As mentioned in another post, KC might be usefull if you have 80+ focus and proc LnL, its generally a bad idea to waste LnL proc on AS and if you use ES, CoS, ES you will be focus capped, in these situations I think ES, KC, ES is the way to waste your gcd

I'm looking for some feedback on some of my napkin math. I've been trying hard to convince myself that the Haste 'plateau' is worth it, math wise. If you have 1.70s vs. 1.66s, you can still fit in 3 CS between ES's, you'll just be pushing back ES by X amount each cycle. This would only be present in full cycles (no AS to dump or BA/SS/Mark refreshes).

Explosive Shot: 32700
Cobra Shot: 14750

@1.70s
32700 + 14750 * 3 = 76,950 / ((1.70 * 3) + 1 (GCD)) = 12,614 DPS

@1.67s
32700 + 14750 * 3 = 76,950 / ((1.67 * 3) + 1 (GCD)) = 12,803 DPS

@1.66s
32700 + 14750 * 3 = 76,950 / ((1.66 * 3) + 1 (GCD)) = 12,867 DPS

This is only DPS for ES -> CS -> CS -> CS and doesn't include any other factors. The difference isn't terribly large because this DPS value doesn't translate directly to your actual DPS, the effects aren't felt as much.

To find the actual DPS losses you'd have to figure out your total delayed time (vs 1.66s) and average it out for a fight, so...

1.70s - 1.66s = 0.04s * 3 lost per cycle, 0.12s
0.12s = 120ms

Assuming a stand still fight like Chimaeron, and ignore sub 20% where you aren't doing full cycles each time... we'll say ~4.5 minutes fight time.

270 seconds / cycle time (~6s) = 45 cycles
45 * 120ms = 5,400ms = 5.4 seconds

So your ES over the course of a fight would be pushed back a total of 5.4s. This isn't enough to get another ES into the fight, and this doesn't account AS (which when fired in our cycle eliminates any haste gains because 1.66 + 1.66 + 1 (AS+GCD) has us waiting for ES).

1.70s CS takes around 500 Haste, whereas 1.66s takes 757. If you factor in secondary stat loss to Haste (~257 Mastery/Crit), it's even more debatable. Then add in Autoshot gains via Haste... I think it's still a pretty close call as to what is optimal.

Thoughts?

Between the 1.7 and 1.66, sure the 5.4s saved by having a tight rotation does not guarantee another ES, it may result in an extra ES depending on where in the rotation this part of the fight ends. That 5.4s window has a 90% chance of including the next ES. Furthermore, even if that extra 5.4s does not include another ES, it does include at least 3 more CoSs (or higher more higher damaging shots). Per your numbers that is 14750 * 3 / 270 = +164 DPS, which is a big gain. Even if we have AS on 1/3 of the cycle, that is at least 2 extra CoSs for +109 DPS. In addition, the extra haste provides more autoshot damage and pet focus from GftT, which is useful considering SV is short on pet focus. It also provides a little higher focus regen and a little more focus to possibly use. On the other hand, the extra crit increases pet focus from the slower autoshots and more Sic'Em.

In FD, if I add 257 crit and subtract 257 haste to my character, it is a small DPS gain of 29 DPS, but that is with me being 35 haste rating over 257 and some dead time before casting ES. If I remove that extra haste rating first to get to exactly 757 and then do the 257 crit/haste swap, the change is -141 DPS. For this case, FD showed -2 AS, -1 CoS, -3 autos, -10% WH uptime, and -0.08 static fps. This shows the 3 extra specials as expected, with the extra focus resulting in some of them being AS instead of CoS.

I'm confused about 'That 5.4s window has a 90% chance of including the next ES', I don't get what you're referring to.

You would get an extra 3 CS's based on a 5.4s gain, but this assumes you're getting a full 45 cycles. It doesn't include reapplying BA, new targets, etc. So we're gaining 0-100 DPS based on how many CS's we gain at the end of a fight (0-2). The thing I take away from this discussion though is that this is 1.70 vs. 1.66. The differences between, say 1.67 or 1.68 vs. 1.66s would likely be almost negligible. This would mean that our goal should not be a hard 1.66s, but more around the general idea of 1.66s. As close to without sacrificing other stats based on per-user reforging and itemization.

In reality there will be too few frames during an encounter that you will be able to get on the track and try to fix X numbers of CoS between ExS cooldowns. And tuning your stats accordingly doesn't sound like the wisest thing to do with all the given procs/temporary buffs/de-buffs. Favoring critical strike and mastery, while caring little about haste caps gives significantly better results with my current gear status both in simulations and live.

I wanna go back to that KC-issue real quick:
Do you think it is worth it to put 1-2 points into Improved Kill Command? Right now, I've got one point in Bestial Discipline, since it was the only point that gave me at least a tiny little dps-increase. I'm using KC instead of AS as a focus dump, which works just fine for me, especially during Heroism or Rapid Fire. I pop KC 3-4 times per boss, on average.

The damage per focus on KC isn't worth using over AS except in the rare occurrence that you're sitting at extremely high focus in the middle of a LnL proc (or if you're out of LoS of the boss and your pet isn't, which is also extremely rare). In the case that you will focus cap in the beginning of the LnL proc and any passive focus + CrS focus will be overkill the extra dpsc of KC becomes worth its extreme focus inefficiency (in Survival spec). In this rare case ES>KC>ES>AS>ES is the optimal sequence (assuming BA is on cooldown). But, considering how rarely you will have the focus to be able to accomplish this without taking focus away from extra AS's it isn't worth taking a talent point away from the definite dps boost of BD. BD means a lot more than it seems when you factor in greater Wild Hunt and Culling the Herd uptime.

If for whatever reason you prefer KC as a focus dump that's all well and good. But it isn't a viable option for maximizing dps.

Seems like they may not want us using 4p PvP in PvE lol. The new 4p PvE bonus seems quite good, letting us reach the next haste threshold without having to actually get haste.

Indeed, it appears Blizzard managed to kill three birds with one stone:

• They managed to actually create a set bonus that will change the way we play. 0.2 seconds less on Cobra, combined with lowered ES damage will most likely make 6 second rotation no longer optimal. It will also make our gamestyle more challenging and insteresting, widening the gap between good and bad players.
• It should make haste desirable once more - just as they promised. Outside of 6 second rotation, we're going to see benefits from haste all the way down to 1s cast time. I'd expect that with heroic gear it will be possible to reach at least 1.33s (3 cobras + instant not delaying ES). Whether it will really be worth doing so remains to be seen.
• Not only will PvE set become worth taking over the PvP one, but the huge boosts it will provide might help somewhat with survival's poor scaling at high gear levels. This should also ease the overall nerf somewhat.

Overall, I'm quite impressed. Doing all that with one move is really neat.
The only thing I'm concerned about is that they might need to overhaul survival (and possibly other specs as well) for the next raiding tier. Otherwise we might run into a situation when dropping 4xT11 might not be worth it even for higher iLevel items and the scaling problems will only get worse.

Bit confused about the haste soft cap calculation.
Is there something wrong with the paperdoll haste display or with cobra shot's tooltip that's known? I have 907 haste at the moment which gives me a paper doll haste of just over 7%. When I look at my cobra shot tooltip the cast time is 1.67 seconds. Where is the 757 haste rating calculation coming from?

Apologies if its in this thread somewhere, I skimmed through each page and didn't see the calculation explained.

I was no where near this on PTR 2 days ago. I did not try MM, but SV on Live was 12K, on PTR 11K. BM on live was 9K and PTR it was 10.5. Maybe your BM rotation is off a bit? Avg item level for me is 347.