Actually, the crit bonus is applied to the mob, so we do not lose this when Demonic Pact is up. I just learned this the other day.
It actually makes shaman somewhat less than optimal for the crit buff too because all casters need to be in range of the totem for the spell power buff and the boss has to be in range for the crit buff. This is somewhat difficult to do in practice especially on movement heavy fights or on fights that force the raid to spread out into several groups. Unless they have extended the range of the totem significantly.
It actually makes shaman somewhat less than optimal for the crit buff too because all casters need to be in range of the totem for the spell power buff and the boss has to be in range for the crit buff. This is somewhat difficult to do in practice especially on movement heavy fights or on fights that force the raid to spread out into several groups. Unless they have extended the range of the totem significantly.
It's 40y, so it's rather easy. On the other hand, crit part doesn't stack with heart of crusader, which will be provided by paladin of any spec.
It's 40y, so it's rather easy. On the other hand, crit part doesn't stack with heart of crusader, which will be provided by paladin of any spec.
unless it's a fight where you have people on all sides of the boss. It's unlikely that people on both sides would be within range of that same 40y totem.
Do the extra ticks continue to increase in damage? Or does the last tick simply get repeated?
I believe that extra ticks are being added in the middle, resulting in no increase in damage at the end. Overall it does seem like a strong affliction glyph, the extra ticks will mean less refreshing and allow for either better timed refreshing of DoTs or a few extra shadow bolts.
unless it's a fight where you have people on all sides of the boss. It's unlikely that people on both sides would be within range of that same 40y totem.
Read all the above posts again, the 3% crit is not a buff applied to players but a debuff applied to the mob so the only requirement is for the totem to be within 40 yards of the boss.
Regarding the Imp:
You guys are complicating the issue. Firebolt costs the same amount of mana, regardless of its cast time. However much mana the imp gets over the course of a fight, from whatever sources, that's how much firebolts he can cast. DP lets you redistribute this damage. DP will never do less damage. It will do more damage if either 1. the imp would not have been able to use all of his mana for the fight, or 2. by redistributing the damage into smaller casting segments you get more OOC regen which translates to more firebolts.
Emp Imp is a concern. There is a slightly higher chance to double-proc Emp Imp within one cast, and lose a proc as a result. However, if your imp will OOM anyways, I think it's higher DPS in either case to stop your imp after he crits until you launch your next spell and have him re-attack.
We simply lack the data...... and it is probably quite a pain to extract now that regen ticks are so frequent.
If people have some good ideas..... please come forward!
We'll need inside/outside the 5SR data.
We'll need data points at different Int/Spi values, etc.
It isn't going to be fun.
I'll take a crack at getting the data; I'm enough of a Masochist that I think these things are fun.
Summary of the following Theorycraft: Generally speaking it is better to let your Imp regen to full mana than to let him keep trying to cast.
To figure out which style of Imp management is better we can completely ignore the period in time it takes for the first OOM. This should be identical whether you plan on resting your imp or letting him sputter. So our theoretical measurement starts with one Imp sputtering and the other passive.
Without any outside help or mana regen while casting our sputtering Imp will spend 7-7.5 seconds in the FSR per cast. That makes the interesting point that even if your Imp is sputtering Demonic Power is a slight DPS increase as it gets the Imp out of the FSR faster. In any case, the sputtering Imp's cast time will be 7 second + the time it takes to regen enough for a Firebolt. Our passive Imp's cast time will be 2-2.4 seconds while casting* once he's full up on mana. So the math looks something like this with T being the time it takes to regenerate one Firebolt outside of the FSR, X being the number of Firebolts the Imp can cast before going OOM, and D being Firebolt's damage.
Damage is immediately shown to be irrelevant, so we can remove that variable entirely. We can also approximate X. Imps have roughly 1/3rd as much mana as their Warlock master at my gear level, which puts them at around 3500-4000 mana. Firebolt costs 180 mana at 70, so that's approximately 20 Firebolts per mana bar. Eliminating D and plugging in 20 for X we get the following.
These two equations can never be equal, and it's clear that letting your Imp sputter is a severe loss of DPS. In order for sputtering to be preferable in this situation the resting Imp needs to have enough mana to only cast 1 Firebolt before going OOM.
Obviously this is a flawed setup, there's no accounting for outside sources of regen. Here's the same equation modified to account for some amount of mana regen inside the FSR. M represents mana regen in seconds per second, and is assumed to be a fraction of T. EM represents the sum of M^Y as Y goes to infinity.
(D)/(7+(T-7*M) = (D*(X+X*EM))/((X+X*EM)(2.4+T)+5)
Damage is still irrelevant. I'm going to assume that M represents a startling .5 seconds/second and solve from there. The value of EM at .5 is a convenient 1.
Again, the equations can never be equal, and again this favors the resting Imp.
Within the constraints of the game and from a mathematical standpoint there is no way for sputtering to be better than resting. At the very best the sputtering Imp can draw even either by going up against a resting Imp that only has enough maximum mana to cast one Firebolt at a time, or by having enough mana regen such that neither imp ever runs out of mana. From a practical standpoint the only time sputtering will be preferable to resting the Imp is in the very last stretch of a fight, or if you're trying to keep Demonic Pact up.
I'll still try and get the hard data so we can simulate the difference, but given the theorycraft above I'm extremely confident that it will be shown the Imp, and by proxy the Warlock, do more DPS when the Imp is given the chance to rest.
What I lack in intelligence I make up for in verbosity.
I still haven't seen a good write up on how the raid replenishment buff works, but I think that's a pretty big issue in this situation. If the imp can be consistently at low mana he'd get a lot of mana regen and probably go to OOM far less frequently.
Regarding sputtering: It's a pretty intuitive result that regenning is better. Whenever an imp stops casting and goes to a regen cycle, he wastes 5 seconds starting that cycle. A sputtering imp pays that 5 seconds every sputter, a regenning imp pays it once. Sputtering imps lose more cast time. It's nice to get some hard numbers on how much better a regenning imp does, though.
Replenishment: That's a tricky proposition. Depending on your raid composition, that could be sniping the Replenisment buff from a needful healer, or a needful mage even. In general, the Replenishment buff trigers off an event (Mindblast on a VTed target, any crit by an SV hunter, or a retadin's judgement). On that event, the 10 targets with the lowest mp (as a percent) recieve the replenishment buff. When that same player triggers the event again, the replenishment buffs are re-assigned based on new mp totals. When a second player with replenishment triggers the buff, the 11th through 20th lowest-mana players get replenishment, and so on. This means you'll want your imp to blow through his mana pool as fast as possible, and then sit OOC until his mana approaches that of the 10th/20th/etc-lowest mana person in the raid. There's probably a small amount of tradeoff of extra time inside 5SR vs extra time in replenishment, but as long as your imp isn't yo-yoing ever 15 seconds it probably won't come out to much. My intuition, however, is that the .25%mp/sec of replenishment is considerably less than an imp's natural OOC regen rate so the benefit of maximizing replenishment is negative due to less 5SR regen.
Regarding my original post that DP is a DPS loss, my logic is more in the realm of spreading Empowerment procs than in efficiency.
Also, he ends up lasting twice as long because the decreased cast time gives me more chances to proc Improved Soul Leech. The time will be variable, but me saying "twice as long without DP" is based on a dozen or so target dummy runs with and without the talent while DPSing along with a ret pally.
It may prove to be more DPS to use DP and just micro-manage the imp when you get procs, and DP or not it seems obvious that letting the imp regen to full is the best way to use him. Of course, how easy will it be in actual combat to micro the imp every time he crits? This is on top of timing conflags and watching your dots while also reacting to fight mechanics.
Still, I wouldn't dismiss my original post without a bit more thought with regard to the Imp's OOM time being twice as long without DP. There's more to it than it may seem.
Replenishment: That's a tricky proposition. Depending on your raid composition, that could be sniping the Replenisment buff from a needful healer, or a needful mage even. In general, the Replenishment buff trigers off an event (Mindblast on a VTed target, any crit by an SV hunter, or a retadin's judgement). On that event, the 10 targets with the lowest mp (as a percent) recieve the replenishment buff. When that same player triggers the event again, the replenishment buffs are re-assigned based on new mp totals. When a second player with replenishment triggers the buff, the 11th through 20th lowest-mana players get replenishment, and so on. This means you'll want your imp to blow through his mana pool as fast as possible, and then sit OOC until his mana approaches that of the 10th/20th/etc-lowest mana person in the raid. There's probably a small amount of tradeoff of extra time inside 5SR vs extra time in replenishment, but as long as your imp isn't yo-yoing ever 15 seconds it probably won't come out to much. My intuition, however, is that the .25%mp/sec of replenishment is considerably less than an imp's natural OOC regen rate so the benefit of maximizing replenishment is negative due to less 5SR regen.
Is it really that much of a problem? I'm under the assumption that all 25 man raids would have at least 2 replenishers, and 10 mans would have at least 1. That covers 20 and 10 people. In 25 mans, you're going to have a small handful of classes/specs that don't use mana, and in 10 mans you'll have one or two, leaving a very small number of people (who are at a higher mana than the rest of the raid) that don't get replenishment.
Regarding my original post that DP is a DPS loss, my logic is more in the realm of spreading Empowerment procs than in efficiency.
Also, he ends up lasting twice as long because the decreased cast time gives me more chances to proc Improved Soul Leech. The time will be variable, but me saying "twice as long without DP" is based on a dozen or so target dummy runs with and without the talent while DPSing along with a ret pally.
It may prove to be more DPS to use DP and just micro-manage the imp when you get procs, and DP or not it seems obvious that letting the imp regen to full is the best way to use him. Of course, how easy will it be in actual combat to micro the imp every time he crits? This is on top of timing conflags and watching your dots while also reacting to fight mechanics.
Still, I wouldn't dismiss my original post without a bit more thought with regard to the Imp's OOM time being twice as long without DP. There's more to it than it may seem.
In relation to DP and Emp Imp, I'd want to see test parses because I don't think skipping DP is that beneficial for practical reasons. Basically you're arguing about the opportunity cost of back-to-back imp crits within the space of a single destruction spell cast. Unless your imp is sitting at 50% or higher crit rate the opportunity cost is low, and gets lower with the more haste you have. During Backdrafted casts, the imp having a higher cast time would actually be a good thing, as your cast speed on certain spells would be faster than the imp's firebolt. It's all about relativities.
In relation to your second point, imps don't proc Soul Leech/Imp Soul Leech, only certain of your destruction spells do.
Overall, I'd hazard a guess at DP being a DPS gain just due to the higher possible output and I don't think it would be worth micromanaging imps to maximise Emp Imp procs even if they did have a high crit rate. Remember, Emp Imp is a 20% boost to crit rate, not a guaranteed crit (unless you're sitting at 80% crit) and subject to RNG like base crit. I'd just let it do its thing and concentrate on more important things.
Regarding my original post that DP is a DPS loss, my logic is more in the realm of spreading Empowerment procs than in efficiency.
Also, he ends up lasting twice as long because the decreased cast time gives me more chances to proc Improved Soul Leech. The time will be variable, but me saying "twice as long without DP" is based on a dozen or so target dummy runs with and without the talent while DPSing along with a ret pally.
It may prove to be more DPS to use DP and just micro-manage the imp when you get procs, and DP or not it seems obvious that letting the imp regen to full is the best way to use him. Of course, how easy will it be in actual combat to micro the imp every time he crits? This is on top of timing conflags and watching your dots while also reacting to fight mechanics.
Still, I wouldn't dismiss my original post without a bit more thought with regard to the Imp's OOM time being twice as long without DP. There's more to it than it may seem.
There's the potential for losing Empowerment procs even without DP. It's a basic risk that you may get back to back Imp crits at bad times. Between Conflagrate and Immolate, or after a DoT refresh/Lifetap you run the risk of back to back proc with or without DP. The only danger unique to DP is that you might get back to back procs between Incinerates. However, you will not be in sync with your Imp's casts frequently enough for back to back procs to sneak into that space often. The amount of DPS gained from trying to micromanage around Emp Imp procs is probably lost from both lost Imp DPS as well as splitting your attention unnecessarily.
Moreover, I don't see how mathematically an Imp without DP can do more DPS than an Imp with DP. Barring a truly amazing benefit from hovering low enough to be a favored target for Replenishment procs (something that can actually hurt the rest of the raid), there is no mana the slow Imp will gain that the Fast Imp won't, moreso considering JoW. The slow Imp will simply be doing identical damage, but slower.
Changing subjects to Imp mana theory, in collecting data for SimulationCraft I found that the Imp actually has some passive Mana Regen While Casting. It appears to be ~30%. The only Imp talent I had for these tests was DP, so it's either passive or granted through that talent.
I'll be back with Imp numbers for SimulationCraft in a bit.
What I lack in intelligence I make up for in verbosity.
Here's the problem:
Demo uses dots, and all key scaling is too far down the affliction tree.
Demo uses nukes (would use the better of incinerate/shadowbolt) but the scaling talents for these are now too far down the destro tree with the ISB nerf.
Meta can't be buffed anymore (we'll be lucky if the dmg increase component doesn't get nerfed) because of PVP balance.
Where does that leave demo?
Soul link dmg modifier is gone and is not coming back due to current placement in the tree.
Changing demonic tactics back to its original TBC form (+5% dmg) won't happen and would go against demonic pact tree design.
Changing Demonic Knowledge back to 15% max rank might help initially, but % increase spellpower talents don't scale all that well (see Arcane Mage and mind mastery) in general.
hmmm....I guess my solution would be to either to allow some %modifier on total dmg that only demo could get for pve (master conjurer and/or demonic empowerment would both be logical choices). One could argue for more buffs for the pet dps instead; this could also be possible, although it would probably heighten the design problem of "useless without pet/OP with pet" that has been alluded to by the devs regarding hunter and warlock "pet specs".
Overall, the playstyle, cooldowns and procs that demo gets are interesting enough, its scaling on the spells the warlock uses (affliction and destruction spells) that are holding it back atm.
I agree with your assessment. The variety of spells that I make use of as demo is one of the reasons I love it, it just feels really versatile, and I rarely feel bored playing it. But you're right, the scaling does start to hurt, though maybe it won't happen as quickly when I'm able to get ruin in the mid 70s. I think adding a +10% to all damage in there somewhere would help a bit, maybe adding it back to soul link and putting it back at the 31 point slot. Demonic empowerment just feels weak to me, putting it lower in the tree would hardly be overpowered, in my opinion.
I think the point you made kinda applies to the other specs as well. Blizzard has made an obvious effort to make all specs use their pets actively, they don't want passive mana batteries or have them sacrificed for buffs. They also want affliction to use nukes as part of a rotation, and they want destruction to use dots, which historically ignored them for the most part. I think it's good that each spec use more than just their own tree, I don't think its right to have to completely ignore part of your spell book because the spells are too wimpy. The challenge will be keeping off spec abilities from falling too far behind, to the point where we stop using them.
Tests were from near zero mana to full. Regen numbers include 4-5 seconds of mana regen within the FSR. Imps don't scale with Spirit, which makes isolating variables easier. The second set of numbers are with Divine Spirit.
What I find interesting is that the Imp gains mana regen from Int such that even with the extra mana he'll fill his entire bar faster. I'd be interested to see numbers at 80, I'll bet that the Imp takes a hit from the spirit regen change.
What I lack in intelligence I make up for in verbosity.
Demonic empowerment just feels weak to me, putting it lower in the tree would hardly be overpowered, in my opinion.
I've run the simulationcraft - Google Code simulator over and over and Demonic Empowerment ends up being a 40 DPS loss (using the felguard anyway). Can anyone else confirm?
*edit* Nevermind, I just downloaded Version simcraft-r698 and now it's showing as a 30ish DPS gain.
I tested Curse of Doom compared to Curse of Agony (skilled out in affli tree). And they nearly did the same damage in one minute (that´s 1 CoD until it impacts vs 2 CoA + 8 sec of a third CoA). If i recall correctly, Curse of Doom did a little more than CoA. Can someone confirm this? If yes, is there any reason to spec and cast CoA? This would make an affliction rotation much easier or would open up time windows for Immolate or another SB.
Does this comparison include shadow embrace and haunt? The last time I checked, they affected CoA but not CoD. Was your test done at level 70?
Last edited by Kalle : 10/09/08 at 6:34 AM.
Reason: spelling
Read all the above posts again, the 3% crit is not a buff applied to players but a debuff applied to the mob so the only requirement is for the totem to be within 40 yards of the boss.
Read the post you quoted again.
Originally Posted by tetracycloide
It actually makes shaman somewhat less than optimal for the crit buff too because all casters need to be in range of the totem for the spell power buff and the boss has to be in range for the crit buff.
He was talking about before everyone has 2800 spell power, the raid will need to be within range of the totem as well as the boss. Unless you're putting it right on top of the boss, and the boss doesn't move, it could be a problem.
And I assume the conclusion is, there may be some fights where the 10% spell power from a Demo lock will be useful even with an Ele Shaman and <2800 spell power (at least for parts of the raid), simply because a 40y range totem is subpar for movement fights, both for debuffing the boss and for being in range of casters.
Originally Posted by Kalle
Does this comparison include shadow embrace and haut? The last time I checked, they affected CoA but not CoD. Was your test done at level 70?
I haven't done Aff testing on my premade, but I can confirm that 60 seconds of CoA was about equal to 1 CoD in terms of damage for a Destro Lock with 2/2 Imp. CoA. I didn't test with the Glyph yet, I'm not sure if that would tip the scales significantly.
Originally Posted by Heeno
I believe that extra ticks are being added in the middle, resulting in no increase in damage at the end. Overall it does seem like a strong affliction glyph, the extra ticks will mean less refreshing and allow for either better timed refreshing of DoTs or a few extra shadow bolts.
I assume you mean there isn't a new "high" for ticks with the Glyph and not that there's no increase in total damage dealt by CoA by the end of the DoT with the glyph?
how is the imp going oom in a raid environment? BoW, Mana Spring and replenishment will be given to him anyway....so how much more mp5 would imp need in order to not run out of mana?
Does this comparison include shadow embrace and haunt? The last time I checked, they affected CoA but not CoD. Was your test done at level 70?
It was done to the boss-level dummy and no, haunt was not up. I thought if haunt increases damage by a fixed value, it doesnt matter if i compare 100% of damage between the 2 spells or if i compare 120% of their damage. But if its true that haunt doesnt affect CoD, i probably should redo that test. Didn´t someone stated somewhere in this thread that CoD was technically a DoT with one tick?
To summarise the Imp/DP issue as I see it (and correct me if I missed something) :
DP means more DPS until oom, and a shorted oom time.
It means more Emp Imp procs until the oom time, and less between the DP oom time and no-DP oom time.
It means a higher chance of getting replenishment / or the imp will start getting it earlier, so (slightly) more mana regen over the course of the fight.
The point about soul leach is that if you are DPSing (and why wouldn't you be?) you will proc soul leach, which gives the imp some 'free' casts.
In my PTR testing just now (average gear ~T5 equiv) an OOM imp with no soul leach or other mana procs does 13 fireballs / min, 4.6 seconds a cast. With BoW, JoW, Soul Leach, replenishment and totems, I don't imagine that the imp being oom is actually much of a DPS decrease compared to no-DP . The micro-management (turning off autocast when its oom and turning it back on once its full) it probably not worth the effort, although it would be an increase.
In the end, I don't think it's going to be a big deal either way, and DP is the closest to a good filler talent to get to Devestation/Imp Imm. (My simulationcraft testing has showed that never casting shadowbolts is just as good as only casting them off nightfall procs, which is better than the default casting priority, so ISB is worthless).
Tests were from near zero mana to full. Regen numbers include 4-5 seconds of mana regen within the FSR. Imps don't scale with Spirit, which makes isolating variables easier. The second set of numbers are with Divine Spirit.
What I find interesting is that the Imp gains mana regen from Int such that even with the extra mana he'll fill his entire bar faster. I'd be interested to see numbers at 80, I'll bet that the Imp takes a hit from the spirit regen change.
Thanks for the details..... I'll start parsing them to see if I can come up with a generic formula.
Were you able to confirm the 30% regen while casting? An interesting method of testing might simply be time-to-OOM for various levels of Int. Since we know the cost of each spell we can compare how much mana was spent vs how much mana the Imp started with. This difference is his "while-casting" mana regen.
I don't want this info getting lost in the shuffle so I will post it here and hopefully get some guidance...looking for advice on level 70 builds after the patch.
This seems to be the most difficult part of figuring out new specs after all we have learned over the current expansion...forgetting everything that worked before and trying out new things. So I am trying some 'Lock builds for the patch on Tuesday (!) and level 70 builds. Let me know what you think. In my opinion everybuild MUST have Bane for the increase in shadowbolt casting speed.
Affliction - I'm not really sold on Nightfall anymore...especially when you can get a Glyph of Corruption to do basically the same thing and not spend talent points on it...although if they stack that would be amazing and OP (so I doubt that they will), but I can't find anyplace else to put points into to get down to Haunt..so there it is. Talent Calculator - World of Warcraft
Destruction - Meh. Don't really like anything here...got the Soul Link because with a 2 piece T5 set it will offset the shared damage with your Imp...I am assuming that will be the pet that is out because the way the Destro Tree is now with a combination of nukes in that build being both Shadow and Fire it will take some time to figure out which pet to sacrifice for max dps. Also...it seems that Blood Pact will be raid wide now so you would have that benefit as well. Talent Calculator - World of Warcraft
Demonology - Gawd I hate this build. I figure you would keep CoA up on the mobs...I put those points over there because Fel Synergy felt too much like T5 bonus to me and Demonic Empathy was 3 points for too little gain in damage. No Demonic Sacrifice because it seems that the bulk of the bonus damage in this tree comes from the pet being alive and in your control. Seems to me that you would be keeping your Felguard out and casting Metamorphosis every cooldown based on your situation. I tried Metamorphosis on the PTR and it was pretty "neat" but felt a little gimmicky to me. Although it is a nice talent to have right at 70 out of the gate. Talent Calculator - World of Warcraft
10/08/08, 6:05 PM
. The micro-management (turning off autocast when its oom and turning it back on once its full) it probably not worth the effort, although it would be an increase.
Similar Threads
