Oddly, the initial post of this thread mentions several unusual talent specs that I doubt see much use (Destruction with Siphon Life?), but there is no reference under either the Demonology or Destruction sections to a build for Shadowbolt or Incinerate spam alongside an imp casting Firebolt, along the lines of either this 0/21/40 or this 0/39/22 (or this variant on the latter). These certainly appear more promising than some of the builds that have been discussed. Unfortunately, finding accurate data for modeling them is difficult.

What's needed:

• the rule for deriving the imp's spell damage
• the rule for deriving the imp's crit rate
• the rule for deriving the imp's hit rate (is it still 83% against raid bosses, unaffected by the master's stats?)
• the spell damage coefficient for Firebolt (assuming that the general formula takes the same form as with a player's spells, adding [(some coefficient) * (bonus damage)] to the base spell damage)
• the order of operations for talents that affect it
• whether the Improved Firebolt talent silently affects spell damage coefficients the way mages' Improved Fireball and Improved Frostbolt used to
• whether Ruin affects the imp

Possible solution: Firebolt damage = (<base Firebolt damage as indicated in tooltip> + <fixed coefficient> * <warlock's bonus damage>) * (1 + 0.1 * <points in Improved Imp>) * (1 + 0.04 * <points in Unholy Power>) * (1 + 0.05 * <points in Soul Link>) * (1 + <imp's crit rate, which is the same as master's crit rate>) * <hit rate>.

Nuking alongside the imp will be higher dps than sacrificing it as long as the imp's dps is higher than 15% of the warlock's. Actually, in the 0/39/22-style builds, it only needs to exceed 10% of the warlock's pre-Soul Link damage (though that damage may be lower than in the 0/21/40-style build). Here's one sample calculation; I hope people don't get tripped up in "but my numbers are such-and-such, so this whole conversation doesn't apply!"; you can perform the same calculation. If a warlock does, say, 1600 dps on a 30% crit rate and 1200 bonus damage, the imp would need to hit 360 damage per 1.5-sec Firebolt (after talents, including both crits and non) to exceed the dps gain from sacrificing it. Without soul link, that's 213 (base damage + C * bonus damage), which is 97 bonus damage, or a 0.08 coefficient.

So, before I attempt rigorous testing with various gear and talent combinations, can somebody point to a definitive source that has already determined the missing variables, which I have not found here or on the ever-spotty wowwiki?

Additional considerations:

• Imps when not phase-shifted can be fragile. Talents can mitigate this. Players can be fragile, too, and many who claim to be serious about min-maxing raid dps dismiss stamina entirely or treat dps vs. survivability tradeoffs very fuzzily. The "standard" stand-and-shoot model is, as we know, hardly standard in current raid content anyway. The best approach is to determine the ideal dps for this case just as for the others and then to work out more rigorously how serious this drawback is for various fights.
• Keeping the imp out allows one to double as a provider of fully improved Blood Pact (as long as the imp doesn't die; see above).
• In the 0/39/22-style builds, Master Demonologist grants 20% reduced threat.
• The imp may run out of mana. In a 0/39/22-style build, it may be possible to replenish it fast enough to keep going indefinitely at the cost of some dps; the gain through maintaining its Firebolts must be weighed against the lost dps from other talents. It is also possible to begin a fight with the imp casting and then sacrifice it when its dps contribution slows, continuing as in a normal 0/21/40 imp-sacrifice fire spec.