The simplest solution is just to wait before attacking, so the tank has time to build initial threat. I have severe aggro issues, and usually wait 10~20 seconds before doing anything at all at the start. This also allows me to use soulshatter to greater benefit, since the total threat it dumps later in the fight will be much higher.

If you are concerned about losing a raid spot over ostensibly poor dps on meters, keep in mind that personal dps is not calculated until you actually start attacking (pet included). Since I do not have to hold back once I do begin, my overall dps and damage done are actually higher, vs. starting off slowly at the start of the encounter.

I have a question regarding the heavy destruction spec rotation in relation to Pyroclasm. Ideal world all of your conflagrates are critting so it wont make much of a difference but would it not be better (dps wise) to cast CoA at the start of a fight and then CoD once you have Pyroclasm up? If that first conflag does not crit then you are missing out on a few thousand damage on your Doom.

This can be taken through the whole fight, putting up CoA after a Doom unless you have Pyroclasm up. Would it be a dps gain?

One other thing, in a poorly balanced raid where you have to use CoE, is it best to use Corruption in your rotation and if so would it be beneficial to work in 2pc tier 7 to add a bit more dps?

A few thousand damage is about the amount of damage you'd do if you had just put up the doom and spent that extra GCD casting an incinerate, though.

Generally (and this varies on a gear basis, so YMMV), corruption has a lower DPCT than incinerate, so you're usually better off ignoring it. The 2t7 bonus can proc off of immolate too, but I'm not sure what the numbers are relating 2t7 in particular to corruption use.

The average damage of curse of doom for Deep destro from the simulationcraft thread is 16745. That will already include some pyroclasm uptime. 6% of that is 1004 damage over 60 seconds. The amount of mental processing time wasted to work out whether to cast cod or coa isn't worth that much damage, when the average incinerate of that spec is 10048.

I have been looking for information on this very topic. I find that with my current haste rating of about 550 (unbuffed) I have time at the end of every other immolate where conflag and CB are still on CD and I don't have time to cast a full incinerate. This jumbling of the rotation is fairly annoying and sometimes I clip immolate (I'm sure everyone is aware they changed the mechanic of immolate not reapplying). I still don't want to clip immolate, but with my current haste I end up with dead time in my rotation.

I heard from someone that around 400 haste is ideal for deep destruction. I understand haste scales better in theory, but forgive my ignorance, but can simcraft account for immolate uptime, conflag CD's, backdraft, and the number of spells you can actually cast during immolate without wasting uptime?

TLDR: I'm thinking of dropping haste for crit to try and smooth out the rotation and hopefully increase DPS.

I have the exact same problem and always assumed that similar to affliction, dead time is not the solution so in that situation I would cast another incinerate at the risk of delaying the immolate refresh and correspondingly conflag and chaos bolt. Personally I doubt that reducing haste is the right solution. With movement and other random factors in fights that force us to deviate from optimal rotations, the type of situation you describe could occur regardless of the amount of haste you have.

On a related note, I find that destruction with its short cast times, all of similar length, causes these types of awkward situations to occur all the time eg chaos bolt coming off cooldown with just over a second remaining on immolate making it difficult to tell if the cast will complete before immolate drops off etc. The general rules I apply that allow my head not to explode are never to refresh immolate unless I am 100% sure it will drop off before the spell lands and subject to that, always casting according to the priority system even if there is a risk of an incinerate or chaos bolt will land without immolate up. Any insight on this issue would be appreciated.

The bonus damage from Immolate (and all other debuffs) is calculated at the cast completion of your Incinerate or Chaos Bolt. It does not matter if you land an Immolate while it is traveling. If Immolate was not present when you finished casting, you will not get the bonus damage.

I´m looking for a 2nd or 3rd pair of eyes to check the following spreadsheet - just not to have made any obvious mistakes.

http://spreadsheets.google.com/ccc?k...Xhqbv9BUrz5xWA

I was curious which socket bonus would pay off for which spec and just ended up doing this.
Gem choices and maths to it was one of the things i missed from the 3.1 Compendium so far - this should help it.

Greetings, fellow warlocks in the EJ community!

I've been reading the forums here for a little while and found that I was unable to locate certain specific topics that I wish to delve into in-depth discussions with in the existing threads. For the things that I found relevant to my own questions in regards to raiding and PvE, I often had to sift through endless pages of one-lined replies that were often unsupported by the presence of tested trials, numbers, or even a logical argument. I hope to consolidate all of the major pertinent Warlock Gameplay Questions and Answers in here.

In time, I hope to move this into a thread of its own entitled "In-Depth Warlock Gameplay Q & As" to address various specific warlock gameplay mechanics and strategies. The reason for this is - the discussions in this thread is too involved to be placed in the existing "Warlock: Simple Questions/Simple Answers" Thread. I will start to build this revolving around PvE discussions. If enough time passes to allow for a PvP section, one will be added.

To get things going and give examples of things that I am interested in discussing with other warlocks here:

----------------------------------------------------------------------------------------------------------------------

Tags: Destruction | Gameplay | Raiding | Haste Threshold
Posed by: Veange
Q #1: Has anyone done the math / simulationcraft to determine whether there is a haste threshold, under the conditions that:
a) assume all raid buffs are present,
b) assume for a fairly standard destruction build (3/13/55 or 0/13/58),
c) assume for access to all available gear,
d) assume stand-still DPS time (with no incoming raid damage to cause push-backs or affect LT decisions) for at least 2 minutes,
e) ignoring haste-procs as they cannot be modeled to have 100% uptime,
such that a "smooth" destruction rotation (between Immolate, Conflagrate, Chaos Bolt, Incinerate, Curse of Doom, Life Tap) can be achieved?

Considerations:
i) Backdraft as well as lowered GCD time with haste introduce a dynamic aspect to this math (duh)
ii) By a "smooth" rotation, I mean the following:
- Chaos Bolt CD coming off syncing with the end of another spell (i.e. Incinerate / Immolate / LT / Conflagrate / Curse)
- Conflagrate CD coming off syncing with the end of another spell
- Always ending a Chaos Bolt or an Incinerate cast while Immolate is still on target
- Never clipping Immolate

I have not fully attempted to find a solution to this problem - it seems a bit daunting.
If anyone has enough mathematical knowledge to at least set up a way to simulate destruction rotations with haste as the independent variable (as opposed to arduously trialing DPS rotations with a series of plotted haste thresholds), please let us know - I'm sure many raiding destruction locks would be interested in your findings.
If anyone is interested in working together to figure this one out, I would love to discuss this question and share the details of the process and our conclusions here.

A #1: (To be discussed)


Tags: Destruction | Gameplay | Raiding | Rotation | Latency
Posed by: Veange
Q #2: There are two widely accepted rotations (debated) in destruction between
a) Immo -> Conflag -> CB -> Incin ... etc
b) Immo -> CB -> Conflag -> Incin ... etc
so let's settle this once and for all - which ends up being more damage?

Considerations:
the rationale for b) being that
i) There is a certain amount of unavoidable latency between Immolate landing and when you are able to cast Conflagrate. Most good DPSers are aware that in most cases, DPS downtime of any form = really bad. If people here still remember the cast change blizzard has made a few patches ago (remember when you used quartz and /cancelcast macros to time your next cast in order to maximize the number of casts you made, given a certain latency?). For people that do not remember / are not aware: currently how it works is that if while you are casting spell "A", and you press the next key to cast "B" in a specific window (the specific window being, where there is 1.5seconds left on "A" cast to ((your latency here)) seconds left on "A"), the server will consider "A" and "B" to be casted back-to-back and you do not lose any DPS time due to latency. If this was TLDR or unclear to you, basically right now spamming buttons circumvents the caster DPS loss problem due to latency. The idea is, you can spam CB or Incinerate as you cast Immo to not lose DPS loss to latency (CB if only it's not on CD), where as if you spam Conflag as you cast Immo, server will not recognize your cast as being valid since Immolate is not on the target.
ii) There is an argument that Conflag -> Incin -> Incin -> Incin is better than Conflag -> CB -> Incin -> Incin period. (Note that these two are only opener rotations, as CDs will line up your cast decisions differently depending on your haste, tactical situations etc). The argument behind Conflag -> Incin -> Incin -> Incin being better is that Backdraft increases the cast time of the next 3 Destru spells by 30%, and thus the buff ends up reducing more cast time for you if it affects 3 incinerates, instead of CB -> Incin -> Incin, whereas CB has a lower cast time than Incin. However, the counter-argument is that Backdraft increases the DPCT of your next 3 Destru casts, so increasing the DPCT of a higher DPCT spell (Chaos Bolt) is more beneficial than increasing DPCT of a lower DPCT spell (Incinerate). To simplify our discussion here, we will not consider the cases where Chaos Bolt goes under 1.0 second cast time and causes you to be GCD locked. Of course, this simple debate can be reduced to crunching out numbers and test runs. Again I'm hoping to find someone with superior math skills than I, or others interested in dissecting the math / simulation behind this with me.

iii) A third consideration, almost splitting hair at this point, is whether it is better to put conflag on CD or to put CB on CD.

A #2: (To be discussed)


Tags: Destruction | Gameplay | Raiding | Rotation
Posed by: Veange
Q #3: I'm sure most raiding destru locks have run into this scenario: a) you are at an open GCD after finishing either a spell or an instant cast spell, b) Immolate will expire on target before the end-cast of your next CB / Incin, c) Conflag is on CD and finally d) CB is off CD, what do you do?
The choices are:
i) The preferable choice is obviously LT for the GCD so you won't have to cast Incin / CB and lose the 15% bonus from Immolate debuff - problem is that it's not always a good time to LT (LT buff already up / High Mana near the end of a fight etc)
ii) Cast Incinerate, lose 15% on this Incinerate so you don't clip your Immolate
iii) Cast Chaos Bolt, lose 15% on this Chaos Bolt so you don't clip your Immolate (but you do get to put CB on CD rather than let it sit unsaved off CD)
ix) Cast Immolate, clip your last tick of Immolate but guarantee the 15% damage buff on your next CB / Incin.
x) Refresh Curse - it would be a rare coincidence to have CoD fall off at the exact time when this happens.
Again, this comes out to be a math problem it seems. I have not done the exact calculation on this but it seems apparent to me that 1 Immolate tick does more damage than 15% of even an Incinerate crit, so the choices come down to ii) and iii) when LT is not a good option. If anyone with superior math skills could enlighten us on what's the most efficient choice, or if anyone interested in discuss this in greater detail to find a definitive answer, please let us know!

A #3: (To be discussed)


Tags: Destruction | Gameplay | Raiding | Rotation | Latency
Posed by: Veange
Q #4: I find that Destruction is different from Affliction in that CB and Conflag both need to be casted as soon as they are off CD to maximize DPS efficiency; There is rarely a situation where I need to cast Haunt on CD. The problem arises, however, when I am going through a cast or a GCD but cannot queue up CB or Conflag as my next cast to eliminate latency DPS loss. The client-server communication protocols simply does not allow you to queue up CB / Conflag while the your client recognizes them as being on CD even while the server already recognizes them as being off CD and ready for cast, the client-server setup would much less recognize them for being castable immediately after your current cast / GCD. For the TLDR: you can't queue up CB / Conflag as your next cast until their CDs are up. The question now is, assuming 100ms, with how much time left (something between 0.1 to 0.5seconds) on either CB or Conflag CD should you actually WAIT for them to be cast-ready (eat the bullet with damage loss in no-cast-time) so that you do not delay your CB / Conflag cast with an Incin or LT?

A #4: (To be discussed)

----------------------------------------------------------------------------------------------------------------------

All answers / comments / discussions / inquiries are welcomed. I will append your contributions to the specific questions / problems!

If you have any similiar questions regarding in-depth warlock gameplay discussions that are not suitable for the Simple Warlock Q/A thread, post it here and I will add them to this post.

I thank you all for your patience with me. I hope this topic will hang around for a long time and stay active to aid all interested players in the Warlock community!

With regards your Conflag -> CB -> Incin -> Incin vs Conflag -> Incin -> Incin -> Incin question, that's not quite what you're asking. What you're actually asking is - Conflag -> CB -> Incin -> Incin -> Incin vs Conflag -> Incin -> Incin -> Incin -> CB. Each will benefit equally from Pyroclasm but the latter takes less time to cast. However, CB hits harder and has a cooldown, so the question there is whether it's more beneficial to cast an Incinerate 30% quicker and leave CB off cooldown for three casts, or just to use CB. CB hits on the order of ~22% harder than Incinerate does based on a SimulationCraft run I did yesterday, so there will be a certain haste threshold where your GCD cap changes things, but otherwise you're looking at:

(0.7*1*2.25)+(0.7*1*2.25)+(0.7*1*2.25)+(1*1.22*2)=7.165
(0.7*1.22*2)+(0.7*1*2.25)+(0.7*1*2.25)+(1*1*2.25)=7.108

Thus, you should leave CB until Backdraft has expired if you ignore the cooldown. However, given how much harder CB hits, a 0.8% increase in damage for that four spell rotation isn't going to make up for leaving CB off cooldown, as even if every spell is GCD-capped and you can cast 11 Incinerates between CBs your average filler spell damage (if you treat CB as a sort of hard-hitting filler) drops by a lot more than 0.8%.

Also, Conflag buffs your other spells and has a higher DPET, so always put Conflag on cooldown rather than CB.

I would like you consider choice ix) again. For me clipping Immolate in such scenario is not big loss. After all we not cry after Immolate ticks lost beacause boss died before last Immolate run full time.
Example:

Fight: 100 seconds
With back-to-back casting you would cast Immolate 7 times dealing 7 times DD immolate component and deal 33 immolate ticks - 2 ticks are wasted beacause boss died faster than last immolate last. For me it means you can clip up to 2 ticks during this fight without dps loss.

If you clip 1 immolate tick during this fight you still deal 7 DD component and deal 33 ticks.

Yep. I recently moved guilds to raiding with an affliction lock only, to sometimes having 4 destro locks in raid. The delay definitely goes away with another person's immo on the target. It can be confusing as the conflag button lights up but isn't actually clickable.

I disagree with your logic. Of course with absolute knowledge of the exact fight length and flawless timing, you could clip immolate to cater to the exact length. However, for example with a 120 second fight, eight immolates, with perfect timing, would be sufficient to have immolate up from the start to finish. Any sort of clipping would hurt you significantly, because you lose the immolate tick, and you would need to recast immolate at the end. You aren't going to know the fight length in a realistic scenario, and it will be better to not clip immolate.

Looking at a recent parse, my average immolate ticked for 1,550. Comparing to chaos bolt, my average chaos bolt was about 10,100, which would mean fire and brimstone accounted for 1,319 to chaos bolt on average. So in the odd scenario that the time remaining on immolate is less than chaos bolt cast time but greater than the immolate cast time, and you have sufficient mana (life tap isn't an option), it would be slightly better to cast chaos bolt rather than clip immolate. It may be that incinerate > immolate > chaos bolt is the best option, but that is a different argument. It is such an infrequent occurrence, and is not going to have a major impact on your dps.

Of course we dont know exact fight length. But i would say there is bigger chance we will waste some immolate ticks at the end than fight will be exactly timed 15X (where X is 1,2,3,4 or n).
In my opinion there is 80% chance immolate we loose at least 1 tick.

It is better to not clip immolate generally but we are talking only about specific scenario.

From the other hand if we clip immolate instead of casting chaos bolt we postpone chaos bolt (it is not on cd longer than optimal).
Looks like it can be dps hurt if through this postpone fight will end before our last CB will be ready to cast.

Exactly

I just wanted to say two things...

1) This last page has touched on the exact question I had been styming over and unable to definitively determine using target dummy testing. Thanks a lot guys for discussing this in-depth.

2) I see the guide recommends spellstone in most scenarios, but I've seen others discussing Firestone taking precendence (non-EJ forums). Assuming 3/13/55 and average gear (iLvl 200-213 gear), which is the preferred choice? Is there a gearing point where one overcomes the other?

Pretty sure, if you are lacking haste, you should definitely go for the spellstone. Using me for example though, I have 499 haste with my current gear (add my scale of fates trinket use, and i have almost 1000 haste lol), so using a spellstone wouldnt help nearly as much as the 1% direct damage + crit that the firestone gives me. Really depends on if you're lacking haste or not.

There is no choice, firestone is better in every way. 60haste is very minor when compared too 1% DD on 3 of your top damaging abilities.

The guide is out of date. All the deep destro simulationcraft runs use firestones and do higher damage with them over spellstones. http://elitistjerks.com/f80/t48311-s...ost_3_1_specs/

Spellstone is better for anything that can get imp stones or for an affliction spec.

I have noticed the conflag button as light up due to another lock's immo; however, this has not eliminated the delay period in casting conflag right after immo in my experience.

Im curious, with the new profession changes in 3.2, what will become the essential DPS professions?

The profession changes are aimed at making (almost) every profession viable for raiding.

Each crafting profession is balanced around the 2 extra epic gems that Blacksmithing gets. Engineering is still going to be slightly weaker in terms of pure numbers in PvE and Tailoring is still going to be the strongest thanks to Lightweave still being amazing (and getting buffed).

The gathering professions still have very little to offer raiding warlocks and are better saved for alts if you are looking to min/max.

Heres something i have been wondering...

If Glyph of LT scales w/spirit and it is going to be in our main rotation, should i stack spirit enchants over spell pwr?

in addition to this is this glyph better to use then the one that prolongs conf.? and if so can someone break down why it is