Even though you pretty much just said it. You could keep Venomous wounds. Remove nature damage on ticks and add 200% crits for poisons. And still retain the same name.

Well, allowing some sort of scaling of the ICD on haste could be a way to fix it. Or, which seems much more like Sci-Fi, allowing HAT to have critical procs that grant 1CP (juste like regular procs) plus a given amount of energy. Ofc the second option would mean removing energy regeneration from recuperate, and also wouldn't fix HAT scaling issues as a free cp generator. But it would at least grant Subspec a greater scaling (crit + haste) than what it has right now, as I read recuperate currently doesn't scale on haste, and consolidate HAT as the main talent of subt.spec.

In fact, I just can't find any other way to fix these scaling issues without losing some part of the spec's nature except making HAT's ICD reduce as haste increases. Sticking free cp's to rupture ticks, or rupture crits as you suggested would make it look like some kind of seal fate.

The challenge of making HaT scale is that because Combo Points are discreet (you can't give a fraction of a CP, and there is a vast difference between 0, 1, and 2 combo points) the only thing you can really play with is the time interval between them. And since there has to be some kind of ICD to prevent OP-ness, we end up with function that is asymptotic to a non-zero value. Contrast this to the functions of other scalable powers, i.e. unbounded and approximately linear, and you can see why it's hard: you would have to design a mechanic that fits the curve in a useful way (e.g. powerful enough at low gear levels, better but not OP at high gear levels).

I played around with the haste model using some different ICDs. The haste model does result in a linear function: the "bonus" combo points (meaning CP beyond what baseline HaT gives you) you receive due to haste scale linearly both with haste and with ICD. E.g., halving the base ICD gives you twice as many bonus combo points, or doubling your haste rating does the same.

Examples:
Level 80, 1000 Haste, base ICD of 2 seconds = ~9 bonus combo points/minute
2000 Haste, 2s base ICD = ~18 bonus cp/m
1000 Haste, 1s base ICD = ~18 bonus cp/m

What's tricky is that although the resulting damage output is unbounded and approximately linear we can't really control the slope of the graph. Which means that if we want to tweak the value upward for high gear levels we change the value at low levels by the same proportion. So which case do we tune for?

We might just tune for low gear levels and say, "Well, at least it scales so that's better than nothing," but given Subtlety's overall low relative EP value for haste (low poison damage, low white damage, no autoattack procs), will we even have any on our gear? Will a few extra combo points be enough to make haste more competitive with other stats?

Would making IP scale from weapondamge help with the scaling issue Assassination has towards weapons? I know that would in a sense make all weaponattacks potentially be dualstrikes similar to how it looks for DKs, with both portions of the damage scaling from the same basic level.
I'm sure there would be ways to differentiate it more, but on the basic level, making poisons scale with the weapon and not AP/crit alone could help. Of course, it would mean that it would only get applied to the mainhand. It would actually feel logical, a bigger hit (due to better weapon) means more poison is applied. Am I completely wrong here?

Also, couldn't HaT effectively have a 'critical effect' of 2 CPs that would follow the Rogue's melee crit? It sound horribly illogical, but if it did it could retain the proc from raidmembers, but still increase in value as the Rogue gears up. This would naturally make Crit more powerful for Sublety, which could be seen as an issue since it appears Haste is the least efficient among the three specs.

One untapped resource so far w.r.t. the weapon damage issue is re-upping the damage increase on poisoned targets for Mutilate, which sort of also falls in line to Blizzard's design goal of decreasing passive damage, and wouldn't be uncalled for due to DWS not being available. And quite honestly the damage is pretty measly nowadays. Mutilate burst in pvp hasn't really been too overwhelming after the Overkill (which one might say was the real culprit to the nerf) redesign and more powerful pvp gear. Obviously they'd want to keep Backstab with Murderous Intent relevant as well, but unless I'm mistaken there's a fair deal of potential wiggle room between the two.

And, while I also appreciate poisons making much of the identity of the tree, it leaves mutilate rogues in a tricky situation if an utility poison is required. Mind you, at the moment the only one that really matters is handled by Deadly Brew in its current form, but there has been a somewhat frequent need for mind-numbing and wound poison at times and it only takes one encounter in an instance to make it quite painful (a la Hydross and Void Reaver).

In the next build Eviscerate has a 10/20% chance per CP to refresh Rupture. Thankfully, this takes some complexity out of the Sub rotation and brings it closer to what Assassination and Combat have.

Energetic Recovery remains the same so currently Sub still appears to want to keep Recuperate up (which means balancing against it ).

I really really hope they'll steer away from the whole idea of incorporating Recuperate into regular dps rotations. The buff to Serrated Blades is excellent news, but it doesn't address Subtlety's long ramp up time and problems with target switching. It only simplifies the standard rotation, but only so much.

Also, the Subtlety tree itself is still a mess. There's 33 talent points' worth of dps talents alone (32 if we disregard Premeditation, which I'm more than willing to do in terms of pve), so not only can we not pick up *any* sort of filler, but we can't even take all the dps talents if we want to subspec 10 points (which we do want, for both Combat and Assassination offer extremely attractive options.)

The other two problems that are still unsolved are Shadow Dance's horrendous clunkiness and Redirect being painfully lackluster.

Of course this has been said countless times, but allow me to reiterate: Making combo points stack on the rogue instead of the target would largely eliminate all specs' difficulties with target switching, which is one of the class's weakest points, both in pve and pvp. It would also be a huge quality-of-life improvement, making it much easier to keep track of our combo points without the need for addons.

"Stunlock" is an obsolete term, even more so in Cataclysm. A potential melee range 6 sec stun every 20 sec is nothing groundbreaking, especially if we compare it to such staple CC abilities as Polymorph, Psychic Scream or Freezing Trap. Considering Blizzard's aim to make pvp encounters longer and more eventful, a medium-cooldown CC ability would be more than welcome and would up a rogue's chances in prolonged fights.

To cut this short, as I doubt an extended pvp discussion is necessary at this point - I believe the pvp implications of such a change are a non-issue.

That said, this idea has been around since roughly 15 minutes after the class was introduced, Blizzard knows they want it, and despite the fact that they clearly are willing to make major revisions to class mechanics at this point, still have made only minor adjustments to combo point mechanics. Which implies that either

a) They think that the fact that we're better on single targets than on target switching is an essential part of the class's character, or
b) They're concerned about the PvP implications, even if you're not.

There probably is some element of the first; the situation as it stands means good players can differentiate themselves by managing swaps better - but I'd be frankly shocked if the second wasn't the major driving force behind the decision.

Paladins and warlocks now have been given a rough equivalent of self-sticky combo points. Neither of those classes has been handicapped at pvp since TBC and arenas hit (and that's a deliberate understatement). Neither of them has had as much trouble with target switching as we do. I fail to see the logic here.

Now is the time to introduce mechanical changes, test them extensively and adjust as necessary. Eclipse, holy power, shards, runes - these are all dramatic innovations being implemented and/or polished as we speak. Meanwhile, the rogue playstyle hasn't changed much since the game's launch and is probably one of the most stale models right now.

But I suppose you're right and there's no point in bringing it up again and again. Or is there.

Redirect, which is the new cp-switch ability, is a 1 minute cooldown.

The current incarnation of HaT only procs off melee and ranged abilities - which I understand means yellow physical crits. That drastically lowers its combo point generation in small groups and in pvp.

In a raid group, regardless of whether we assume combo points being generated on the rogue or on the target, we're still generating 1 cp per 2 sec (ideally) without any effort, so I don't understand why you brought that point up. The switch potential is of course there - that's the whole point of the idea, to strengthen our switch capability - but the most it can mean is that we get a "free" finishing move on our new target - which, depending on our spec, would be an Eviscerate (for Combat), an Envenom (which requires at least one stack of DP on the new target) or Rupture (for the 6% damage boost from Sanguinary Vein), or most probably a kidney shot in a pvp/add management situation.

And that's it. That's the only difference it'd make in practice. If we are already allowed to do that every minute with Redirect (which in a raid encounter requiring a lot of switching, such as the aforementioned Lady Gaga, is never going to be often enough, and in pvp is almost never worth the gcd it uses,) there's just a small step from there to perma-sticky combo points. Like I said, at this point there's little more to it than quality of life.


Edit: And of course a situation where we start a fight with 5 cp is inadmissible - and that could be addressed just as it has been for holy power: make combo points decay out of combat. Presto.