I just think Blizzard simply isn't interested in giving rogues significantly more utility-like options without nerfing damage output. They know the rogue community, and they know they'll squeeze every bug/cheat/exploit whatever you wanna call it out of any spec (see weapon swapping) for 3 more dps...or 600. While I believe they wanna see us do top dps in PvE, they're not gonna let us become too absurd in PvP, if we're not already there.

Mavanas, is it possible to rethink about HaT again with Patch 3.3 and or would it be a waste of time? I was thinking about using 8/18/45 with the Tier 10 set.

Nothing is changing in 3.3 that increases the viability of HaT relative to any other spec.

That's not entirely true. This bonus: Gives your melee finishing moves a 13% chance to add 3 combo points to your target, will favor a spec with higher frequency of finishers. HaT has a natural tendency to use finishers more often because of the extra income of combo points, even after the nerf. Even though my gut feeling says that 13% chance is not enough to fix anything for HAT, when I get some free time, I will put that into my last working version of HAT simsheet. Mutilate is also another spec which has a higher number of finishers used, and I doubt HaT will get anywhere close to it in terms of dps despite the set bonus.

HaT now also proccs from poisons. Should this be known already just ignore me (didn't find anything about that, but maybe I should check for more than 20 seconds)
So does the pyro rocket, engineering grenades and DMC: D

I'm wondering if there are any rumors about HAT resurfacing as PVE viable for Cataclysm down the road..

I have to disagree with you here. Sub has more or less been an entirely dead tree for some time now. A Sub build really isn't even good in PvP, let alone PvE. I wouldn't be surprised in the slightest to see a large overhaul of the Sub tree in Cataclysm; there may not be a PvE spec built around HaT, but there may be a viable PvE spec there in some capacity come the next xpac.

Honestly, with the focus of moving away from passive stat talents that Blizz has stated they plan on doing in Cataclysm, I wouldn't be surprised to see the Rogue trees shaken up a good bit. Rogues really don't have much at all in the way of choices to make when it comes to speccing their toon. You can either pick up one of the great talents, or one of the really bad talents. There are very few places in our trees where we have to make a choice between two good or even interesting talents. Just because Rogue damage output has been pretty well balanced doesn't in any way mean that our talent trees are even approaching what Blizzard's stated game plan is for the expansion. I, for one, am excited about the chance to actually make talent choices based on playstyle and circumstances beyond going cookie-cutter Combat if I have to move or cookie-cutter Mutilate if I don't.

To be honest, the reason rogue damage is so well balanced is because of talents which are balanced on a razor's edge. I see what you mean about subtlety being largely ignored these days, but we have 2 dps specs which are closer to each other than the vast majority of other dps specs (ele/enh, arms/fury, arcane/fire, aff/destro, etc). It's not the fact that rogue damage is well balanced that suggests rogue trees are in line with blizzard goals, it's the fact that blizz has stated that the rogue trees are in line with their goals. GC has frequently alluded to the state of rogue trees when talking about things they have gotten right in wotlk. He has also frequently mentioned that they want to fix subtlety, but I still maintain my opinion that subtlety will not receive a huge overhaul to make it raid viable because they already have other issues to work on. Blizzard always has to pick and choose what they can address; with 2 raid-viable trees almost identical in dps and being chosen for flexibility or specific fights, rogues are in a good place.

I don't think the subtlety tree will be identical when cat launches (particularly because it can't be, since they're redoing all of the trees). And no doubt they will have made some changes to make it more viable as a dps spec - and hopefully as an arena spec as well. But I think it will still be the red-headed stepchild.

This is obviously all augury and prognostication at this point.

Okay, let's stop it with the mindless speculation. There has been no official announcements of what they're doing with subtlety in Cataclysm. It will presumably receive some changes, as they've more or less stated that almost all trees should expect some changes. But trying to comment on whether it will be viable (or not) is a complete waste of time, as we have absolutely no information. If all you're going to do is blindly speculate, we don't need to hear it. So let's just drop it already. Thanks.

I copied my character over to the test realm to mess around with Subtlety. I am not versed in it or the preferred spread of the remaining non-sub branch points, so I went 10/10/51 backstab. My approach was to weave in ghostly strikes off cooldown, and an occasional hemo when time/energy permitted. I didn't settle into anything that felt natural though and didn't have time to experiment. After about 20 minutes of target dummy testing I was sitting at about 5k dps. My rotation was pretty rough, so I expect someone who knows what to do with all the cooldowns would sit a bit higher.

I realize Mavanas may have an "official" answer for this soon, but until then, are there any guesses on what the best expenditure of those remaining 20 points would be? Does anyone know if the new talents require a MH dagger (when they say "has a dagger equipped"), or would an offhand count? If so, perhaps fist/dagger with an SS or Hemo combo point builder would work better...

For reference (only), my Mutilate dummy numbers are just over 6k.

Since you have your character there already, you should actually test if putting a dagger in offhand allows you to do 160% weapon damage with a hemo and 180% with ghostly. You can use a dagger offhand first, check average hemo damage, then put something else in your offhand and try again.

If 160% damage only applies when dagger is in MH, then my findings so far indicate that slow/fast with hemo does more damage that dagger/fast with hemo or dagger/fast with backstab. I tried 10/10/51, but I also played with puncturing wounds instead of precision, and tried putting precision into imp snd. The reason I haven't tried every combination carefully is because they all seem to fall behind mutilate by at least 2k dps close to BiS gear.

What I am not yet able to analyze is the effect of weaving in ghostly strike and I have no way to model shadow dance for now. Especially if slow/fast with hemo turns out to be the best way to go, shadow dance will involve a swap to dagger for ambushes, which is a lot of work in the simsheet to model. I guess I'll try to add some napkin math to sim results when I have time.

Also don't forget that part of subtlety viability may depend on two rogues trading tricks with filthy tricks. There is a lot of value there. I haven't taken that into account yet.

I assume the spec would include 2/2 Blood Spatter, Hemo usage to build CPs. High rupture and garrote usage... Cheaper hemo and the very high benefit from both 2p and 4p t10 may in fact allow subtlety to creep up to being "viable", or at least viable once raid dps increase from hemo/frequent tricks are taken into account, although I'm not holding my breath yet.

Is there any interaction between energy reduction from Filthy tricks and the 4pt10? From the wording of the set bonus, I'd assume not, but it should be fairly easy to test.

Professor Hurt: There's no reason to use poisons other than IP/deadly since the interaction between full-stack deadly and MH poison increases IP proc chance to where it's more effective than wound in all circumstances.

I've noticed most people are doing 10/10/51 for testing Subtlety, but I was just curious if someone who's setup on the PTR had considered giving something along the lines of 13/7/51 a shot instead. I suppose with the effects of SftS reducing Backstab's energy cost Combo Points are really not a problem, but it's worth investigating. As for the 10/10/51 I can only imagine this being the spec. I was also wondering if taking the 3 points that I assumed were thrown into Precision and dumping them into Lethality would result in a dps increase. This is of course just speculation but if anyone get's around to simming\testing it, it's worth consideration.

I hesitate to participate in this topic for fear of embarrassing myself for even mentioning HaT, but I've been obsessed with it all day and since there's no current spreadsheet...

What I've been thinking is that:
1) 2 piece T10 + Filthy Tricks + Tricks glyph = the 50% uptime of tricks on your best friend (or worst enemy if you're trying to get him killed) but also a 15 energy return to you each time.
2) I've been assuming backstab is the way to go, but with hemo that extra 15 energy would mean more raid buffage.
3) With the (possible) incoming Rupture change, Serrated Blades + Blood Spatter (+Mangle/Trauma) will tick for stupid amounts.

Assuming you and your tricks target(s) do similar dps, is the personal dps loss from HaT currently <=> than 7.5%, given the above factors? At what inflection point does it make sense for the worst geared rogue to start tricksing the snot out of the best geared rogue(s)?

Although it's really hard to test this spec without a raid, I've been playing around with some different variants and "rotations" (if you can even call them that) are by far the most complicated to master that I've tried. The erratic combo point generation is hard enough; optimizing both Prep/Vanish and Shadow Dance is just nutty.

My analysis is not over, but I will say some things I've found out already:

• The only way HAT dps COULD be competitive is with tott glyph, filthy tricks, tott exchange every cooldown. I am saying could because best current setup I've found is still about 1k dps below mutilate. But I am not done optimizing.
• Dagger/dagger setup using hemo seems to be pulling a head. Most likely the winner is going to be using backstab glyph and extending the life of each rupture with 3 backstabs, but main combo point builder is hemorrhage.
• Rupture is in the rotation and shadow step should be used only with rupture.
• Arpen is better than agility, but hard cap is not reachable since you also need expertise cap and spell hit cap for max dps.
• So far 10/10/51 with improved snd seems like the best HAT can offer.

The things I am still working on is estimating the impact of shadow dance and possibility of glyphing for it; evaluating the value of hemo debuff for raid dps; checking if 2/3 HAT is enough for 1 combo/sec income.

That seems really wierd to me - We spec into SR and Blood Spatter, include Rupture in the rotation, want to use Shadowstep to maximize Rupture damage but Gem and Gear for something that doesn't increase Rupture damage?
Also Sinister Calling and Deadliness are huge for scaling with Agi (guessing they double-dip?) so how come ArPen still outdoes Agi?

I'm no math genious but that just seems absolutely awkward to me.

Also Hemo increases the damage of each physical hit / ability by 105 with the glyph - With 10 energy / second, you'd be using a Hemo every ~5 seconds including finishers (?) - if all charges (10) get used up (should be the case in a 25 man I assume) that'd be 1050 additional damage every ~5 seconds, resulting in 210 additional raid dps (everyone will realize that this is very rough napkin math)

Considering the above, do you still think it's worth using Backstab?

If each backstab (with glyph) extends Rupture by 2 seconds, figure out how much your Rupture ticks for on average and add that value to your backstab damage. So the inequality, assuming you only backstab 3x for each Rupture application, is:

(avg_backstab_dmg + avg_rupture_tick) / backstab_energy_cost <=> (avg_hemo_dmg + raid_hemo_benefit) / hemo_energy_cost

Don't forget that you have to factor in (or at least estimate) the raid's crit rate, and other bonuses, when computing hemo benefit. For example (I THINK...please correct me if I'm wrong) if you Ambush a target with hemo debuff you actually get 169 * 2.75 bonus, and it's probably a crit on top of that.

How does my napkin look?

EDIT: Changed 105 to 169.

EDIT AGAIN: Looks like this discussion is happening in parallel in two threads. The 20 second shadowstep/rupture cycle is really interesting, but definitely need a spreadsheet to model all the variables here. At least my small brain does.

I am well aware of multipliers of subtlety spec. I have been modeling it since last year, and I modeled it back when we had t8 bonus. Just like then, armor pen pulls ahead of agility and AP despite all the multipliers. The reason for that is 75% of your damage is affected by armor penetration: melee, eviscerate, and hemo, and you do not have poison enhancing talents, such as imp poisons, vile poisons and even LR. With all the multipliers agility is not going to reach 4 dps/point (although it may reach 3 dps/point), but heavy physical damage specs, such as combat, have Arpen at 4 dps/point.

What I said there was that the main combo point builder is hemorrhage, and backstab is used only to extend the life of rupture. The glyph of backstab is worth about 170 dps. However, thinking more about it, if you use hemo, you are doing a hemo every 3.8 seconds (based on simsheet). To extend the life of a rupture, you need to do 3 backstabs every every 22 seconds. In those 22 seconds, you normally do 5.8 hemos, but now 4.5 of them are replaced by backstabs in terms of energy, so you are left with a mere 1.3 hemo per 22 seconds if you want to get the 170 personal dps benefit from backstab.

Ok so now I can fill the gaps in hemo and shadow dance math.
Hemo raid dps benefit: Per hemo = 788 damage, I assumed the charges are consumed by autoattacks, subject to glancing hit modifiers, average crit rate of 50%, crit multiplier 1.03, average armor pen of 400, and resulting damage reduction of 27.8%. Without backstabing, the frequency of hemo is 0.263, which results in 207 raid dps increase. With glyph of hemorrhage, the benefit increases to 290 dps. With backstabbing to extend rupture life, the hemo raid benefit is only 47 dps, so together with personal benefit of 170, you get 217. Based on this it's better to glyph hemorrhage and never use backstab at all. In that case, you also go slow/fast for max hemo damage.

As for shadow dance, you get an opportunity to use garrote and ambush, of them ambush has higher dpe because of all the armor pen, so I'll just assume ambush is used. I am also going to assume that ambush is talented, and the player can pull energy before shadow dance to fit 3 ambushes in 6 seconds. In terms of DPE analysis, you gain 22482 damage (after taking into account loss of hemo buff). From combo-point perspective, with full initiative and 1 premeditation, you gain 7 combo points + 6 more from HAT, for a total of 13 combo in the 6 seconds. Without shadow dance, for same energy you get 4.5+6=10.5 combo points, so the gain is 2.5 combo points. 15 combo points allow you to spend 30 energy and do 29346 damage worth of eviscerate, instead of 5772 you can get from a 30-energy hemorrhage. So the benefit of 2.5 combo points is 3929 (note this is pretty simplistic as it does not take into account poisons and 4pt10 bonus among other things). Anyway, the total benefit of a shadow dance based on this napkin math comes down to 26411 damage every 60 seconds, or 440 dps. Note however that in order to ambush, you need to swap in a MH dagger twice every minute. Due to swing reset, you are losing 5231 weapon damage and 3052 instant poison damage. So the benefit goes down to 302 dps (in fact the loss from weapon swapping exceeds the gain from using slow main hand of 111 dps based on simsheet).

Final conclusions:
If you want to enjoy the full benefit of shadow dance, due to weapon swapping losses, it becomes optimal to stick with dagger/dagger setup and use hemo all the time except for the shadow dance. Hemo rotation with armor penetration setup gives you dps of 13148. My napkin math shows you can gain another 440 from shadow dance, and 290 raid dps from hemo debuff, so your total raid dps becomes 13878. This number takes into account the damage gain from trading tricks with another rogue every cooldown. It assumes tott, hemo and eviscerate glyphs.

Now compare that to mutilate dps of 13855 without the benefit of tricks in 3.3.2. So with tricks, in 3.3.3 mutilate will have at least 14548. Which means subtlety is a raid dps loss of 700 dps if you trade tricks with another rogue. In order to cover for 700 dps loss, you would need to be tricking someone who has 9333 dps more than you, i.e. someone around 20k dps.

Link to 3.3.3a shimsheet.

Thanks for the Info,

Just a quick question for you Mavanas, did your sim math for Sub include the extra 2 talent points the default Simsheet has? (its a 10/10/53).

10/10/51 seems the most viable so far but I am playing between two of them. 5/5 Precision vs Imp SnD + 3/5 Precision.

5/5 Precision
Glyphs - Rupture, TotT, Eviscerate
Rotation - 5s-5e-ShS+5r-5e

SnD + 3/5 Precision
Glyphs - Rupture, Backstab, TotT
Rotation - 5s-5e-5e-ShS+5r-5e-5e

Not a fan of the latter as timing with Shadowstep and Rupture are off. (20 sec CD with 26 sec Rupture.) Not to mention Backstabbing only 3 times per Rupture cycle seems weird.

ArP does seem to be better with more Evis weaving.

I need to play with it more.

Edit: Wrong Glyph. Wrong Talents.
(This was submitted before Mavanas's above post.)

Mavanas, do you think using Vanish+Preparation to get the extra CPs with Premeditation is worthwhile? Or is that already included in your analysis?

I added two points in imp ambush when I was doing the napkin math on Shadow Dance. You can take those two points from setup, which is largely useless (only helps when you have a chance to dodge or resist something).

I also realized two more things: Shadow Dance benefit that I estimated at 440 dps has to be multiplied by an additional 1.075 for tott if you cannot time shadow dance and incoming tricks. So you get additional 33 dps form that. If you can time shadow dance to occur during incoming tricks, then you get 66 additional dps. Given that tricks last 10 seconds, as soon as you see incoming tricks, you can pull energy for 3 seconds and then blow shadow dance with at least three ambushes and at least 2 five-point eviscerates. So the new estimate is 13944.

Also in my previous analysis I assumed MOS is not triggered during shadow dance. If you get MOS benefit during shadow dance and 6 seconds after it, there needs to be another modifier.

P.S. Simsheet assumes when you Vanish, you use premeditation if it's up and then garrote (with arpen I should actually change it to ambush). Preparation is used after the second garrote debuff runs out to allow you to do another vanish.