Why does haste, in Grim13's spreadsheet, show as being 20% superior to crit on a rating to rating basis?

If that is actually right, the compendium needs to be adjusted.

I've done some napkin math, and it seems logical that haste should toast crit in SEP. You need 40% more crit rating to get a 1% crit versus 1% haste comparison. Now how much of your damage is white/heroics?

I also don't know the full impact of haste. Some day it increases the PPM of enchants, some say it increases overall rage generation more.

Basically, we need a solid 1% haste versus 1% crit comparison and crit has to win by a 40% margin.

Btw, I'm talking very high end gear here.

Technically, haste increases rage generation, decreases Flurry uptime (under the same idea that slower weapons = higher Flurry uptime) and has no effect on PPM mechanics.

Crit increases Flurry uptime, increases rage generation (in a spiky way however) and also has no effect on PPM.

Things to note about both:
Crit is always at least a 100% increase on damage for that swing (more with Impale and RED)
Haste is guaranteed extra damage, the extra attack gained from 1% Haste is subject to the entire attack table and therefore could be a glancing, or a miss, or alternately could be a crit.

Haste affects all of your white damage, none of your instant attack damage, and a minor effect on Heroic Strike (due to increased rage generation) damage unless you have enough rage to Heroic Strike every mainhand swing. Giving a very rough estimate, I'd guess it affects around 40% of your damage, with an additional 1% of your Heroic Strike damage for every 1% of Haste. I.e 1% Haste should average to around 1% more rage generated long-term, therefore 1% more Heroic Strikes.

Crit affects all of your damage that isn't already a crit, excluding the 25% of your white damage that is glancing. This should be something like 50% of your damage(?) assuming 40% is from white damage and you have a 40% crit chance.

The way I see it, crit has a slightly higher value than Haste, and affects slightly more of your damage than Haste, however it comes at a higher cost. 22 rating vs 16 rating. I'm too tired right now to try and actually validate any of the above or to go into any further detail and thus there's a good chance that most of the above is blatantly wrong. I'll sleep on it.

Wouldn't the uptime be the same comparatively speaking? I guess you would get less instant attacks in between Flurried swings under additonal haste effects than without, but the increase in Heroic Strike speed (and a greater chance to crit as opposed to a glance or miss) should somewhat even that effect out shouldn't it?

My expertise drops to 5 (orc expertise) when offtanking and depending on when the add(s) I'm tanking are killed I usually swap back to a 2-hander so slamming etc and so I can also lose that 5 expertise bonus.

However, the two slam dodges are recorded at Gurtogg Bloodboil and Reliquary of Souls where I can guarantee I had the same values of expertise.

So doesn't look like I was dodge capped after all!

Finally managed to get the log browser working WWS Dodge Report for Actor Cavein.

For fury warriors there should be a decrease in flurry uptime as haste only effects their white hits (assuming they have enough rage to keep their specials on cooldown).

For slam warriors the whole rotation is hasted as they do not chase global cooldowns until the 2.5 second swing speed soft cap. So until that point relative flurry uptime should be the same (ie. in terms of time flurry uptime will be lower but in terms of affected swings it should be the same).

Beyond the point of the 2.5 second swing time there would be a decrease in flurry uptime but this should only happen under special circumstances (bloodlust / heroism, haste potion) and the relative benefit of flurry would be diminished if you kept to a normal slam rotation.

You neglected to factor in WF procs. This is important as it is our biggest external DPS boost. Arguably it's rather moot as Flurry increases attack speed, and thus more opportunities for procs, while haste provides a static increase of the same.
I also think your aside on Heroic Striking deserves some further exposure. Again, it falls into the senario mentioned with crit producing Flurry and thus increasing attack speed, whereas haste provides a static gain.
I think it does deserve some mathmatical modelling, as it really does seem rather close, and has been highly argued but never dis/proven.

The 2.4 Fury dps thread is also discussing the effects of haste. The impact of this research could turn the warrior DPS world upside down. Heck, even the melee dps world.

http://elitistjerks.com/698186-post168.html

I'd like to know where the 3300 number on all bosses except Mother comes from... His full report shows 1831 hits (which includes glances) 997 crits, 15 parries and 6 dodges. The total doesn't even come to 3000.

Can you do a confidence interval with those numbers instead?

As you can see in this WWS i got only 1 dodge the hole evening in Sunwell and i have 6% anti dodge (Trinket, Talent & Orc). If the % of dodge remains 1% after cap, wouldn´t we see more dodges?

Sorry, my estimate was really sloppy. Will redo and edit my first post. Was to lazy to go through the report and add up all boss fights so I included trash in my figures which is obviously wrong.

I thought the 1831 figure (landed hits) did not include crits, glances, parries, dodges or misses? My 3316 figure comes from adding 1831+997 (crits) + 511 (glance) + 146 (misses, dodges, parries) - (the same in the mother fight). Looking at the wws again I see that Landed hits does include glances but not crits nor parries. I need the total number of swings (i.e. normal hits+glancing hits+crits+dodges+parries+misses) for all boss fights, does the final figure you give include dodges, parries and misses (I guess it does I can't get the numbers to match up otherwise)?

Using these figures we get a 95% confidence limit interval of [6.27%,6.67%]. Need more swings to pinpoint it but 6.5% is well within the limits for now and 5.6% is definitely out.

It's not crazy at all. The whole point of hiting the 9% hit cap is so your specials never miss. However, they can still dodge/parry. With shard and WM you can reduce dodged specials by 4.75%. Now you're dealing with 1-2% of all special attacks that do not connect with a mob for w/e reason. The proc for AP is the tradeoff for a DST's haste proc. Obviously the latter is superior, but bundled it's Shard ftw hands-down.