While I understand that haste effects wont gain from it, there are always ICDs and trinket on-uses to consider as well. Those were the specifics that I was referring to.

Allow me to save some time on speculation here: Recuperating will be a DPS gain. The speculation around "what if it's not" is sort of pointless, as it's almost impossible to imagine that it won't be. Consider: it gives 115 energy for 5 combo points. If nothing else, you can spend that 115 energy on 4 Hemos, which will restore 4 of the 5 combo points. Hence, worst case, it's a finisher that gives you the damage of 4 hemos for 1 combo point. And I pretty much guarantee that that will always be worthwhile. Alternately: the Combat spreadsheet right now attributes damage values to combo points and energy. In my current gear, a combo point is worth 1051 damage and a point of energy is worth 217.5 damage. So 5 combo points is 5255 damage, and 115 energy is 25012.5 damage, for a net increase of 19757.5, which works out to 659 DPS. That is: recuperating ever 30 seconds would increase your DPS by 660 relative to using those combo points on Eviscerate. To put that in perspective: it would be worth doing even on Anub'arak, as the damage you gained would more than offset the extra healing you give him via Leaching Swarm. Now, admittedly those numbers are for current combat and not 4.0 Subtlety... but still. It's hard to imagine that it won't be worth doing on cooldown.

And, frankly, this seems stupid to me. They've stated that Recuperate should be a reasonable tool in our arsenal for survivability in raids; thus it seems odd that one spec gets to have it up continually at all times as part of its regular DPS rotation. If it's enough healing to be every worth using for Combat/Assassination, isn't having it up 100% of the time as Subtlety going to be a little bit on the powerful side? And if 100% uptime on it isn't potentially a balance concern, how will it every be worth worrying about for the other specs?

Moreover, lets consider what it does to our rotation. Recuperate means we generate 14 energy per second; if we use this all on Backstab (which will almost certainly be optimal), we thus generate .35 combo points per second via attacks, and another .5 per second via HAT, for total CP generation of .85 per second. We then, ideally, want to keep SnD (5 CP per 21 seconds), Recuperate (5 per 30) and Rupture (5 per 20) up, which leaves precisely enough CP generation leftover to eviscerate... once every 25.6 seconds. Now, this is an approximation; since the Evis will occasionally refresh the Rupture we'll be able to keep Rupture up slightly more cheaply, and thus get a few more Eviscerates; on the other hand, this also assumes that combo points can be spent with 100% efficiency (which they can't), assumes HAT procs exactly every 2 seconds (which it won't) and ignores the energy cost of the finishers themselves. So realistically, with having to keep all this stuff up, we're only rarely going to be able to squeeze in an Eviscerate.

Using Hemo makes the situation a little better - it makes it up to about 1 eviscerate per rupture - but it still winds up that we're spending most of our CP on sustaining DoTs and self buffs and very little doing direct damage with our finishers... and given that that's supposed to be part of the shtick of the tree, that seems a bit off.

Haste will of course help the situation a bit in either case... but not very much. We're unlikely to have more than 10% haste anytime soon, and an extra .03 combo points per second just isn't going to make enough difference to change the fundamental conclusion.

In short: I think tying energy regen to a continuous self heal is a bad idea in the first place, and doubly so for Subtlety. It would make far more sense to tie Sub energy regen to Rupture (as is done in Assassination) and allow a reasonably straightforward 3-finisher rotation. This would require coming up with an alternate mechanism for energy regen in Assassination, but I'm sure there are other things they could use.

Just two weeks ago, GC said (here) that rogues will not be balanced around 100% uptime on Recuperation so I can't imagine they would make it viable for Subtlety to do so shortly thereafter; Blizzard changes their minds all the time but not so quickly. The regen numbers have to be a first pass just to see if it works; I have to believe that the purpose is only to offset part of the cost and not to provide meaningful regen making this more of a utility talent and not a true DPS talent. Hopefully they know this and will take Recuperation out of the rotation in the next build.

"Sadism: Every time you deal poison damage...."

EDIT: On second thought, the name "Sadism" is too cool; save it for the Sub talent. :-)

I can get behind a reboot of HAT so that it scales with gear but I do like the dependency on a group/raid dynamic that it brings and would hate to lose that. It's very unique in the game and if there's a way to make it scale with gear and keep that dynamic, I'd be all for it. Perhaps it's as simple as having Haste lower the internal cooldown (although I don't know how "simple" that is) or maybe balance it to be a proc chance that scales with paperdoll crit (although proc'ing on a crit of someone else based on your own crit rate gets complicated fast).

As for poison scaling in the Assassination tree, wouldn't 200% poison crits further devalue Mutilate and Backstab (and maybe even Rupture due to the Envenom proc) and push more value to Envenom and auto-attacking? Don't we want to get away from that or am I missing something?

In terms of assassination there are two issues, which are important to keep separated.

One is how passive/active the damage sources of the class are - this is the "skill" consideration. At the moment, Mutilate has a 2-finisher rotation with Rupture and Mutilate, and your rotation shifts dramatically at 35% when you switch from Mutilate to Backstab. Moreover, there are subtle optimizations that can be done in terms of Envenom timing to maximize buff uptime. So in this respect - particularly relative to Combat - I think Mutilate is doing fine. Oh, ideally it might be nice if Envenom timing mattered more than it does, but the fact remains that the cycle has a decent degree of complexity to it. So I'm not worrying too much about this one.

What I'm concerned about is the "scaling" consideration - that is, will the spec improve at the appropriate rate relative to other classes/specs. At some level the mastery mechanic makes this a bit less relevant than it was in previous expansion (as they can just change 20 to 25 rather than needing to invent new mechanics), but that's not a complete solution. It means that the class can't be balanced simultaneously for well geared players and poorly geared players - if the mastery is set such that top end players do the right amount of damage, undergeared people from non-cutting-edge guilds will be too strong, and if you balance things for those players, well-geared players at the cutting edge will be too weak. Moreover, it will also imply that gear matters less for us than for other classes - you can screw the rogue over on gear and he'll still do fine on damage once his next periodic buff comes through. This is not ideal.

So: Mutilate scales poorly with weapon damage (as only half of it's damage - if that - comes from weapons), poorly with crit (as 30-40% of its damage comes via poisons, which scale poorly if at all with crit), not as well as usual with agility (as the crit component of it does less for assassination than for the other specs), and so forth. So there is a "scaling problem". But note that it's not limited to "our weapons matter less" - it's "our weapons matter less, and a couple other stats stink as well". One solution would be to make weapons matter more - but I think that robs the tree of some of it's identity. I think the idea of a rogue spec that does half nature damage is an interesting one. Hence, I think the better solution is to allow it to continue to do that, but allow the poison damage to scale as well as it needs to for the class to keep pace. It will mean that weapons - particularly OH weapons - matter less for Mutilate than they do for Combat... but that's not necessarily a bad thing. Again: different doesn't mean bad.

Also note that increasing the crit multiplier on poison doesn't necessarily increase the amount of damage that comes from poison - they could reduce the AP coefficient by a commensurate amount, such that total poison damage, AP scaling, haste scaling, etc. all remain the same, but crit scaling improves. And that would make the class scale strictly better than it does right now.

HaT procs based on autoattacks in some way would probably also require an ICD (with the potential scaling softcaps that can imply), or else burst could get out of hand. A lucky string of attack-based CP might be awkward in PvP.

There are two substantive differences I can think of between HaT procs on Rupture ticks as opposed to haste-shortened ICD for raid-crit HaT. First, it requires Rupture to be up on a mob, which will increase ramp-up time at the start of a fight and seriously hurt CP generation on short-lived mobs. This is pretty clearly a nerf, as one of the strengths of HaT is it's ability to provide CP on your target when you might be out of range, or so forth. Being able to toss free Deadly Throws while a boss is out of range is a very nice (possibly unintended) feature of HaT, as well as the easy that you'll have in putting up an Expose Armor or refreshing SnD when switching mobs.

Second, it will allow a Subtlety rogue to generate the same amount of CP without a group as with one, which breaks the original intent of HaT which was to provide sub with a larger PvE damage boost than it does a PvP damage boost. It could be the case that rupture is rarely used in PvP and thus this is irrelevant, and I don't know if the "better with a crowd" design goal is one worth preserving.

The two methods would have approximately the same CP generation on a fight where you could keep near-perfect Rupture uptime. I think I'd prefer the same basic HaT mechanic as live with the ICD can be reduced by haste. It seems cleaner and one less thing to worry about, between keeping three finishers up as well as the Hemo debuff, under the presumption that BS will wind up stronger and be the energy dump/cp generator.

Actually, I just thought of a third difference... the potential to keep Rupture on two mobs at once, but that'd really be a mess to dig through, so I'll just pretend that it's impossible.

Before my level 82 got stuck in the Moonglade portal (which is where about half of all the beta toons are stuck now it seems) I was able to get some looking at the mastery for Assassination and well, frankly, the tuning they have on it now seems incredibly weak. I saved a bunch of gear from quests so I could compare non-mastery gear with mastery gear. I didn't use my raid gear for comparison because the differing budget for stamina might have made things a bit weird (same weapons though).

Now, I'll admit it's too early to get worried about mastery being a junk stat or anything like that but I think it's fair to talk about how well things scale right now. When I looked at the paperdoll, my mastery quest gear gave me only a slight bit more IP damage on the tooltip but with severely reduced crit and haste numbers than with my non-mastery quest gear.

When I get home I'll pull out my notebook and write some stuff down but in the meantime, has anyone else looked at the tuning on mastery and felt the same thing?

About Mutilate and poisons : The idea of Scourge Strike-like talent isn't a bad idea, but since the tree already allows us to proc poison more often than the two other specs, we already got that. Under the Envenom buff, each auto attack is quite to 90-95% proc of the poison, and since you have your Deadly at five stacks, this is 90% proc for both hands to proc Instant, which makes over the double of the white attack (when non crit), and even more when it crits. I think making proc crit damage at 200% in the baseline of the tree will be more efficient than more nature damage from auto attacks.
The "Improved Poison" would so looks like : Your poisons rate is increased, and your poison crits makes now 200% of the base damage.

About Sub and HAT scale : They could make the internal cooldown of HAT on the same mechanics that the GCD for Caster, meaning Haste reduce that ICD. The problem will be if it would be a limit to the ICD, and what the limit could be.

Edit : The Improved Poisons could also increase the Venomous Wounds crit damage, which is logic since the nature proc comes from the poisons debuff

And focused attacks is disappearing, of course, making crit fall even further from current value.

I don't believe increasing crit damage bonus on poisons would really fix the scaling problem unless they also allowed instant (and deadly, assuming it can crit) to go off physical crit rate. You can already see examples of this in paladins and death knights. But maybe that would make agility too powerful?

Obviously the AP coefficient would have to get reduced quite a bit. If it WERE made physical, they'd probably go the whole way. (200% crits, melee crit chance, 8% miss chance)

I don't think there'd necessarily be any reason to go all the way to melee hit mechanics with poison; I mean, if you look across the DPS caster classes and specs, almost all of them get 200% spell crits either baseline or via an easily accessible talent - even DKs have 200% crits (and still use spell hit mechanics) even though they only have a very spells that can crit. Hence, the suggestion is making the spell-damage focused rogue spec have the same 200% crits as every other spell damage spec.

They *could* do it for all rogues in general, I suppose, but as it feels less necessary I was initially thinking of it in terms of a talent that makes sense for Assassination. With the significantly smaller amount of poison damage that combat and sub will be doing (probably less than 15% in both cases), it just isn't as large a scaling problem as it is when it's almost half your damage.

In terms of Venomous Wounds in particular: I've never been a particular fan of the nature damage component in the first place - it always struck me as existing mostly to try to encourage us to use rupture more than because it was actually interesting or flavorful. And now that Rupture gives energy regen, we no longer needed the pathetic nature damage proc to want to use Rupture. So it honestly wouldn't surprise me to see the nature damage proc removed from Venomous Wounds entirely (which would probably require renaming the talent as well).