Doing Leo with, and without Searing Totem is a world of difference. If nobody runs anywhere stupid during whirlwind (or the damn hunters don't stop dps when they should), he goes right back to the tank spot and it makes DPS in elf form so much easier.

What +hit do you assume in those calculations? The math in the OP assumes 15% total (talents + ~95 rating, which is admittedly a bit low for a raiding shaman but possibly not low for early Kara-level), and comes out with < 98% UR uptime and 57% flurry uptime with 30% crit. (Though the Flurry math leaves out other sources of crits, like Windfury and Stormstrike, so it will actually be higher. I don't have a good way to model it at work, but I can't imagine it would be over 20% higher with 10% less crit chance by adding WF and SS.) Don't want to say you're absolutely wrong, but those uptimes seem unreasonably high for very low crit% without some math to back it up.

[e] Put another way, your autoattack Flurry uptime (i.e., not counting WF and SS) with 20% crit will be 42% with 13% total miss, using the formula on the front page (1 - (1-Crit%)^3)*(1-Miss%). I'm saying that I'm skeptical that adding in WF and SS, which for me are ~0.35 hits per second on average compared to ~0.78 autoattack hits per second with flurry down/~1.1 hits per second with flurry up, will almost double that uptime even though they can't miss.

That's actually not that surprising - I don't remember if it was in this thread or one of the other shaman threads around here, but supposedly Searing Totem is basically the one thing in the game that doesn't get its aggro reset when mobs do a "full aggro reset," which makes it both useful on some fights to give the tanks a little time to track down the mob and build some threat, and horribly bad on fights where you don't want it running after a totem.

Thanks guys. I am switching my main to an enhancement shaman as my guild currently raids with 0 shaman. This forum has been a huge help.

There are a number of mobs in Zul'Aman that split in two...if the tank isn't ready, or lags a bit, it's really easy to lose the split mob and with it a healer or two. Last night even Consecrate wasn't keeping the bears with their riders.

It is the only time I've ever found a use for the Stoneclaw totem. Dropped at 21%, every time the split mob went for the totem and remained within taunting distance of the tank.

With the sunwell being constantly fubard my guild has been hopping into the t5 instances and even gruul's for more DSTs in the past couple weeks since after the t6 clear we're all bored out of our minds. I happened to pick up a Shoulderpads of the Stranger this week and as far as my personal AEP goes, it's significantly higher depending on party makeup/buffs and my weapon choice. I tested it in raid and as a draenei shaman without any other expertise gear the .50% reduced dodge equated to somewhere around .7% over about 4 hours in hyjal and bt last night compared with the Hidden Predator shoulderpads last week.

The simulator indicates an approximate 3 dps increase with stranger over predator; real time testing indicates that they equate to about the same dps on trash with the leather giving a slight increase on bosses (10 dps average over 8 bosses). I've ran sims with and without kings/might/lotp and they all indicate the same pattern.

My question is this - at what level of expertise does the curve in dps gain justify equipping expertise at the loss of some crit? I.e. the curve stays level for a certain amount of expertise, starts gaining slope and hits a point where the slope is maximized, what is this point when the possible loss of crit is taken into account?

I seem to recall SR being a PPM buff. So any variation of 2h/DW and speed is irrelevant.

I'd like to see some evidence of this.... sometimes when I use it I get 1k mana back and other times 7k, if it was PPM I would think it'd be alot less streaky.

This was a claim first made by Snorkle on Blizz EU shaman forums, IIRC. He claimed he had tested this over a long time and concluded this.

I did some data parses of some shaman tests of SR and found that it was a chance on hit. However, my analyses weren't that great (though Oggie did a great job of gathering the data). So, I thought it was a flat chance on hit.

However, snorkle did longer tests and came to pretty convincing conclusions about it being PPM. His tests looked good enough that I didn't feel that it was justified to spend the time to confirm or deny - I've taken it for granted since then that it is indeed a PPM effect.

But, as always, blasted lands is --> that way if anyone feels bored and wants to do some testing for further confirmation or a possible denial :-)

@Yo!
hello, i'm a long-time lurker of this thread, but this is the first post i've made.
since enh/ele was found to do comparable dps to enh/resto, i've been running with a 16/45 spec, and it works out pretty nicely. however, when running the simulator, if you have 0/3 or 3/3 elemental devestation, everything transpires exactly as you'd expect, but if you put in 1/3 or 2/3 (i currently have 1/3), then your crit chance drops to basically 0, you have 0 UR uptime, and your dps goes down by around 35-40%. i even tried adding an extra 0 to my crit chance, and at 300% crit, i have a UR uptime of 2% :[

for now i've just been testing with 0/3 elemental devestation, but i just thought i should make this bug known.

also, awesome simulator Yo!, mega-useful, despite this small discrepency :]

Well, we operate under the assumption that SS is a "two-roll" attack. Technically, it is a "three-roll" attack in that each hand has independent crit rolls.

Well, apparently the healing debuff has been removed from Flametongue now, according to mmo-champion and worldofraids. I wonder if they're just abandoning the idea, or if it will be reintroduced in some way again before the patch is released.

I haven't seen the post myself, but apparently Kalgan directly stated that it was intentionally taken out, according to several people on the Blizzard forums.

WoW BlueTracker: Lifetap change reverted
As an aside, there is some interesting info in there on the under-/over-representation of certain classes at 1850+ and 2200+ ratings, for each bracket. Take it with a grain of salt though, it's based on the total number of max-level characters of each class, so classes with lots of max-level characters (Warrior) are "under-represented" even though they are found on many arena teams.

It's a legitimate question and an "I don't know if anyone's taken the time to do this" would have sufficed instead of telling me to go do it.

My point in asking was to not waste 5 hours of plugging numbers into spreadsheets and then simulating 10 different sets of stats and gear if someone had already done it and was willing to share what the median line there is. You can't just play with numbers in the simulator and get correct dps, you have to take into account every piece of gear and every stat change that you would gain the expertise from or the AEP values mean nothing.

If I have time over the next few days I'll try to model and graph something as a relation between Expertise AEP and dps throughput with the new t6 and shard of contempt. While I'm doing that, if anyone has input regarding the relationship or general comments about when expertise becomes optimal, regardless of effect on other stats, i'd appreciate it. (If you don't know, then don't waste your time typing an answer)

Easiest way to compare different style of gears is put it them to two different warcrafter sheets. Then put these stats to YO's. Take about 5minute and you can see your looks too. Don't try manually calculate AP relationships. Just sim it and see dps. Simple.

There's a quest in Blade's Edge where you have to fight off waves of mobs that use only AoE attacks. By dropping stoneclaw, which is immune to AoE damage, the quest is trivialized. It's one of the quests near the "other" dark portal.

Also, the mobs in mechanar that toss aoe grenades all day are easily handled with stoneclaw.

Edit: I'm tired, and I'm not sure that the below stuff actually conveys my meaning: PPM or no, the proc rate is high enough that you still benefit more from multiple, faster weapons (as far as mana regen is concerned). I don't think I'll be swapping to 1.3 speed daggers any time soon for mana regen purposes, but hey, you never know.

If it is PPM, it's significantly more procs per minute than I'm able to hit with a slow 2H, since I get more than double the number of procs when using my DW weapons compared to a 2H. I haven't timed this out and taken exact data, but I'm getting back more than double the mana DW than I am 2H.

That would mean that, even if it is PPM, it's such a high number of procs per minute that you will still benefit greatly from using DW instead of 2H. Possibly worth swapping in 1.3 speed weapons (like Finkle's skinner) during raids if you're really low on mana? I remember doing this when 2.0 first came out, but I don't think it would be much of a benefit given how few mana problems I really have in current raids.

I remember some testing I did long ago, when the taletn appeared for the first time.

With a very slow 2-hander (3.9 speed) it proced almost on each hit, while with a fast dagger (1.4 speed) it proced like one third of the hits, so it always seemed to me that it's a ppm with very high procrate (and of course as a ppm you will get twice as much procs when dual-wielding).

Well it was more than a year ago so its mechanism may have changed in between....

I never thought of that. Seems to me on heroic the mighty Stoneclaw totem could only take two of their AoE blasts before melting, and chances are I'd still be close enough during that period to take splash damage.

Obviously if you dual wield, you have twice as many chances to proc. That has to do with the number of chances, not the speed of the weapons.

You won't gain anything from swapping in faster weapons. Indeed, you'd actually have LESS mana returned. Here's why:

Let's say it's 10 PPM. With a 2.6s weapon, you swing 23 times a minute. Your chance per hit would be 43.4%. With a 1.3s weapon, you swing 46 times a minute. Your chance per hit would be 21.7.

Let's assume a really nasty miss chance: the full 15%, plus 6.5% dodge. With two arms swinging, at 2.6s, you'd have 18 successes inside of the 30s window. With two arms at 1.3s, you'd have 36. In either case, you'd have an average of 7.8 successful swings.

However, in both cases you'd also have a chance to proc off Windfury and Stormstrike. Assuming WF every 5s, and SS every 10, this is an additional 18 (minus dodge, 16.8) chances to proc SR returns -- at the same rate as your autoattack swings! With 2.6s weapons, this is an additional 7.3 procs. With 1.3s weapons, this is an additional 3.65 procs.

In conclusion: when nothing is normalized, slow weapons are a good thing. Didn't you hear?!