Ok, just tested it and the result was 5750 DPS. As you see WW>BT>Slam! remains the best rotation to use

I ran a test of Berserker uptime on MH white damage.

How the test was conducted
I stood behind the boss target dummy in IF and hit it with white damage only attacks. Berserker was on my MH. I was wearing my full DPS gear. Unfortunately, I was not hit capped at the time. So, I had to do a couple calculations to adjust for misses (see below).

I use an uptime meter mod that basically has four numbers.
--Total downtime of the buff in minutes and seconds
--Downtime as a percentage of the total time dpsing.
--Total uptime of the buff in minutes and seconds
--Uptime as a percentage of the total time dpsing.

The uptime meter is very simple and I feel confident in its accuracy -- I watched the appropriate numbers tick for about an hour as the buff went on or off.

I ran the test for 2 hours.

Results
Total Berserk uptime over a period of 2 hours on MH white damage was 24.36%. Adjusting for not being hit capped, it was 24.94%.

Conclusion
I think we can safely assume that white damage only uptime for Berserker is about 25%.

I think the real question Blizzard have to ask themselves is what percentage of our DPS do they want Deep Wounds to be in terms of what the average tick should do over the course of an encounter and how many ticks per minute occur (taking RNG into account). 15-20% seems to be an acceptable ballmark, and the best way to achieve that in the short term is almost certainly going to be to lock it out of special attacks. At the moment I'm averaging just over 40 ticks per minute on boss fights with the average on fights excepting Malygos/Loatheb/Thaddius being between 1,000 and 1,500. That's normally 20% of my dps, second behind white Melee swings. My interpretation of recent blue posts is that they would much rather it sat in the 10% mark, but then they need to change how the mechanic works. They know as well as us that our rotation radically changes at 20%, and that at that percentage DW's damage increases as well.

Just looking at last reset's parses from raiding, on Grobbulus I do 18 executes with 61.1% being critical strikes, accounting for 16% of my 889,402 damage out (142,368 according to stasis) and is the third highest sourse of damage behind Melee swings and Deep Wounds. Melee swings (the only constant throughout all rotations) are 20.4% in first, Deep Wounds 16.0% in second. BT (10.1%), WW (15.6 MH+OH) and Slam (1.9%) account for just over a quarter of all damage dealt. Assuming you time Trinkets/Death Wish and pop Death Wish when Mirror procs (alongside say Loatheb's Shadow in my case), you get larger Executes and white hits which fuel DW, so where as you might get 500-700 average ticks at the beginning of a fight you're going to easily get ticks in the 1000-2000 stage at the last 20%, which pushes up the average and total percentage of damage. Stop Execute causing DW, percentage should drop by a notable number (2-3% at a guess without pumping it into a spreadsheet). Stop all specials causing it and it stays in the 4-500 mark and comes down to 10-15%, which is probably what they're hoping for.

E: Didn't realise how big a wall of text it was.

Heroic Strikes are also "melee swings". Not sure if you're not counting them or what but Blizzard seems to be ignoring them. Even on a Thaddius fight where DW ends up being 26% of my damage, Auto + Heroic ends up being 38% - 40%.

How about BT > Slam! > WW? I ask because delaying a WW rotation by .5s is less of a DPS loss than delaying a BT rotation by .5s.

5797 DPS so its basically the same as WW>BT>Slam!.

What!?!?! White Damage being about 40% of our damage????? Something is bugged.

This is why WWS is bad.

You know what's awesome though? When GC will ignore "what other people tell him" when it goes against what he firmly "believes", but when the situation benefits his cause, he will gladly pay attention.

We need "several other Warriors" to claim they are rage starved and that Rampage does not light up enough either. Maybe he will believe that too. I'm sorry but 15% is a nice amount of damage, but it's in line with WW/BT usually so I don't see how it's doing too much. You could try to play the "it's 5% per talent point and so early in the tree" card, but we actually give up more than three talent points for it and it's also accessible to two DPS specs and even a tanking spec. I'm pretty sure they purposely put specific talents where they do so they are in reach of every spec.

What I don't understand is how Deep Wounds's unmitigated damage is supposedly going to present a balance scaling problem if it's 15-20% of our damage. Feral cat bleeds are doing nearly 30% of their damage, so how does that not also present a problem?

Theirs is an ability that they constantly have to keep up, using both Energy and rotation slots. Ours is a passive, innate effect of a talent. That is the major difference between Deep Wounds and Rip/Rake.

A possible fix for the DW problem could be that it is allways a fixed percentage of the crit damage done, like mages ignite and not manipulated by secondary effects like Deathwish anymore. On this way it will scale with armor pen because armor pen will have an effect on the crit damage and resulting from that on the deep wound damage.

Well our trade off equation changes with the change to beserking(25% armor reduc -> 5% armor reduc), and your looking at this forumula...

Y = your armor
X = beserker uptime

[110 + Y/1200]/400 < X

So for...
armor class vs required uptime
11,000 29.8%
12,000 30%
13,000 30.2%

Essentially the OH %uptime on beserking needs to beat 30% for it to be better then massacre. I don't think any hard data has been posted for OH %uptimes from a warrior yet so we'll need someone with extra enchanting mats to test it. The closest relevent data I can find in this thread is Alamir's auto attacking MH uptime. He found the uptime to be approx 25% which could be resonably assumed to apply to OH also if only auto-attacks were being counted. The question is then if the off hand hit from WW pushes that uptime above 30% or not.

I planned on doing several more tests. It's just a matter of finding the time to do them. So, stay tuned.

is this with or without keeping in mind the other 5% penalty proc from MH ?

It's also been established that the OH hit from Whirlwind does not proc Berserking, so I very much so doubt that Berserking will ever be worth it in the OH.

The MH armor penelty should be independent from the OH armor penelty from what I read and what is on the wowhead notes. Meaning that if both MH and OH beserking proc your armor is reduced by 10% not by a stacking effect which would leave us with 9.75% armor reduction. I could be wrong on this and it should be tested, but even if they do stack the consequences will be very small.

The proc within a proc formula would be...

[110 + .95Y/1200]/400 < X

which for the 12,000 armor case results in 29.875% uptime needed, which is just .125% off from the non-stacking proc case. Taking into account that this case has be to be weighted by the amount of time it occures versus independent procs and your looking at a difference small enought that it can be ignored(a fraction of a point of ap).

Ok i tested this one and realized i had forgotten to check if bloodsurge runs out.

Slam!>WW>BT=5783 DPS
WW>BT>Slam!=5832 DPS
Slam!(1,5s)>WW>BT>Slam!=5837 DPS

So you are right, but the cases where the Slam buff runs out seem to happen only in rare cases.

Yes, although due to human reaction times and the changes coming in 3.0.8 where it can proc off a heroic strike right before both BT and WW come off of CD, it can happen. Really, I'd say human reaction time could be the main culprit.

That's cool that you can test based on the time left on Slam; I didn't think that Rawr.Warrior could do that.

Its pretty common knowledge that on many accounts GC is more or less in the dark, and tends to listen to horribads complaining about non-issues, or things that make the game even easier.

Plus yeah, DW is only absurdly high on damage on certain fights, and only with luck, I've seen it 2nd on Patchwerk, usually 1st on modifier fights (Maly, Thaddius, sometimes Loatheb). Its a little silly of Gc to be saying things like its above most skills when in reality it isn't. Of course you could show him several WWS parses but he'd say that its only one of many. Where deep wounds is now is fairly decent, and in terms of it being a significant part of arms DPS, I don't know where hes finding arms warriors that are DPSing right now, I was fairly sure that they were more or less non-existent at this point, though I could be wrong, haven't really be keeping up to date with PvE builds lately.

There are plenty of Arms PvE warriors, and there is a good discussion going on in this class subforum. Best in slot Arms is pushing ~5k DPS these days. Look over the last few pages of the Arms thread for some WWS links.

Some minor 'confirmation', i've just done some testing with 10.13% hit on a lvl 83 boss and after doing 463 yellow attacks there were 0 misses. So for anyone still wondering if the base chance to miss yellows is 8% or 9% this indicates that it's 8%, even though it's a rather small sample.

I built my "own" rawr.dps warrior module based on a simulation and checked it in, but it won`t be released because it is a Simulation which has some problems with the rest of the program, i.e. its too slow and sometimes 4 agi is better than 6 through its simualtion based nature. But for testing rotations, talents etc. i find it good enough.