There used to be a problem where certain levels of Haste (just before the thresholds for extra ticks) reduced the duration of Immolate to such an extent that it was very difficult to refresh it using Hand of Gul'dan. This meant that there were haste values we wanted to avoid. This is no longer the case because with the newly increased duration of Immolate it's impossible for this interaction to occur, so there are no more Haste thresholds for Immolate: Haste now linearly increases the dps of Immolate as long as you can refresh it with HoGd.

There are still haste thresholds for our other spells where they gain extra ticks and the important ones are noted in the OP, but they're much less significant. They represent a dps increase (perhaps higher than what you could have achieved with other stats) but getting near and failing to reach them can no longer be a dps loss.

Immolation Aura and Shadowflame are both spells that particularly benefit from the haste thresholds because their cooldowns are much longer than their durations. With other spells it is more a case that they become suddenly more efficient at the thresholds because you have gained a tick period's worth of time where you don't have to refresh them, clearly less time refreshing DoTs leads to more time on filler.

TL;DR - The value of haste may change because some of our spells have thresholds where they gain significant value, but gaining haste can no longer be a dps decrease. As to whether Haste is worth more or less than other stats before or after any thresholds, you would have to generate your own scale factos to find out, but doing so near the thresholds might be a good idea.

Answering a quite old post here :

I don't have simulations to support this either, but many raid tries.

In my experience, starting the fight with the felguard and then swapping to succubus does give a huge dps boost, however, swapping in-fight back to felguard to get his demon soul ability seems to be a dps loss (since you have to stick with him for 45s or hard cast the succubus summon).

The best way to handle this, for what i know :

Soulburn roughly 4s before pull, in order to still be able to use your potion before the pull,

Start your usual spell cycle (Meta > Demon soul ... ) while using felstorm and pop the succubus when you're about to start casting some shadow bolts (you should still have the soulburn buff) and go on with your fight normally.


I usually get a 40-45k burst at start, sometimes going over 50k.

As previously said however, this is only for single target fights, since you'll keep the felguard in aoe situations.

I've been having an ongoing argument with my raid about the demon spec,

Generally they see demonology as a lesser spec or "the worst spec for raiding" compared to afflic or destro. I agree affliction seems more suitable for some fights, Namely any of the ones with multiple bosses that you can take advantage of glyphed soulswap. Affliction doesnt seem to do any better on single target fights though. At least not significantly better than demon spec does.

My argment to the raid is that the 10% spell power boost clearly outweighs the 1-2k more dps that an afflic spec can bring to the table.

My Question is this then, has anyone determined/shown just how effective the demon spec spellpower boost is?

My feeling on the matter is that using demon spec is clearly better for the raid in the event that your raid doesnt have a a spell power buffer (6% for elemental totem or mage buff) already in your group, and it may indeed still be better for just the 4% extra spellpower your 10% buff provides.

if this has been done, great, please link it for me. If not I would be happy to try to chart this out to show exactly what the cost/benefits are and post it.
I'd need to know just exactly how much 1 point of spell power helps particular classes/specs, even if only as an estimate. has this simcraft been done? I'm guessing yes, to determine stat weights, so please post it or point me to it if the info exists

Thank you.

TL;DR - It's almost certainly better for the raid for you to be Demo.

This used to come up a lot in discussions, but frankly the answer to "is 10% vs. 6% spellpower worth more than another spec" is now very obvious and very clear from even a cursory consideration of the sim results. There are other related questions such as "should I bring an elemental shaman or a demo warlock?" or "I have several warlocks, which should go demo?" that have much less clear-cut answers, but that's not what you asked so I assume that those aren't options.

Firstly, precise stat weights can only be direclty applied to very specific circumstances (the ones for which they were generated) thus making generating precise stat weights for an entire raid essentially impossible, or at least hopelessly specific. As such one cannot "know just exactly how much 1 point of spell power helps particular classes/specs". It's generally not especially useful to do so anyway because it's the relative worth of stats that is generally important when making gear choices, but I digress.

Generally speaking, however, Spellpower is (I believe) simmed as the second best throughput stat for all casting classes (behind only Intellect). Thus, the answer to how much it helps is: "a lot". Do not forget that the effect that the spellpower has on healers is also important, but rather challenging to weigh against the personal dps gain that a non-Demonology spec might conceivably bring you.

However, the theoretical difference between the different Warlock specs is now so small that it currently sims better for you to be demo if you're the only source of the buff even if you're the only caster in the raid.

If you want some numbers, let's ignore a swathe of dangerous assumptions and use some from the Simcraft thread. You are either the 359 demo from the simcraft thread, or the best 359 in those results: Destruction. You do 23863dps in demo and you do 24304dps as the destro warlock before we take off the spellpower buff. Spellpower simmed as being worth 2.3013dps per point and you had 8597 of it when raid buffed. Losing the buff would lose you 313 spellpower, thus a loss of 719 dps, putting you at 23585dps, which is less than the demo spec. Remember: this doesn't even consider everyone else in the raid. The numbers shouldn't really be used this way, but it serves to illustrate that the buff is worth orders of magnitude more to the raid than the difference in your spec.

Now, your raid might be better served by someone else going demo, or someone else bringing an elemental shaman or many other options that only you can know or decide about, but to raid in another spec and go without a 10% spellpower buff at all is, in my opinion, doing yourself and your raid a grave disservice.

As an aside: I am astonished to hear that your fellow raiders would complain about you choosing to buff them at what they perceive to be your own expense!

If you had an elemental shaman in your raid he would already bring a 10% spell power buff, so you being demonology would have no additional effect.

I wanted to highlight this before we go further into "what should my raid do" land, but this is all very subjective and doesn't add a lot to to the discussion. Raids are unique, figure out your own buff situation. And heaven forbid you have to do a little math along the way (Hint: logs are useful, if your raid isn't full of the perfect players that simcraft simulates).

I still don't understand why it lists Demon Soul (Imp) as well. Does this mean you start of with a Fel gaurd, switch to succubus, and then switch to Imp for the second Demon Soul?

The whole pet swaping idea seems like a dps increase, but only once at the start of the fight like someone already noted. The biggest annoyance with this concept is when you do it during combat you lose Soul Link and have you recast it. I don't know if it's a bug or not, cause if you do it out of combat you still have it after you swap pets, but it sure is frustrating.

Typo. Was meant to be succubus.

If you were to do any pet swapping I believe starting with felguard and switching to Succubus after Demon Soul/Fel Storm would be the way to go about it.

Given that we are a more cooldown oriented spec than the other warlock specs; are on use professions more valuable to us?

Earlier discussions have touched on this, that there is a multiplicative effect on stacking cooldowns, so much so that 2 minute items/abilities used can have a greater effect on dps than what thier simple stat averages equate to.

For example, I'm an orc warlock and I get a 585 boost in spellpower on that all-important 2 minute cooldown that we've got going. On a 2 minute cooldown that averages to a 73 spellpower boost. I would argue that I'm getting more out of the spellpower boost than its simple average... Specifically, I'd think I'd be getting at least around 30% more damage effect from the ability since I'm using it during cooldown time (demon soul from FG or succubus, and metamorphosis). That would equate to an effective spellpower boost of 760 ((585*.3)+585) and an average benefit more like 95 total spellpower with even more effect added for mastery.

I guess the question boils down to whether getting a guaranteed 480 haste or int boost at that critical cooldown time is more valuable than a straight +80 int provided by a different profession? This has broader implications for engineering; the tinker would coordinate with our cooldowns meaning that your getting more out of that +int boost than the straight +80 average. Thus making engineering clearly our second best profession, and it also boosts the value of herbalism.

Using the math above +480 haste for 20 seconds every 2 minutes would actually provide us with around 104 effective haste instead of 80. The math for the engineering tinker is more complicated because its on a 1 min cooldown, but it averages to ~92 +int instead of +80((1min =+80 plus 1min=+104)/2). Again leaving out mastery effects for simplicity... An average of +92 int clearly makes engineering our second best profession without the added mastery benefit. The herbalism haste probably doesnt make up the difference from a +80 int profession, but its definately more effective than the OP would indicate..

Is my math right? Should the OP reflect eng as clearly our second best profession choice because of this?

P.S. This probably means some "on use" trinkets are more valuable to demon warlocks than we'd give them credit for too... and it also probably means, mastery is more valuable as a stat, the more "on use" abilities you have.

Blood Fury seems to give 585 spell power, not 173.

I find that often, Meta is up at the 2 minute mark if your using Shadow Bolt or incenerate as your fillers.
However if its not, your right you are losing some effectiveness of your on use items, but thats difficult to quantify.

What your really talking about here is the cooldown doesn't refresh in time. If meta refreshes at 15 seconds before your cooldowns do, your golden, because you have glyphed meta, and none of the use CDS last longer than 20seconds. Its extremely rare that I see meta refresh at 30 seconds before the 2 minute CDS, but when it does happen, I suspect its still better to wait 15 to use the CDS together.

The discussions on this subject that I have read have said that you should wait and use the CDs when the meta is available if its going to be up in a reasonable amount of time to when your CDS are on. Like around 15 seconds. but if its going to be longer you should just pop your cooldowns when they refresh. We really digress on this though, I think this debate about the optimal Meta use is still ongoing...

More to your point, after the 2 minute mark, your only getting the 10% damage boost from demon soul if your unlucky. Accounting for what your saying, you'd get 30% from the first two minutes, 20% in a 4 minute fight, and 16.66% added benefit in a 6 minute fight (averaged percent increases). That would be minimum values, assuming your completely unlucky with the Meta refresh RNG, but for me its rare that they're only up together at the beginning. Usually they will hit together at least twice in a 6 minute fight. The numbers I posted are Ideal numbers, numbers for a fight where you have meta up every 2 minutes. It doesnt change the statement that we still have more effect from on use items, it just means that the longer the fight, the less additional effect we get.

Don't forget though, my numbers leave out mastery, so our actual percentages are going to be higher.

Just to clarify, I think the OP should be updated to say engineering is the 2nd best profession for warlocks because we get more effect out of its on-use int boost. And herbalism's numbers could be updated, to show its more effective for us, your right its not best, but its not just a straight +80 haste for us either. I never stated that either should be considered our top profession. Tailoring is clearly the best at the moment, by a clear margin.

P.S. The engineering tinker is on a 1 minute CD, meaning if your completely unlucky on meta reductions from sb or incinerate, those two cds would still be together.

I don't think that's such a great idea. Between Engineering and Jewelcrafting (it's not up to any debate, that those 2 are the candidates for the second best profession for locks) JC seems like a better profession even more if you apply your own logic.

Engineering bonuses include BiS 359 head and a dps int cooldown (also a nitro-boost/plasma shield, but let's forget about those for now), while JC bonuses include 81int and BiS 372 trinket, which happens to be one of the best "dps cooldowns" for locks there is (here, i am using data from simcraft thread for locks, it could be inaccurate, but i found it to be reasonable within my experience). "Added cooldown benefit" is much better for the "1485sp for 20sec" proc, than it is for the "480int for 10sec" proc, that seems pretty obvious.

One would say "I can use engeneering trinket AND some other trinket that is not so much worse than figurine and get the same good cooldown benefit from that trinket", but no, you can't, because Synapse Springs trigger "on_use_trinket" cooldown. Those 20 seconds make it nearly impossible to align any good trinket effect AND Springs during one meta (your cooldowns would need to align perfectly, or you would need to have drastically less uptime on them). Ofcourse you cannot always mesure the benefit of having a BiS head item vs some other item, that varies with each player's loot-situation.

To conclude: I concur that as a spec, that revolves around metamorphosis, on use effects are a bit stronger for us, than just plain "average over time" number, but it's up to personal preference and your gear setup, whether you would want to use Engineering or JC.

That sounds a bit quirky to me. Succubus' demon soul is nowhere near "10% damage boost", while felguard's could be around that number, but we don't get to use felguard all that often in current content.

I got a question regarding pet swapping in combat.

As the demon soul effect from the Felguard is a bigger benefit than the one from the Succubus we tend to start the fight with Felpup out, popping meta, cd's and demon soul and soulburn out the Succubus.

My question now is:
Is it worth to start the fight with the Felguard's Felstorm and after that switch to the Succi? Because as far as I observed, in-combat pet switching makes you loose your Soullink, which results in recasting it. So you sacrifice one GCD for Felstorm.

The alternative is to pop meta and demon soul and soulburn the Succubus before combat starts. You would loose Felstorm but gain one GCD.

So what's the most ideal way of switching from Felpup to Succubus (assuming you want the Felguards demon soul)? In-combat or pre-combat? Does the use of Felstorm compensate the loss for one GCD?

Would you mind clarifying what trinket you are describing that is JC only, especially one that is heroic ilevelled. Could you link to the object and perhaps the pattern from which it is obtained? As a demo lock only using the Gems, yet doing heroic content I would be very interested in getting it.

He meant Figurine - Jeweled Serpent - Item - World of Warcraft
It's not ilvl 372 but it surpasses most of the available trinkets except Darkmoon Card: Volcano - Item - World of Warcraft

Source: http://elitistjerks.com/f80/t110222-...n_4_0_6_4_1_a/

While this probably is true, sometimes speccing out of Demo is not an option. (for example I raid 10-man lately and our setup doesn't include an elemental shaman or even a mage) So min-maxing your Demo output in most fights is a valid discussion, even with the dualspec option =)

It was answered by Nokwar already, just to clarify, by "372 BiS" I mean "Best in Slot with all of the gear available, including 372 iLvl pieces". If you are a Jewelcrafter there is practically no reason why you'd want to skip figurine, except maybe some dire need of hit.

Swapping felguard precombat equals at the very least 3 seconds lost on the demon soul uptime, probably more. It feels much better for me to let him spin, while I get max uptime of the buff I actually bothered to summon him for.

But that's pure feeling. The math is easy though, if "2 gcds (of fillers probably, since both casting succubus and sould link comes at around the time you are done with DoTing, HoGing and SFing anyway) > 3-5 seconds of Demon soul buff + Felstorm - ~7 seconds of succubus' dps" holds, then you should swap them before fight, otherwise just stick to the felguard and his storm.

Keep in mind that the trinket list is generated at ilvl 359, not 372. I would be very very surprised if for example, Heroic Bell is not better than Figurine at 372. Ratings scale better with the higher levels of intellect/spellpower available on he 372 gear.

Just a quick comparison between scale factors for 372/359 demo shows that the value of crit has increased by 8.3%, mastery by 14.3%, hit by 13.9% while the value of intellect only increased by 3.9%. So logically the double int/sp trinket is going to be the poorest scaling trinket.

1.5 seconds if you prep it right. If you summon the succubus before the fight and cast soul link, then summon the felguard, she'll still have soul link when you summon her again.

I see lots of posts about the felguard/succu fight starts and no decisive answers, I though my post was clear enough, clearly it wasn't, so I'll try to explain this further.

Succu start :

-Preburn
-Prepot
-Curse of the elements (I am the assigned lock for this in my raid, skip it if you're not) (should be cast right when the tank pulls)
-Meta + various on use cooldowns + immolate cast (usually macro'ed)
-Hand of Gul'dan
-Bane of Doom
-Immolation Aura
-Corruption
-Shadowflame
-Soulburned SF
-Incinerate (Molten Core)
-Demon Soul
-Shadow bolt

Felguard/succu swap start :

-Preburn
-Prepot
-Curse of the elements
-Meta + demon soul + various on use cooldowns + immolate cast (usually macro'ed)
-Hand of Gul'dan + felstorm
-Bane of Doom
-Immolation Aura
-Corruption
-Shadowflame
-Soulburned succu summon
-Incinerate (Molten Core)
-Shadow bolt

Felstorm should not be on the meta macro, bear in mind your felguard has first to get to melee range, so don't use it right on start, but not too late either since you want don't want him to have his felstorm unfinished when you swap to succu. Casting it at the same time as HoG should do the trick.

So, let's sum it up :

What you lose with the felguard/succu swap start :

-succu's dps for roughly 10s
-succu's demon soul
-one Soul Fire (you need one gcd to summon the succu in the succu/felguard swap and you use one gcd to cast a soulburned SF in the succu start so I sum it up by stating that you are losing the damage of the soul fire)

What you gain :

-felguard's dps for roughly 10s
-felguard's demon soul


Firstly, with the felstorm, the damage output of the felguard and the succu on a 10s basis is roughly the same so let's not consider them.

What we need to weight here is on one hand the felguard's demon soul and on the other hand the succu's + the damage of one soul fire.

For that, let's say you are unlucky and don't get any molten core procs (a 10% buffed shadow bolt is STILL way under a molten core'd incinerate), that way, you get the most out of the succu's demon soul.
Shadow bolts account for around 50% of your non-pet damage, so the succu's demon soul grants you roughly 5% bonus damage for 20s.
On the other hand, felguard's demon soul grants you 10% damage (not only on your shadow bolt !) AND 15% haste.

So, comparing the two demon souls, you gain 5% damage and 15% haste with the felguard's.



Last question : does a 5% damage + 15% haste boost gives more than one soul fire damage ?

In this section, consider that I don't take into account the pet dps.

(Bear in mind that at that moment, you're in meta with all of your cooldowns)

Let's say that in meta, with all cooldowns (except for the demon soul one), you do 20k dps.
The 5% damage boost ONLY gives you (20 * 0.05) * 20 = 20k damage bonus, which is roughly the damage output of one soul fire (in meta + cds).
That shows that 5% damage + 15% haste for 20s gives way more than one soul fire.

And finally, we can see that the succu/felguard swap start is way better than the succu one.

Above 1993 haste, Mastery has a higher stat priority than Haste, would it be beneficial to gem Artful Ember Topaz instead of Reckless Ember Topaz? (sorry in advance for not linking, but artful is 20 int/20 mastery verses 20 int/20 haste)

It is much much superior to soulburn summon Succubus after poping your cooldowns in most fights. Remember, Succu has to run to encounter and if you stand far away in a ranged camp, you automatically lose a lot of dps potential and if your raid pops bloodlust in the first seconds, it is definitely better to buff Succubus with haste.

Another thing is Master Summoner. You can put 2 points into this talent and the result is 4 seconds hard summoning. It makes sense to hard summon Felguard and soulburn summon Succubus after cooldowns. Optimal way is to do it in a Shadowbolt filler phase. You lose two shadowbolts in avg 2k dps and win avg 6k dps from felguard demonsoul. That makes a 4k dps win with 23-24k dps. If your dps is higher you win much more from your felguard demonsoul.