No it doesn't mean that sebudai's wwstats prove that.

But how about next week, when i will face brutalus again, to play with the enhancement / resto spec

use drums and haste potions the same way i used tonight,. (our first kill hooray!)

Wow Web Stats

and see how it goes?

would that help me determine even for me personaly what's best?


PS i still don't see searing totem dps on the wwstats. My recount showed 1899dps :/. WWStats shows 1806 and no searing totem pff. I even used the last 1min50sec my fire elemental pffeh

i won't deny that, i have encountered these times of very low and weird dps, but i wonder if we could reverse engineer the WWStats, cerrtifying the simulator in development with real results.

Brutalus fight is very useful for these tests as it's a non-stop nuke for enhancement shaman for 6 whole minutes.


What we need from WWStats is

Initial shaman stats
Stable buffs on us and debuffs on boss on whole fight
Temporary buffs on us and debuffs on boss, time of use and length of effect
Final dps

Then use the simulation of testing with WWStats info and see if the result (final dps) is similar to the real one. After fine tuning the simulator we could, i guess, get some real results out of it for specs / sub specs

Not really. The amount of WWS parses that would be required to obtain a statistically significant sample is nigh impossible, not to mention the number of uncontrollable variables.

Here is what I found using Yo!'s sim

Stats:
AP - 1568
Crit% - 30.61%
Hit% - 14.88% (used in the neutral and enh/ele sims); 17.88% (used in the enh/resto sim)
Haste% - 0%
Armor Pen - 553
MH DPS - 97.5
MH Speed - 2.6
MH Expertise - 10
OH DPS - 96.5
OH Speed - 2.6
OH Expertise - 10
Spell Hit% - 0 (used in neutral and enh/ele sims); 3% (used in enh/resto sim)
Spell Crit% - 5%
STR - 242
Agi - 382
Boss Armor - 7700

Under Procs & Stuff my OH has crusader (yes I know I should have mongoose, but I'm cheap and currently not raiding). For trinkets it's the darkmoon card crusade, and abacus. 2/5 Concussion for the enh/resto sim, Concussion, CoF, and reverb all maxed for ele test, and all of them set to 0 for the neutral test. 24 shocks in 2 minutes for ele, while only 18 for resto and neutral. I didn't change anything in the buffs and debuffs tab.

My results were 1432 for enh/ele, 1406 for enh/resto, and 1370 for the neutral test. I ran each test 5 times and it was set to 1000 hours. Although some things about this is that on the updates under February 12 it says "- Higher ranks of Reverbation produce slightly less dps with same number of shocks per 2 minures due ti the limited Global cooldown simulation. This is offset if you increase number of shocks (that is why you choose reverbation anyway)" but you can twist w/o ever hitting a GCD issue (SS, shock, WF totem, GoA, Shock, repeat). This means that the sim is penalizing Reverberation even though it doesn't need to. Although when I set the rotation to 24 shocks but 0/5 Reverberation it only comes up with 1415 dps. If I set it to 0/5 Reverb and 18 shocks I get 1390, and if I set it to 5/5 Reverb but only 20 shocks it comes up with 1390. So while I don't know exactly what it is going on, it can be seen that the sim penalizes you for getting 5/5 Reverb even though you wouldn't hit a GCD crunch.

So the total difference that I see per the sim between the 2 specs is about 26dps (or 36-62dps difference vs just 45 in enh), BUT given that the sim penalizes the ele spec it makes me wonder if it isn't actually a higher difference. Also the sim assumes that you are beating on the target 100% of the time. This means that on any fight where you have to move out of melee range enh/ele has an even bigger advantage as your shocks and searing totem will make up even more of your total dps than they would in the sim (granted not a whole lot more, but still more).

While I can't try to parse this in game due to not being able to raid anymore, when I was raiding I had seen about a 50-75dps increase on the same fights in MH and BT the following week that I went from enh/resto to enh/ele . Some of my guildmates actually commented on how they noticed my dps suddenly jumped up from one week to the next. Now I know this doesn't "prove" anything as there is no actual data to back that up, it is what I saw in game.

EDIT: doh I had completely forgotten I had already run tests before =_= Although this post is a bit more thorough than my last one was.

We ran the sim as enh/elem having 9% hit and enh/resto at 12% hit to account for the difference. It's a much better test to use the same gear set and notice the difference the spec change has on its own.

You are going to be hit capped given just how much gear has +hit. enh/resto is hit capped from the start with +9% (you forgot about DW Spec adding 6% in your original post) from talents, while enh/ele needs about 45 hit rating to be capped since it only has 6% from talents. Getting 45 +hit is nothing. Also it isn't like you are just magically going to find gear that doesn't have hit rating on it. The only thing that you could really control is enchants and gems, both of which should be the same for both specs. This means that even if you could magically find gear for all slots that had 0 +hit, ele would still out-dps it since it would only need 30 from gear (as your helm enchant should give 16).

By setting hit% to 9% in both ele and resto, and changing the AP to 1626 for the resto test I got 1348dps for enh/resto and 1389dps for enh/ele. So actually if you were to drop it down to the min hit needed you would actually hurt your dps at certain gear lvls (this has been stated before, but given the amount of +hit on gear never really been explored for where +hit becomes more valueable than other stats).

EDIT: Something else I just realized. Totem Twisting. With 5 second shocks it is a lot easier to twist as your shock rotation, SS, and twisting are all on 10 second CDs (well shocks are 5, but you do 2 per 10 seconds). With enh/resto you will run into your shock timer landing on the totem twisting timer more often, or have the GCD caused by the moving shock timer lock you out of twisting occasionally which could cause WF to fall off from time to time.

Crazy Shaman's DPS & AEP calculator (c) Yo
Uusing the simulator and only changing the +hit and the elemental talents, no procs and no buffs other than wf, it shows with every simulation that enh/ele is marginally better. The minor differences in DPS would go to play style and not what spec is better. With personal gear the totals will be different. A draeni will be better at enh/ele, and an orc with duel axes will be better with enh/res. Any other race will be the same any way you spec as long as you have a minimum of 43 talent points in enh.

Here is my personal and new favorite spec:
WorldofWarcraft.com -> Info -> Classes -> Shaman -> Talent Calculator


Here is the simulation results on a es fs rotation.

ap=1500
crit%=25
hit%=6 enh+enh/ele, 9 enh/res
mh+ohDPS=100
mh+ohspeed=2.6
spellhit%=0 enh+enh/ele, 3 enh/res
spellcrit%=5

enh
613
wd=33% 206
wf=25% 156
ss=9% 59
sh=24% 150
st=6% 40

enh/res
639
wd=33% 214
wf=25% 165
ss=9% 61
sh=24% 155
st=6% 41

enh/ele
644
wd=32% 206
wf=24% 156
ss=9% 59
sh=27% 175
st=7% 46

Even if I use top end stats the gap between enh/ele and enh/res just gets wider. Its all dependent on AP. If you have Mental Quickness then you should have elemental offspec. Using the simulator is easy and you can spend 10 minutes to find out what I've found. At 3k AP and 40% to crit enh/resto is gonna do around 1700-1750, and enh/ele will do 1750-1800 DPS. Your total spell casting totals to around 16-17% of total DPS. However resto has some utility that 16/45/0 does not have. So again the minor differences in DPS would go to play style and not what spec is better.

When your at 3k ap and have 1100+ spell damage do you really need to worry if your totem is gonna reach the other meele characters standing around the same boss? Or is being able to shock every 5 seconds worse than ankhing every 40 minutes?

I'm hesitant to engage in theorycrafting here, because I'm nowhere near as brilliant at math as most of you, but there seems to me to be an important mistake that has been made throughout this entire thread. Perhaps I'm wrong, but let me lay it out. We're working on the theory of the difference between the two sub-specs assuming perfection. Which is easier to handle, I'll admit, and I understand the desire to remove equal variables from the equation. That having been said, neither spec can be executed perfectly. For instance, let's look at the 5s shock spec.

0sec
FS
1.5sec
WF
2.5sec
GoA
3.5sec
SS
5sec
ES


Except you're never actually going to see that. With imperfect button mashing and latency, you're not going to have a perfect transition from GCD to GCD. Now I realize this seems rather minor, but isn't this likely to A: put you in a race condition at some point and B: keep you from achieving your perfect 24(20) shocks per 2 min? The same thing goes for the resto subspec, but it's more eliptical, which is why I chose to use the more repetitive, "cleaner" spec for example. I'm not entirely certain where to go from there because I'm not sure how time lost to latency would stack (sometimes you get lucky and mash so that you're early on your screen, but on time when it reaches the server), but it stands to reason that it's not only possible, but probable, that one subspec or the other would be affected more by imperfection in the timing. Also, I know people have said they don't run OOM with the ele subspec, that SR keeps them topped and whatnot, but I'd like to see a sliding scale showing mana return averages on SR for different AP values at certain hit ratings to see how this works in practice. I know someone threw out the figure of 600 something MP5 back on page 6 or 7, and it seems to me that since a 600mp5 cost works out to 14,400 mana over 2 minutes of SR CD, that's either blatently wrong for a class that sees maybe 150mp5 inside the 5sec rule (which we'll never get out of), or someone's not fessing up about chugging mana pots. The 600 figure isn't what was posted, it was actually 600 and some change, but you get a rough idea. I cheated and skipped some pages in the middle because it's bedtime and I wanted to get this down before I slept and lost it all, so I apologize profusely if this was somewhere between pg 7 and 13. I'm sure the sim takes into account the difference in UR uptimes from 3% hit to 5% shock, I can't imagine we're talking more than a fraction of a percent, but I haven't seen that factored into the theorycrafting either.

Am I just being too nitpicky or are these things valid topics of research? Again, I don't claim to be God's gift to theorycrafters, so please educate me.

Personally I use this rotation:

SS - FS - WF - GoA - ES

That allows for more slack since there's going to be only the 2 totems or 1 SS between the 5 second shocks instead of your SS + 2 totem + gcd left by shock which is 5 seconds, having no slack at all.

Second, you're gonna need to have a ret pally or a pally dedicated in keeping up JoW, otherwise you're gonna run OOM about a minute into SR cd

Well first, this isn't the theorycraft thread. But the spell-queuing latency rehaul in 2.3 may have eliminated that consideration by letting the spell go off as soon as computationally available if you hit the button a little early.