It would be totally awesome if you could add Scarlet Fever and Blade Barrier to the priority for Blood.

Wild magic potion are not included, however I don't see how 200 crit for 15 second could be better than 500 haste for the same time.

with a haste pot and bloodlust I am put over the gcd cap of 1 second, which greatly deteriorates the value of haste. As a result I'm not sure if wild magic or speed would be better for pre-potting in this situation

The main use of haste is the increased auto attack speed, as far as I know. For that it is even more beneficial to use speed pot and Bloodlust together than after each other as their haste bonus is multiplied not added together (haste from haste rating is multiplied with %-haste buffs and multiple %-haste buffs are also multplied by each other).
As the only GCDs that are reduced by haste are spell GCDs and our special attack output is still restricted by runes/runic power which don't come in faster with more haste (yet) and haste is such a good stat regardless (which is because of its increase in auto attack damage), I don't see how 200 crit could possibly be better than 500 haste just because with the haste you would be reaching the GCD cap for the few spells we use.

The online client isn't working for me as well, but only on Firefox (tried with deactivated Advertise Blocker too). It seems to be full functionable with GC or IE.


@Herrfugbaum

The easiest way to enter the correct stats is to import your armory via chardev.org, reset talents and add flask&bufffood (because it isn't implemented yet). You can identify the green AP value via RangedAP-Agility.

This is what I get for your gear:

SpellHit | 3
CritRating | 1,97
Haste | 1,78

Therefore you should go for Str/Hit, though I'm not sure if the Sim is right in here, to confirm it just try two different gear setups: one with crit ametrines, one with hit ametrines, and choose the more powerful DPS.

P.S.: Everytime you change gear, the EP values change. Because of this, normally one can't determine universal EP values. Especially with Frost I recommend to generate the EP values on your own.

The Cata simulator is giving me wildly inaccurate results for Unholy DPS (I have not tested frost yet).

Here is my character.
The World of Warcraft Armory - Elretardo @ Stonemaul - Profile

I'm using more or less the standard unholy spec, only difference is I dropped the 1 point in butchery so I could have my beloved chain death grip in PvP.

On the target dummy, self buffed only, my DPS ranges from 9100 to 9500, yet the sim predicts only ~5500:

Ability	Damage done				hits			Crits			Misses		Glances			TPS	Uptime
	Total	%	#	Avg	#	%	Avg	#	%	Avg	#	Avg	#	Avg	%	%
ScourgeStrike	767578150	19	175677	4369	125359	3398	68	50318	6790	27	9738	5	0	0	0	0	0
MainHand	667331709	16	283486	3151	125413	1916	42	86352	3832	29	15738	5	71721	1341	24	0	0
Ghoul	560169107	14	553956	1330	260672	833	47	160395	1658	29	0	0	132889	582	24	0	0
DeathCoil	464611101	11	124841	3722	91061	2930	73	33780	5856	27	0	0	0	0	0	0	0
ScourgeStrike+ScourgeStrikeMagical	298393874	7	175677	1698	175677	1698	100	0	0	0	0	0	0	0	0	0	0
BloodPlague	294013565	7	240171	1224	175278	965	73	64893	1925	27	0	0	0	0	0	0	99
FrostFever	293409729	7	239873	1223	175180	964	73	64693	1926	27	0	0	0	0	0	0	100
Ghoul: Claw	226660998	6	186047	1218	132267	1040	71	53780	1658	29	0	0	0	0	0	0	0
BloodStrike	216654812	5	68003	3186	48050	2465	67	19953	4923	28	3790	5	0	0	0	0	0
PlagueStrike	91965950	2	24466	3759	17365	2910	67	7101	5836	28	1381	5	0	0	0	0	0
IcyTouch	75844816	2	24480	3098	17738	2428	72	6742	4863	28	0	0	0	0	0	0	0
Gargoyle	65744325	2	22019	2986	22019	2986	100	0	0	0	0	0	0	0	0	0	0
UnholyBlight	46483428	1	124841	372	124841	0	100	0	0	0	0	0	0	0	0	0	0
Army of the Dead	41556491	1	491040	85	466466	85	95	24574	80	5	0	0	0	0	0	0	0

DPS 5597(+/- 814)
Total Damage 4110.42m
RunicPower Total runic power used: 2876202 (2 wasted)
Threat Per Second 3786
Generated in 37.13 s
Template uh80
Priority Unholy
Presence Unholy
Sigil
RuneEnchant
Pet Calculation True


Today I did a 25N Saurfang as a benchmark for raid DPS. In this raid there was no windfury, no bloodlust, and my pet was dead for the last several seconds. But my DPS was about 15000. However, the sim predicts only about 11k, even with full raid buffs and debuffs (and yes, with the ICC buff set at 30):

WoL Parse: World of Logs - Real Time Raid Analysis

Fully buffed simulator results for fight of equal duration:

Ability	Damage done				hits			Crits			Misses		Glances			TPS	Uptime
	Total	%	#	Avg	#	%	Avg	#	%	Avg	#	Avg	#	Avg	%	%
MainHand	1530115054	19	360447	5684	119272	3168	31	149907	6337	39	20403	5	91268	2218	24	0	0
ScourgeStrike	1367570637	17	179661	7612	109764	5482	58	69897	10956	37	10153	5	0	0	0	0	0
Ghoul	1334724404	17	705697	2490	258326	1436	37	277729	2858	39	678	0	169642	1002	24	0	0
DeathCoil	799501778	10	134150	5960	84562	4352	62	49588	8702	36	2398	2	0	0	0	0	0
ScourgeStrike+ScourgeStrikeMagical	547666450	7	179661	3048	179661	3048	100	0	0	0	0	0	0	0	0	0	0
BloodPlague	500886045	6	240322	2084	151535	1523	63	88787	3042	37	0	0	0	0	0	0	99
FrostFever	500100513	6	240256	2082	151640	1522	63	88616	3039	37	0	0	0	0	0	0	100
Ghoul: Claw	412345018	5	185884	2218	112949	1799	61	72935	2868	39	172	0	0	0	0	0	0
BloodStrike	380949200	5	69178	5507	41539	3938	57	27639	7864	38	3942	5	0	0	0	0	0
Gargoyle	190167572	2	47212	4028	42453	3659	88	4759	7322	10	851	2	0	0	0	0	0
PlagueStrike	160413305	2	24456	6559	14919	4713	58	9537	9448	37	1409	5	0	0	0	0	0
IcyTouch	124053957	2	24480	5068	15233	3676	61	9247	7360	37	432	2	0	0	0	0	0
UnholyBlight	79927138	1	134150	596	134150	0	100	0	0	0	0	0	0	0	0	0	0
Army of the Dead	60785964	1	792968	77	713669	76	90	79299	86	10	760	0	0	0	0	0	0

DPS 10879(+/- 3331)
Total Damage 7989.21m
RunicPower Total runic power used: 2928590 (8 wasted)
Threat Per Second 7012
Generated in 37.24 s
Template uh80
Priority Unholy
Presence Unholy
Sigil
RuneEnchant
Pet Calculation True

I have the simulator configured to use unholy presence, the default Unholy.xml priority, NoIntro, and the default Scenario.xml.

Are other people seeing better closure between the sim and target dummy and raid parses?? Are these results due to a bug on my side or is this kind of behavior reproducible. Since people are already using this sim to produce tentative statweights and draw conclusions about different specs, I assume someone somewhere has some reasonable validation data...



Also, the ranges on the simulation results, especially the +/- 3000 in the fully buffed case, seem absurd. I suppose it is possible that one simulation trial maybe got 20 dodges in a row or something ridiculous like that, causing a low DPS number, and maybe another trial got runic corruption procs on every DC and/or a long crit streak...but if this is the case then the RNG range of the reported DPS really ought to be expressed as a standard deviation or variance statistic (or better yet, a histogram), rather than a simple min/max range. The way things presently are, it is impossible to get a meaningful idea of what kind of DPS variations can be expected in practice, which makes it impossible to determine whether or not a DPS delta is statistically significant (e.g. when evaluating stat/gear/spec changes)

If you're getting a 3k damage range, I can only assume you're doing something wrong. I can't find a way to get a gap that big, with any level of gear or with any spec. Generally speaking, the range is +- 10%, which is perfectly logical for RNG variance.

At any rate, the simulator does appear to be currently calculating the magical aspect of SS improperly... but after some testing, I can't find any rhyme or reason to it. Most of the time, it the magical portion hits for 40% of the physical portion. That's nothing more than .36 * 1.3 * 1.08, and that's what the sim assumes. Sometimes, however, it hits for seemingly random amounts like 63% - and it's not some odd anomaly, as it happens fairly often. Analyzing combat logs, I can't seem to link it to any sort of specific buff or what have you.

I'm at a loss as to what's going on there. Something strange.

Ghoul damage is absurdly off the mark; way too low. Likely something to do with how SI/DT are applied, but who knows. It should be over 50% higher than it is.

Aside from that, some small issues; ghoul scales with crit when it still doesn't (likely a bug), FF crits for 2x when it actually does 1.5x (likely a bug), and death coil's damage is too low (beats me why).

I'm sure the kinks will be worked out by Kahorie soon enough, and I can all but guarantee even once this is all fixed the order of stats won't actually change, anyways. How Unholy appears to perform compared to Frost may be currently misrepresented because of all this, but with the current Obliterate bug, that hardly even matters!

Thanks for pointing errors. It's easier for me to look into the thousand code lines where there could be a problem.

I don't have time to make a hot fix cause it's late here, but I have found that the coef of DC is 0.15xAP in the sim instead of 0.3. I don't see anything wrong with SS except that the overall damage is low. And it's maybe the default priority that make Festering Strike not cast enough. Maybe putting it higher in the prio would make the trick as it won't use a death rune, just pure blood/frost.

I have also to check in game if the AP/Str coef are correct for the ghoul. Your WWS claw is making 1/2 of your ghoul damage where mine is 1/3.
If someone can check what the AP to STR coef, base AP/Str of the ghoul, I could correct that tomorrow or it will wait until this weekend.

It gets 2 AP per 1 Strength, still, and gets 1.4x of our strength with the glyph. A simple means of testing is Horn of Winter; it adds 506 AP which, based on the fact that it gives us 155*1.15=178.25 str, almost perfectly fits the simple formula of 1.4str * 2.

Claw deals exactly 1.25 times the average auto-attack (and 1.5 times when under the effects of Dark Transformation, of course). Looking at the above sims, the average Claw is way, way below what it should be based on the average (non-glancing) auto-attack. Not sure how it's being calculated, but something's up there.

Pure speculation regarding scourge strike: The 62% SS magical is 40% expected x 1.5, which would indicate a sort of undocumented crit or mastery based random bonus. One option would be the Glyph of Scourge Strike - Spell - World of Warcraft. In the effect part of the spell there is a damage modifier and "Unknown Effect (Effect #0) Value: 9" where the Glyph of Obliterate - Spell - World of Warcraft just have the damage modifier.

A quick sim of my character showed the following:
I notice a few things that don't fit what I know about DKs here.
1) Hit Rating is given a higher value on the 'Hit Capped' scale than the non-capped scale.
2) Haste rating has 0 value. (Edit: whoops, make that no haste value in the scale-format output. The long-form feedback clearly lists one. I've been staring at numbers too long.)
3) Hit and Expertise are worth considerably less than Strength.


I hope this can help get things tuned in accuracy-wise. And if it's off because of some option I missed, feel free to let me know.
Character - Seraphel, Stonemaul US (I know my gear's shabby, I just started this DPS set.)
(Edit: Wow, I don't know what's up with Armory's version of me. Here's a Wowhead profile.
Profiler - Wowhead)