Eviscerate won't be any better than it already was, as it would require you to materialize three talent points out of nowhere to take Improved Eviscerate (i.e. you'll have to drop good talents later in the tree) and it still only gets 5% of your AP per combo point, while also being reduced by armor. Envenom is not reduced by armor, gets 7% of your AP per combo point, and increases your poison proc rate for several seconds. There's no contest. The difference is in three-digit DPS, if my sheet is to be believed.

Couple more questions:

1) Does Prey on the Weak affect poison damage, or just physical attacks?
2) Has it been determined how the Rupture bonuses from Serrated Blades and Blood Spatter stack?
3) Am I correct in assuming that the Rupture Glyph gives 2 more ticks at normal strength? That is, it makes Rupture doe do 25% more damage over 25% more time, rather than the same amount of damage over 25% more time?
4) Similarly, has it been confirmed that Imp SnD stacks with the SnD Glyph the same way it did with the T4 2/5 set bonus?

1, 3, and 4 are all questions I've been wondering about myself, but I haven't gotten around to testing. As for question 2, judging by the way Find Weakness, Surprise Attacks, Aggression, Opportunity, and T6 4pc all work additively with one another on their various affected abilities, and Find Weakness being additive with Vile Poisons on Envenom, I'd say it stands to reason that Blood Spatter and Serrated Blades are additive with one another. However, I'll confirm that tomorrow as well as the others.

So I ran a bit of testing on the Heroic Training Dummy in Dalaran. Unlike the regular Training Dummies in Stormwind, the Heroic one does in fact have armor, which makes it substantially more useful for DPS testing.

I ran some tests with a single Dirk on it and 0 talents spent. My character sheet damage range was 222-225 my actual damage range for hits was 129-131. This puts the armor reduction at about ~58%, which actually is very close to the 11085 number on Vulajin's sheet.

Next, I decided to do some testing with the following build:
51/13/7

For my cycle, I kept HfB up at all times, refreshing it only when there was <4 seconds remaining on it. I used exclusively Envenom, only doing 4 CP or higher finishers, making sure to pool energy as reasonable. My SnD did drop once for a second or two, but given the duration of combat (~5 minutes I think), that should have a negligible effect on the numbers. I had no buffs whatsoever from other classes on me, no consumables, and no debuffs on the target other than Master Poisoner.

Here's what the damage looked like in recount:


My actual DPS for the test was 1626.9 DPS, while my predicted DPS via Vulajin's sheet (assuming I used it correctly, of course) was 2332.4 DPS.

Type	Predicted	Actual	%Err
White	409.32	383.95	6.6%
Mutilate	317.55	300.99	5.2%
Envenom	203.44	178.96	13.67%
Instant Poison	801.60	597.07	24.25%
Deadly Poison	244.37	152.92	159.7%

I'd consider White/Mutilate to be within the % error given the relatively small sample size, and it's not that surprising they're accurate given nothing has really changed for them mechanics-wise.

However, the Instant Poison has a fairly large error margin, and the Deadly Poison is off the charts.

For Instant Poison, I thought at first maybe the Envenom bonus proc chance wasn't actually implemented yet. However, upon further calculations it did appear that the sheet correctly predicted an approximately 45% IP proc rate when taking into account the Envenom bonus chance uptime. However, while scrolling through my combat log I came across something I found interesting. Instant Poison can be partially resisted. Now unless I'm completely missing it (which is certianly possible), it does not appear this is currently modeled in the WotLK sheet.

This brings up an interesting question though...what is the average boss resistance percent, and thus how much do we reduce the IP damage by? Now if we go by the TBC model I believe once you hit cap (which I'm very close to getting to), you still will have approximately 5% of your damage lost to partial resistance. So that makes up some of the difference, but not all of the damage difference. Unless the global partial resistance amount has changed, I'm not sure where else the discrepancy comes from, but it does seem like your proc rate calcuations are correct. If it helps you guys out, here's the Recount IP breakdown:


And if you thought IP was off, Deadly Poison is incredibly off. I think the problem here is the uptime modeling, which says I should have a 99% uptime with 5 stacks of DP. In pretty much every fight I've been in anecdotaly so far with 1/2 Deadly Brew, I'm just not seeing nearly that high of an uptime. Most of the time I'm only hitting a 5 stack, if at all, right before the Envenom. It seems like most of the time I'm 3-4 stack, and once I even had my cycle breakdown where I couldn't Envenom because I had 0 DP stacks since the previous Envenom (and this was where my SnD dropped FWIW). I don't have time to test it right now, but if my suspicions are true, 2/2 Deadly Brew is far, far more valuable than at least I previously thought. Tomorrow I'll try running some tests with 2 points instead of 1.

Finally, Evenom is also off by a decent amount, I don't really have the time to analyze it further, but here is some more Recount detailed info:


Finally, my gear/stats/enchants. Note that most of this stuff isn't socketed/enchanted/optimally etc., yet due to a multitude of reasons (finding people to do some of this stuff is hard on Beta). Here's what I've got at the moment:
MH: Knife of Incision w/ Executioner
OH: Omen of Ruin w/ Executioner
Ranged: [Distracting Blades]
Head: Helm of the Vast Legions w/ Glyph of Ferocity, currently no gems
Neck: Chain of the Tolling Bell - Chain of the Tolling Bell
Shoulder: [Shoulderpads of the Silvermoon Retainer] w/ Inscription of the Blade (yes not greater)
Back: Embrace of Sorrow - No enchant
Chest: Skarvald's Dragonskin Habergeon - No enchant
Wrist: Shackles of Dark Whispers - No enchant
Hand: [Grips of Deftness] w/ Assault - yeah I know they suck
Waist: Ley-Whelphide Belt w/ Perfect Accurate Huge Citrine (7 hit 7 expertise)
Leg: [Trousers of the Scryers' Retainer] - No enchant (don't ask), Rigid Sun Crystal (12 hit), Balanced Shadow Crystal (9 sta 12 ap)
Foot: [Nyn'jah's Tabi Boots] w/ Cat's Swiftness, Shifting Tanzanite, Bright Crimson Spinel
Ring: Band of the Kirin Tor
Ring: Kurzel's Rage
Trinket: [Shard of Contempt]
Trinket: [Romulo's Poison Vial]

Sucks that WoW Armory doesn't work for Beta. Anyways, other relevant information is I'm an NE, and my professions are Skinning and Mining, so I have +25 crit rating over what I'd normally have.

Let me know if there's any further info that might help.

Just a note, Deadly Poison can also be partially resisted, on a per tick basis, and fully resisted on application:
WWS log

Also, did the improved poison talent nerf made it on beta yet (+3%, only for instant / deadly per talent) ?

This is certainly true. However, since there seems to be no other raid buffs active (and Rogues no matter what utility they've received are still very much a buffee class) for these tests, it looks like a baseline is being established, for the basis of future comparison. As such, no outside buffs should be used in the forming of the control point.

Once this control has been created, applying CoE is simply multiplying the DPS provided by Poisons by the factor of CoE (1.13). Any other way is chaos.

From the standpoint of Resistance, I was under the impression that virtually all non-immune bosses had very little resistance, and thus the CoE debuffing of resistance values would have no effect on partial resists.

I was just pointing out that there will be buffs increasing poison damage available. Folks on the beta can answer the question about boss resistences, but I expect since CoE has been updated we will see more nature resistence on bosses going forward.

There are exactly two bosses in the whole BC that have any resistance at all.
Also, CoE may get phased out by their DK/Moonkin equivalents which do not remove resistances.

This makes it even less likely for bosses to have any meaningful resistance since it would force you to:
A) Bring a warlock specifically for CoE or
B) Tell your casters to respecc to a non-resistant school
C) Kick them for more melee.


All kind of magic damage loses about ~6% of it's damage to partial resists on average vs. mobs 3 levels above you.
(This cannot be overcome in any possible way.)
This has nothing to do with mob resistance and we hopefully won't see any of it in WotLK.

Maybe somebody thought about that too? And published the results of what talents are worth it! Something crazy like on the second last page perhaps!

No, that wouldn't be possible, would it.

I doubt anyone (even developers) can give you this sort of info. Atm mutilate builds own but poison damage is way off the record, while combat builds have not-working combat potency and stuff. Just it is too early. It is not a bad idea to stock on fast daggers, from GB for example just in case, there are the badge ones too...

Honestly it's a little too early to start planning 3.0 specs given that we're likely going to see at least one more significant talent pass prior to release. However, if talents were as is as (with the assumption Combat Potency were fixed) Combat Swords you'd end up going 5/51/5 at 71 most likely.

Interesting, I believe full resists however will go away once you're at the spell hit cap. For this control case in particular I have 303 hit rating w/ Precision which leaves me with like 0.47% chance to fully resist my poison attempts. I wouldn't be surprised if the partial resistance damage reduction ends up being the same as IP, which at least by TBC standards would be 5-6%?

Yes. While it still has not been updated on the actual talent tree, the tooltip was updated to reflect it, and additionally testing has confirmed this.

Right, unless the Heroic Training Dummy for some reason has higher than normal NR (no reason to believe that but it's always possible), there's no reason to believe that CoE's -resistance will have any effect on the rate/amount of partial resists. As you said, generally modeling buffs is extremely easy as they just add to stats, or are simple multipliers on damage. The far more difficult problem is developing the combat model and then empirically testing the veracity of that model. However, with this addition of the Heroic Training Dummies, that part of the test has gotten a lot easier...assuming it does turn out to be the valid test I do believe it to be.

If they have Nature resistance -- or worse, Poison Immune bosses I will be sad. Hopefully they've learned from their mistakes (and it seems like they have, given that Sartharion is the first black dragon who isn't particularly resistant to Fire).

This is what I've heard for TBC mechanics as well. I'm curious if this number has changed at all in WotLK, as several other caster mechanics seem to have changed (like the amount of +hit% you guys need, if I'm not mistaken).

The spell miss rate is the same (17%), but they got rid of the last 1% that couldn't be overcome. This is on +3lvl.

Misery or FF will you give you +3% hit.

Using rank 8 Rupture my tool tip shows 2013 Damage over 16 seconds.

Here are Results without glyph:

8x264-265= 2117

With Rupture glyph:

10x 264-265 = 2646

If you want any thing else let me know.

I'm having problems getting the log up on WWS so I can't say definitively, but it sure didn't look like PotW was affecting poison damage. I do know it's no longer affecting the Scryer proc on the Shattered Sun Pendant.

Just throwing this out there since I've finally been able to play on the beta and I currently specced 54/10/7 picking up MP and TtT as opposed to putting the last 3 points in CQC. I don't know why I did it, it's just beta. Anywho, TtT seemed to be proccing in solo play and I was wondering if anyone had mentioned that yet. I was always under the impression that one had to be in a party or raid for a member of said group to activate the talent.

Turn the Tables does work solo.

As for Master Poisoner, I'm actually curious whether or not this talent works. I haven't had a large sample set so it's hard to say, but at least Blood Poisoning seemed to have a separate debuff it applied. I've never seen that for Master Poisoner however. Unfortunately I haven't ran enough tests to empirically know, but that's something I can possibly test out tonight as well.

I fixed some bugs in my calculations today, which brings the DPS for a full-Sunwell rogue at 80 down to 5k - which is still high-end, but a bit more reasonable.

I also ran some numbers on Combat Shiv builds; the first challenge of such a build is figuring out what cycle one should be running. I penciled in an Xs5r5n build, which seems to perform reasonably - although still 1000 DPS behind Mutilate. I get about 5700 DPS for a full-Naxx25 rogue. That might be improvable via better cycle choice, however.

Stat weights for both Shiv and Mutilate builds still strongly favor AP; only Expertise Rating crests 1.5 EP, and even it falls short of 2. AP is pretty clearly the best benefit-for-item budget stat, though that will admittedly change as the quality of gear improves. In terms of the other stats - Crit is a little bit better for Mutilate, and Hit is a little better for Shiv, which is more or less what you'd expect.

The other thing of note stat-wise is as that ArPen Rating is a horrifically bad stat, a fact which owes somewhat to the fact that poisons account for a large portion of the damage of both builds, but I think is also symptomatic of a larger issue, namely: the change to ArPen is actually a pretty strong nerf, and while this does prevent ArPen gem stacking from taking over, it's also disappointing to see a formerly good stat rendered so utterly useless.

13 energy per second (10 base regen, 3 from Combat Potency) as a baseline:

1s/5r with both SND and Rupture Glyphs is an 18s cycle with a 20s rupture - this obviously has some flaws in it, but it's pretty damn close to a very compact cycle.

With a 1.5 speed offhand, Shiv costs 35-3(Average Potency Procs off Shiv) = 32 energy ... 32 x 6 = 192 Shiv Energy + (.8 * 25) Finisher Energy = 212 total energy, with average energy generation during that time being 234 energy in 18 seconds. If you dump the Rupture glyph, you can do 1s/5r cut for a 17 second cycle with 94% rupture uptime (221 energy regenerated on average).

That would be more appropriate for a 7/51/13 build - the question is, does that level of Rupture uptime actually come out as a good thing? What would it take to make such a build actually viable - perhaps adding shiv to Agression?

Keep in mind that with the Shiv glyph, the cycle can be even shorter than that. Getting 100% Rupture Uptime isn't even the problem with this sort of build; the question is what you do with all the other combo points. I mean, 1s5r3n has 100% SnD uptime and actually clips the Rupture, which is why it got extended out to 3s5r5n for my estimates. The shorter cycle might be better, but I do think you want to think in terms of what you should do above and beyond 100% SnD/Rupture uptime.

In terms of the actual spec, the best I've found so far is 23/41/5+2, which is basically a testament to just how powerful the poison talents really are. The fact that Vile Poisons is good doesn't really surprise me; the fact that Improved Poisons to increase your white poison procs trumps all other available options was a bit more surprising to me.

Damnit, all that and I totally missed the idea that made it worthwhile - yeah, -5 energy on Shiv means you HAVE to go to a 3 finisher cycle.

In terms of poison damage, I think our DPS regulation (and everything coming down the pipes sounds like a nerf to rogues, but fairly in this case) is going to come from that angle. I think poisons are pretty overtuned at this point and to assume they'll stay the same is a bad idea. That being said, poison damage is pretty significant in a shiv build regardless, due to the IP on the offhand going nuts.

How does haste stack?