Given that HaT is basically terrible now that pet crits don't proc CPs, I don't think the change to Shadow Dance makes the spec any more desirable.

What will the point of Norm 25man if it drops same ilvl as 10man hard mode?

10m Hard requires talented players, 25m Regular requires a lot of players.


Here's a link to the T9 set (highest iLVL).

Unless you're sporting both shoulders and helm, you won't be over the cap. You can easily achieve the 4pc bonus without being over the Expertise cap. Even if luck is poor for you, it is better to wear Helm + Shoulders and have both points into Weapon Expertise. You're still gaining expertise from these points (unless you have 214+ from gear alone, which is a mistake in itself), even if a portion of the expertise on gear is wasted. Your alternative is wasting many more ilvl points on expertise on gear, to pick up a talent that might help pad your trash meters.

HaT is hardly terrible. With 1CPS HaT pulls 7k+ sustained dps, is better at target switching, works very well with 4pct8, and has the added utility of cheat death, shadowstep and prep. Combat and mutilate rogues contribute more than .3CPS each in recent parses.

I would be very interested in seeing how a backstab/shadow dance spec would perform.

Ok, maybe terrible is a strong word. It's competitive in ideal circumstances, yes. But part of the problem with adding Shadowstep to the build is you lose quite a few sustained DPS talents from Combat in the process. I'm not sure it's worth it.

As for a Backstab build -- I'd have to imagine the best route for that still would be Combat. You just miss a lot of talents if you go that deep in Sub but still intend to use Backstab as your CP generator.

Isn't that the whole point of HAT, though? HAT is your CP generator. 45e Backstabs hit hard as hell and if you don't have to rely on them for your CP generation, they probably become really good "filler" between all the Eviscerates.

Also, how do you lose a lot of sustained DPS talents by picking up one (or two, even)? If you're going Backstab subtlety, you're likely going to Slaughter in the Shadows anyway, so you might as well grab Shadowstep along the way, and Shadow Dance afterwards.

I'm not saying the build will work -- just not sure I follow why you're saying what you're saying above. It might be fun to play around with it in Naxx sometime just to see how it fares before investing any real time into number-crunching or whatever.

This. One thing you notice looking at SWP gear versus, say, Tier 6 and below, is that the T6 stuff still follows the model of having AGI/AP + 2 stats to fill out an item. Most of our stuff is Hit/Crit, Crit/Haste, Hit/ArPen, and the T9 set is throwing in a few variants, but they all follow that same model.

The Sunwell Plateau Slayer's items were _so_ good (take a look at [Slayer's Boots]) in large part because they moved up to 3 stats + Agi/AP to avoid dim. returns on stat increases, and I think we might see that in Icecrown. One large benefit of this is it's easier to add expertise in more granular chunks since you can get 4 items with ~40 expertise each instead of 2 with ~80, plus you get a lot more bang for your iLvl "buck" allocating 3 stats instead of 2. Easy to make very attractive Icecrown upgades that way (plus they'll probably swing back towards lower STA allocation on T10 + more red sockets ala Sunwell) especially relative to what we're seeing of tier 9.

Doesn't really solve the problem of overshooting the expertise cap using 5pc T9, but aren't we all pretty used to using 4/5 with an optimal offset piece anyway?

So, we now know the stats on the new professions - lets take a look at how they stack up.

Alchemy: 80 AP.
Blacsksmithing: 2 epic gems, at 40 AP or 20 of any other stat, meaning your choice of 80 AP or 40 of stat-of-choice (usually Agi or ArPen).
Enchanting: 80 AP.
Herbalism: More healing, handy for questing and soloing stuff, I suppose, but useless in raids.
Inscription: 80 AP
JC: New Dragon Eyes are 68 AP/34 stat of choice, an upgrade of 28/14 over epic gems; thus, the total benefit is 84 AP and/or 42 of stat of choice.
Leatherworking: 80 AP
Mining: 60 sta
Skinning: 40 crit rating

So, among the easily-valued professions, we have a pretty obvious hierarchy:

1. JC: 84, or more if there's a stat with EP > 2.
2. BS: 80, or more if there's a stat with EP > 2.
T3. Enchanting: 80
T3. Inscription: 80
T3. Leatherworking: 80
T3. Alchemy: 80

Okay, so what about the two that are a bit harder to assess? Where do they fit in?

Well, Tailoring gives Swordguard Embroidery, which has a 45 second internal cooldown and a 25% proc rate; thus, un average, it's going to proc about once every 47 seconds, giving it uptime right around 32%; thus, the average-case benefit is 400 * .32 = 128 AP or so. Compare this to the best available alternative, which would be 22 agi for most of us, which has a value right around 45 EP. So the overall benefit is right around 83 EP - give or take. In practice, it might be slightly higher than this, as the proc will tend to align pretty well with your trinket procs, leading to some degree of synergy. Note, however, that as the value of agility increases (and it probably will, with time), this margin will be eroded somewhat.

Engineering, on the other hand, gives a multitude of benefits which we're going to have to add up:

1) Hyperspeed Accelerators now give 20% uptime on 340 haste rating, or 68 average-case haste; this works out to a benefit around 91 EP for Mutilate and 101 for combat; compared to the baseline glove enchant which is 44 EP for both, the benefit is roughly 47 EP for Mutilate and 57 EP for combat.

2) Flexweave Underlay is 1 agi better than the regular cloak enchant. 2 EP for both.

3) Nitro boosts are now identical to Icewalker, with the added benefit of boost movement at need; in an ideal fight, where they can be used on every cooldown, these reduce transit time by 2.5 seconds once every 3 minutes; 1 second of DPS in in the neighborhood of 7000 damage for most of us, so we're looking at something like 17500 bonus damage every 180 seconds, which is a full 100 DPS benefit, right?

Well, no. Because in practice the number of fights on which you need to to move far enough to get the full 2.5 second benefit, and need to do so more often than you can with sprint, are few and far between. The only fight in Ulduar where I can see getting full benefit from them is Yogg phase 2; one can argue you receive partial benefit on Mim, XT Hard, and some modes of Iron Council, but it's definitely the exception and not the rule. And even in those cases, it's only during certain phases, so in reality you'll use them once a fight at most in the majority of cases. So a more realistic assessment is, perhaps, 2 seconds of movement saved across a 10 minute fight, which works out to more like 20 DPS, which is around 25 EP. And on most fights, it will be less than that. So we're looking at, average case, perhaps 10 EP, I would argue.

Finally, we have bombs. A bomb does 1325 base damage, increased 13% by raid buffs, reduced by 6% due to resists, and increased by around 15% due to spell crit. Thus, we estimate the average damage in reality is around 1600; with a 60 sec cooldown, this works out to around 27 DPS, which is somewhere in the neighborhood of 30 EP if used on cooldown.

Thus, the total theoretical benefit of engineering is around 47 + 2 + 10 + 30 = 89 for Mutilate, and 99 for Combat... assuming you use everything on cooldown. If, however, you are consistently 5 seconds late on using Hyperspeed Accelerators, or don't feel like burning through a hundred bombs a night, or otherwise fail to use the profession to it's absolute fullest, it's value is somewhat lower. For instance, if you use bombs and accelerators every 65 seconds instead of every 60, the total benefit drops to 80/90; and every fight you skip using bombs entirely on drops engineering to the bottom of the heap.

So, in terms of optimal professions, what do we find? Well, if you're absolutely commited to using engineering, and use every single ability on cooldown on every fight - including going through very large numbers of bombs and putting up with a lot of extra cooldowns to keep track of - it's going to be the best profession, and thus the optimal pairing is JC/Eng.

However, for those people who aren't willing or able to put up with that level of crap on engineering, it's a weak profession; at that point, JC is the best, and either Tailoring or Blacksmithing will be second. In particular, if there's a stat who's EP value is 2.1 or higher, Blacksmithing is likely best; if not, Tailoring is probably the way to go. However, none of these professions have a significant advantage over LW, Enchanting, Inscription, or Alchemy, so I suspect a lot of rogues will not find it worthwhile to swap professions for the 2 DPS difference between professions.

Edit: Assorted updates and corrections based on comments.

Popped a flask on ptr. It's a total of 260 ap, so 180 base and 80 bonus from Mixology. (0/0/0 so no additives from talents.)

Also, not sure why you listed the glove enchant as "~41 EP", as crusher gives 44 AP, and therefore should be exactly 44 EP (if you're not referring to Crusher, then whatever you're referring to is lower EP than crusher anyway unless it beats 44 EP).


Other than that, thanks for running the numbers.. definitely think I'm sticking JC/LW. Wish they'd make Skinning worth a damn, but I don't think I'm going back down that path.

Sorry, I didn't see the difference. A 51-point Subtlety build that doesn't have HAT is like a 51-point Combat build that doesn't have Combat Potency -- you could do it, but why the hell would you?