I was out grinding ogre beads last night with 5/4 and I had one case where I had 4 procs on one hand followed by 4 procs on the other hand about a second later. I had multiple other times where I'd get 4 procs or where I'd get procs from one hand followed by the other in less than 3 seconds. I don't think anything has changed yet.

This number is higher for Shamans due to WF weapon procs causing (in theory) an average of 40% more damage than a normal swing, but Brissa's comparison is pretty accurate otherwise. More hit is better!

Yes since the only shaman attack that does not depend on tohit is stormstrike (you only need 5 to make sure SS never misses) and ES and those tend to be around 20% of total DPS, tohit is by far the best stat to have as an enh shaman until you max it out. Unlike rogues and wars where special attacks can usually be somewhere around 50% of your total damage and it only takes 5 hit to make that damage cap out on tohit.

Note that you only need 5% tohit to eliminate the possibility that the WF attack itself will miss (which comes from talents). However since you cannot proc WF on a missed attack your wf can be considered part of your white damage for the purpose of figuring out what tohit does for you (as per brisa's calculation above). All in all, if you already have the crit to be flurried almost all the time and have UR up (in my mind 25ish buffed is about enough), then tohit is better then crit until you hit the cap, even before you consider that getting 1% hit is cheaper then 1% crit.

The cap however is 24% for the miss rate on DW against an equal level mob, so with 9 from talents you'd need 15% hit to reach it.

Again...you can never make it so your attacks NEVER miss. Even with ~16% hit I get SS and WF misses, although they are admittedly quite rare.

I'm not sure when other melee stats (namely, AP and +crit) begin to scale better than hit rating for shamans...I believe for combat rogues, once you start getting your +hit to somewhere in between 16% and 18%, +crit scales better. Well, that's how it worked in WoW 1.0; I would assume those numbers are slightly smaller in 2.0 since crit has scaled down a bit (for rogues anyway).

For shamans, I would think that the optimal +hit is a little smaller simply due to the fact that nearly half our melee damage comes from WF, which does not suffer from the DW miss rate. But...this is merely a supposition on my part.

Well, I will still maintain what I said about missing with WF and SS; I am certain they can still miss regardless of how much +hit one has...if only I could find something that confirms this.

Either way, changing the present topic a bit and touching on something you commented on earlier, I would also like to see nonhastened and flurry graphs for both WF and white damage with OH weapons that have 90 or 95% the speed of the MH, since those combos the most viable (in terms of itemization) in the game.

Untrue. There is a very, very small chance to miss no matter how much hit you have. A Hunter in my previous guild with 10% hit reported a miss, and it was the only one he had in a span of several months. I'm reasonably convinced that +hit that reduces your chance to "0" is actually just putting a reduction on your miss chance that essentially adds a digit of significance. 99.9, 99.99, 99.999, etc.

Also has lines for the effects of a 10% miss rate, which didn't turn out to be very interesting, and lines for 10% haste, where the number of procs does actually go down when hasted for a few speeds, although not by much.

Where does that idea find support? I agree that ~0 is close enough to 0 for all intents and purposes, I just wonder how people came to this conclusion, why the server might do that, etc. Curiosity, if you will

*shrug* whether it's that or the other theory that it's a .01 or .001 chance (the limit of accuracy for the number or whatever), the point is that to all practical purposes, it's 0. If I had to guess I'd agree with the server messing up, since spells do infact always have a 1% miss chance, so if blizzard wanted that functionality for melee it could have been included, and there is no reason I can think of why the number couldn't be set to 0 if one wanted to.

It's very easy to find something to prove it, run a personal damage meter that lists your misses/parries/blocks/dodges and records them over a long period of time (like kombatstats). If you find a miss on wf or SS, then grats. My data hasn't been reset in probably weeks and I have 0.

Thanks for the graphs, unfortunately it seems to confirm what I was guessing that there is less of a leveling out effect for speeds higher then 2 when your OH has a slightly lower speed then MH (as opposed to same speed or half speed), and slower is always better.

Thanks for the graph Shalas! <3

One quick question though: is it representing the same damage as the earlier graphs? I noticed it measures "arbitrary damage units" instead of "windfury damage" as before.

*WARNING* completely theoretical *WARNING*

If I had to guess what the real problem is with windfury and why it has a shared cooldown, the extra attack mechanic would be the culprit. I don't believe the quad hit bug was random at all. Extra attack procs seem to work from last known information. Because procs are slightly delayed, there is a time period in between the hit and the proc for your other weapon to hit the target.

For example:
OH hits (Proc)
Delay (MH hits, procs)
Proc goes off (MH delay)
OH proc uses last known, deals MH damage.
MH proc goes off dealing MH damage if OH hasn't hit exactly between last WF and the next one.

It seems WF proc hits count as hits for the sake of tracking last known info too. Normally there is not enough time for any type of weapon attack to go off between a hit and a proc, except that DW uses completely independent weapons. I'm willing to bet the reason this hasn't been noticed by anyone before is:
A) a human couldn't react in the time between a swing and a proc even with 0 latency & knowing the proc would happen to take advantage of it and
B) any accidental instant strikes which would be placed between a hit by an weapon and an extra attack proc are probably simply pushed back to after the proc by the server specifically to prevent that issue.

Windfury is a unique instance of an odd interaction. There's no real way to tell for other classes if extra attacks are being calculated correctly because of the lag times and seeming randomness of the combat log due to server parsing of the attacks. However, quad windfury immediately tells you something is wrong because of the known 3s cooldown between procs making the chain proc impossible otherwise.

So under the current implementation of WF and with the downrank bug, you will quite often see streaks of 4x MH and 4x OH windfuries. Its not in fact a bug at all but an interaction between WF mechanics and the downrank bug that gives a separate cd.

The interesting thing is that once you hit a dual proc, both WF are on a virtually identical cooldown timer, resulting in a higher than average chance of another quad hit if your weapons were identical speeds (or given the lag time and WF counting, within .2s I'd guess). Obviously enough, stormstrike adds yet another way to set both WF cooldowns to near-identical timers.

I'd be willing to bet this is why they had to add a shared cooldown. The fix would basically be rewriting the extra attack mechanic, unlinking the MH/OH windfuries would lead to the same problem as downranking did.