I'd need the full data set to do a full statistical analysis, but off the top of my head:

If for our null hypothesis we assume that the reduction is uniformly distributed between 65% and 85% and work out the standard deviation we'd expect to see on 256 glances, we get a variance of .36% - that is, we'd expect to observe an average glancing reduction between .243 and .257 about 95% of the time. Since we're in that range, the data is consistent with that probability distribution and mean. So 25% damage reduction seems to be in the right ballpark, anyway.

As for glancing rate... well, in 1002 attacks, we got 256 glances, which means that the true glancing rate is 95% likely to be between 22.8% and 25.8%. So, again, consistent with the 24% number we've been using, but insufficient to pin it down further.

All in all? I think we can safely say that glancing blows haven't changed too much, and while eventually we should pin it down a little better, for the moment we're fine using the same assumptions we have been.

More fascinating news: it seems like when they increased the Ambush weapon damage multiplier from 250% to 275%, they inadvertently increased the amount of static damage contributed to the ability as well. The Wowhead entries for Backstab and Ambush have the following pattern: they provide for "normalized weapon damage plus [some flat damage amount]", and "weapon damage % multiplier." For all ranks of Backstab, if you multiply the specified flat damage amount by the weapon damage multiplier, it checks out with the tooltip. However, for all ranks of Ambush, the same is only true if you multiply the flat damage amount by the old multiplier of 250%.

I confirmed this by logging in to a fresh copy of my rogue on beta and performing Ambush 7 several times on a Training Dummy. If my suspicions were correct, the expected damage on Ambush 7 with a [Dirk] and 1963 AP should have been 1029.50, or crit damage of 2120.77. These values matched what I experienced on actual Ambushes.

Conclusion: the tooltips for Ambush ranks 1 to 7 are all incorrect; the correct flat damage amount is 10% higher than what is listed.

(edit) Interesting corollary to this finding -- it's likely that Sinister Calling also slightly increases the flat damage provided by each of these abilities.

Okay, so with the information I know for sure filled in, and reasonable guesses plugged in for the rest, I put together some DPS estimates for a fully-raid buffed Mutilate rogue in the best known raiding gear (Mostly Naxx 25 stuff). The first thing to note is that the damage estimate for such a rogue is in the ballpark of 6k-6500, which is quite a bit. And there's definitely some oddities in terms of stat weightings and such, which makes me think that a number of things still have a very good chance of changing. But with those caveats in mind, lets press onwards.

The baseline build that I'm comparing to is this: 51/13/7. This may or may not be the best Mutilate build in existence, but it's certainly a reasonable one, so lets go with it for the moment. One of the major objectives of this undertaking was to answer the question of what is best.

For the moment, I'm assuming optimal cycles are either 4+pt Envenom spam, or 2-Mutilate Envenom spam, whichever has higher damage. Rupture might be worth weaving in, but I haven't had time to work out the numbers yet.

The way I'm going to approach this is to list the damage change by adding/subtracting points in any given talent, so we can figure out which talents are important and not. So, the list:

Malice - 164 DPS for 5 points, or 33 DPS per point.
Ruthlessness - 124 DPS for 3 points, or 41 DPS per point.
Puncturing Wounds - 118 DPS for 3 points, or 39 DPS per point.
Lethality - 108 DPS for 5 points, or 22 DPS per point.
Vile Poisons - 571 DPS for 3 points, or 190 DPS per point.
Improved Poisons - 841 DPS for 5 points, or 168 DPS per point
Quick Recovery - 19 DPS for 2 points, or 10 DPS per point
Seal Fate - 187 DPS for 5 points, or 37 DPS per point
Deadly Brew - I don't have exact estimates here, but it appears to be on the order of 1000 DPS for two points.
Focused Attacks - 232 DPS for 3 points, or 77 DPS per point.
Find Weakness - 82 DPS for 3 points, or 27 DPS per point.
Infectious Poisons - 230 DPS for 2 points, or 115 DPS per point
Turn the Tables - 47 DPS for 3 points, or 16 DPS per point
HFB - 325 DPS for 1 point

DW Spec - 387 DPS for 5 points, or 77 DPS per point
Precision - 300 DPS for 5 points, or 60 DPS per point
CQC - Same as Malice, 33 DPS per point

Relentless Strikes - 330 DPS for 5 points, or 66 DPS per point
Opportunity - 171 DPS for 2 points, or 85 DPS per point

So, what we gather from this is that the initially proposed talent build is not actually particularly good. In fact, this would appear to be the highest-DPS build (or perhaps the analagous build with 1/3 TtT and 3/3 Find Weakness in place of Murder, given the low density of Murderable mobs in Naxx). It also gives a pretty clear idea of which talents are over- or under-tuned. In particular, all the poison talents seem pretty strong right now, and Turn the Tables and Find Weakness notably underperform.

Most of the conclusions from the latest version of my sheet suggest the same things, and the final build you provided is very similar to what I've been working with. Worth noting regarding cycles: I found that using Glyph of Rupture (which, despite the tooltip, increases Rupture duration by 4 seconds) and Glyph of Slice and Dice made trading 2 points of Close Quarters Combat for Imp Slice and Dice and 2 points of Ruthlessness for Blood Spatter almost worthwhile. The DPS loss was something like 20 (about 4850 with Envenom-spam to about 4830 with Envenom/Rupture).

Another interesting note regards the Shiv spam builds which had been discussed several pages back. Although I can't state this as a specific answer to the builds discussed at that time, I can say that, in general, Shiv will earn a small marginal benefit in poison procs as compared against other combo moves. This is because with the increased effect of Improved Poisons and the poison proc rate buff from Envenom, most builds will have a decently high proc rate to begin with; Shiv won't be effectively granting you that much extra poison per use. Its actual gain will be more than offset by the big drop in energy damage as compared to using Sinister Strike or Mutilate.

I am running that build on beta currently, but with 3/3 find weakness and vigor. I know, not the cleverest combination but it was the best i could come up with and not having to crunch numbers, due to the fact that Infectionous Poisons can't be specced into currently. Clicking it has no effect on the talent window.

Awesome, thanks for your great work again.

What I'd love to know is the impact of weapon speed.
Daggers seem to come in 1.4s and 1.8s variants, from PvP and instances, if Daggers - Items - World of Warcraft is any indication.

Could you pick a fixed set of talents and gear and check the DPS for:
1.8s/1.8s, 1.8s/1.4s, 1.4s/1.8s, 1.4s/1.4s?
Having some quantitative values would be nice, whether absolutes or relatives.


As for specs - I wonder how fast 1/2 Deadly Brew gets 5-stacks on the target, compared to 2/2.
I'd guess that the relative poison damage increase from the second point isn't much, but the absolute DPS is likely still a lot higher than any of the other 1%-talents there.

Also, does your spreadsheet have a damage summary like "30% white, 45% IP, 10% DP, 10% Mut, 5% Env" or something?


Sorry for having so many questions at once, Rogues are just too interesting right now!

Master Poisoner is a raid debuff. You get it if you have no Ret or ToW (unclear whether it does/will affect melee crit) in the raid, otherwise you don't.

As for AP scaling - if you plug a level 70 rogue with blues, cheap enchants and a few Kara items and only a few raid buffs into a sheet, you'll see agility as a better stat than hit.
It takes some gear for hit to outscale agi/AP.

The same happens in Wrath, just with a higher mark.

Correct, but the chance you'll be bringing an elemental shaman or a retri paladin is high

Has anyone done testing regarding Tricks of the Trade? Specifically I'm interested in whether it uses the rogue's threat modifiers or the target's.

Perhaps that will be possible to test with blizzard's threat deal thing?

You don't really need a Ret Pally. Our Holy Pallies are likely gonna take Heart of the Crusader (3% raid crit) and will be judging every time they can cause of Judgement of the Pure (increase haste by 10% for 30 sec on judgements). Plus they're able to do it from ranged with Enlightened Judgements (20 yard increase). So just have to coordinate if your pallies are doing that.

I have a question to those who used Tricks of the Trade: did you have any problem with target selection? Did you have to switch raid or party leading during an encounter? I really like the idea behind this skill, but its usage sounds a bit weird to me. I'd prefer to choose directly its target, just like misdirection.

Afaik, it works exactly the same as misdirection. Hand reticule lights up if you have a invalid target and you can use /cast target=targettarget or target=tankname to be specific.

I'm going to take a wild one here and guess that the main damage difference is due to poisons.
Something similar happened when I swapped Shiv of Exsanguination with Fang of Kalecgos, with Fang showing quite a bit better dps.

Try unselecting poisons on the offhand and see if it's still the same difference.

At level 70, switching from a 2.6 sword to a 2.4 sword would net you approximately 20 DPS (with Combat Potency and Deadly Poison). So it must be the poisons, since you're actually losing a significant chunk of damage by dropping Sword Specialization.

Why is everyone skipping sealfate for fleet footed + quick recovery in what i presume to be PVE builds?

Is it completely out of the question to suggest MHing and OHing fast speed daggers for more energy generation for the current WotLK mutilate builds? Wouldn't speed increase the amount of poison and energy while maintaining the same white dmg?

Historically, SF hasn't been a notably good talent, so the early hypothesis was that it was one of the points most able to be cut without hindering DPS - and Fleet Footed, at least, does have significant PvE value. However, with the increased damage of finishers, the value of SF has been rising (although it's still mediocre) so it's start to creep back into more builds.

No, it's not. A fast OH is definitely worthwhile, and my preliminary estimates is that a fast MH is probably worth it as well. This assumes that poison stays at it's current DPS level (which would surprise me a bit), but as things stand, yes, fast MH/fast OH is potentially worthwhile.

It adds nothing if you assume that someone else in the raid is applying a debuff that doesn't stack with it (ret pala or ele shaman apparently).

I just noticed this on PTR, CttC only refreshes SnD, which probably isn't a issue, since we can start with Shiv->SnD->Mutilate->Mutilate(x2?)->Envenom.

I was pushing ~~1900 dps with Dual Bombarider's Blade (SSO Revered blue daggers).
32% Melee, 25% Instant Poison, 21% Mutilate, 10% Envenom and 10% Deadly Poison.
I was also refreshing HtB early (~~5seconds remaining)
Mutilate spec was: This due Infectious Poison being bugged.

So I went and respecced to combat to try and see how it compares to Mutilate. Spec was: This.
However, when I started off trying to maintain 5/5, I ran into a problem. SnD would run out before I had 5cp again.
Confused, I tried 5/4 and ran into a similar problem, only at 4/3 I was able to barely maintain it.

Which led me to believe that Combat Potency is somewhat bugged at the moment.

With 4/3, best I could maintain was 1420dps with Brutal Gladiator's Slicer/Quickblade.

Thinking that this was a issue of me lacking Relentless Strikes I specced: This.
Which resulted in 5/5 being maintainable, however the dps was lowered to 1350 average.
55% melee, 24.6% SS, 15.2% Instant Poison, 3.3% Rupture, 1.2% Sword Spec.


While I didn't do extensive testing, it clearly shows how out of whack poison talents currently are.

Edit: I went back and tested Mutilate again (this time using AtoL and SoTC) and managed to break ~2000dps unbuffed.

5s/5r is not going to be maintainable without Relentless Strikes. Relentless Strikes ends up regenerating an amount of energy within the same ballpark as Combat Potency.

Well I'm pretty sure that there's something wrong with Combat Potency as well as Focused Attacks at the moment.

Edit: Missed that Vulajin.

Except that Combat Potency and Focused Attacks are widely known and acknowledged in many places to not currently work in beta.