Then it's not so hard to test... Use some 1.4 dagger (or even 1.3 from vendors) as a MH and ask for help from shaman. You will need haste totem, food and a Heroism/bloodlust. Make 100-300 autoattacks with buffs and same amount without = profit )
P.S. Can't do this right now myself =/

Is there any possibility crit will become useful for combat in 4.2? Or is there a main reason crit will always be bad that i'm missing.

The critical strike modifier for Rogues is currently only 150% and the % gain per rating is far from impressive. Additionally, neither our talents nor mechanics offer an alternate reason for crit stacking to become viable.
None of these factors are going to change in 4.2 and, as such, there is no reason to believe crit rating is going to become a dominant stat any time soon.

Basically what I'm saying is that crit isn't so much 'bad' as that it just doesn't really do anything all that impressive, especially compared to the alternatives.

Critical strike modifier for rogues is not 150%, it's still 200%. It is poisons who has a damage multiplier of 150% when landing a crit.

However as you say, we, as of patch 4.0, no longer have pray on the weak, which made all critical strikes hit 20% harder on targets with lower health than you, percentage-wise. And as for assassination, no focused attack, that returned 2 energy for every melee critical strike. These two talents made crit partially better before the redesign of the talent trees, than it is now.

This results in no additional gain other than raw damage from crit, which yields haste, mastery, expertise and hit up to their respectively caps, to have higher ep values than crit.

The value of crit might increase some as we get higher tiers of items, but so will also haste and mastery, which in the end will leave crit as a sub-par stat.

Well, this post shows two Main Gauche procs at the same time, so...

From last night I had simultaneous procs at various points:

[20:41:48.471] Jinglejangle Main Gauche Onyxia 4424
[20:41:48.471] Jinglejangle Main Gauche Onyxia *7937*
...
[20:55:51.626] Jinglejangle Main Gauche Onyxia 3242
[20:55:51.626] Jinglejangle Main Gauche Onyxia 3146

and at least 15 other times throughout the night.


ICD doesn't exist if it can proc itself, but further proof it is not 1.5s:

[20:48:57.018] Jinglejangle Main Gauche Onyxia 4203
[20:48:57.791] Jinglejangle Main Gauche Onyxia 4346 -- .78 seconds

and

[20:50:41.216] Jinglejangle Main Gauche Onyxia 4549
[20:50:41.382] Jinglejangle Main Gauche Onyxia 5170 -- .17 seconds

So I ran a quick test for if Main Gauche can proc itself, and came across this in the log:

The sectioned area appears to show a Main Gauche proc off a Main Gauche, noting the time between the swings.

This information is most likely an anomaly in the combat log. See later posts for details.

Sakuratei show more of the melee attacks previous to the pasted combat text. I know looking at a specific example of two MGs at the same time looks convincing, but we should examine a larger data sample. During my test, with a 35% proc chance and 300 MH melee attacks, if MG proc'd itself we would expect to see something like 140 MGs. But there were only 97, which is much closer to the expected 105.

May I suggest that 300 autoattacks is not really conclusive or beneficial? Expected procs are a good way to guess. Seeing two procs with no autoattack anywhere close is proof. With such a high amount of mastery in Assassination gear we should be able to produce a few more examples. It's probably a good idea because the timing of the combat log can sometimes be wonky. Note that the period in between Sakuratei's white attacks varies quite a bit. 2.6->2.0->3.0->2.3.

A better test would probably be to stand sideways so that you're hitting specials but not white attacks and then using sinister strike no faster than every ~3 seconds. Two mg procs in between any of those sinister strikes will be indisputable. I'll try to get one now.

Edit: Small sample size (250) but I was unable to produce a double proc using SS. MG still procced in singles so direction isn't a factor. I believe this is the best and most conclusive way to test it. I tried a small sample size because if I had produced a proc during that period it would have been an answer, but not getting one obviously isn't also an answer. Hope someone else has more time to test it.

The log that my section was taken from was deleted by one of my officers (he considered target dummy parses as spam...) so I will have to record a newlog at some point. As for combat log timing disparities, I agree that the timestamps may not always be accurate, but in this case it's really hard to think that has come in effect. Attack 2 triggers a MG, a second MG is shown 0,4 seconds after the first MG. Attack 1 hit 3.1 seconds before Attack 2 and 3.8 seconds before the first MG, and Attack 3 hit 1.2 seconds after the second MG. Even with timestamp disparities, I don't think personally that the combat log could be so far off as to change the order of which attacks occured (attacks and MG) with those timers between potential proc attacks and the MG pair.

However, I agree completely that this is not definite proof of anything. It is more of a data share to hopefully motivate more testing on this, as I am now definitely unsure if MG procs off itself or not. I will be running a new test of autoattacks with MH only for roughly 3 hours and see if I can duplicate these results, hopefully multiple times in the same log.

I ran another test, only ended up as a 30 min parse as the game disconnected me for inactivity. With 12.85 Mastery, equaling roughly 25.7% chance to proc Main Gauche, 783 melee attacks landed and 193 Main Gauches were performed. 193/783 = 0.2458 = 24.6% (Within reason of variation).

There were no events of a possible Main Gauche->Main Gauche proc in the log. You can read it through yourself here.

Between this and liftir's results, I think it's fair to assume my previous result was an anomaly with the combat log and that Main Gauche cannot proc from itself.