Waiting is averaged. Since it is extremely close to 0 most likely you are never really waiting in the "normal" rotation (being at the gcd cap, so the unholy switch is a huge benefit) and you only have short waiting times in case you get very unlucky with procs.
Being at the gcd cap 90% of the time and only waiting 10% of the time (numbers randomly chosen to illustrate the point, not to be taken literally) should (and apparently does) put Unholy Presence ahead of Frost Presence.

I double checked this and this was another bug reintroduced with the Cata revamp. Fixing rime procs generating runic power from rune abilities puts Frost Presence back ahead of Unholy Presence by about 300 dps for our 372 DW profile.

I've fixed the bug in SVN and it will be out in the next release. Unfortunately, we just released two hours ago, so I'm not sure how long it'll take before we release again.


[e] Running some sims and scale factors, prioritizing haste as your secondary stat, even with FP, results in more DPS than stacking crit or mastery.

Quite the opposite. On average you are never waiting on rune cooldowns (that's the wait time, right? or am I reading that completely wrong?) which means you are pretty much always at the gcd cap and wasting resources.

FP vs UP is a fine balance. At its base level, the 15% damage increase from FP is superior to the 10% haste from UP - as the haste bonus is both a smaller percent and affects a smaller overall number of our attacks.

The deciding factor is therefore how over the GCD cap you are discounting the benefits of both of these presences. If the .5s reduction to the GCD cap allows you to use a number of attacks greater than the difference between 15% damage/10% RP and 10% haste/4% RP, then UP will become superior.

This is why such small differences between the amount of RP generated and even the GoHB/Outbreak rune bug can cause a swing in one direction or the other.

It's hard to quantify, but I felt that the biggest benefit of being out of FP was not having the extra RP from the presence. Too often have I found myself forced to use FS while in FP in order to avoid wasting RP, and frequently this has resulted in either wasted RE procs (if there were no pairs on CD, or if one of the pair was <1.5sec) or wasted KM procs. When in FP this happens a lot less; firstly the higher ambience of Rime means less chance for the dreaded Dry Streak, and secondly when all your runes are gone, using up all your RP takes a lot less time, which in turn makes it less likely that KM proc will go to a FS. The faster GCD of FS is also much more efficient for RE procs: You gain more value if you gain the extra runes faster, as it drastically reduces the chances you'll have to sit on a Rime proc, or re-proc KM.

And of course it goes without saying that on mobility fights the 15% run is a tangible advantage.

My impression on UP vs FP is that at the low tiers (iLvl 359) you gain more from UP than from FP. Only by switching from FP to UP I made 1k dps more against Argaloth although I was unlucky at procs. This might change with iLvl 372 but for my gear UP feels better.

Sadly the latest fixes are yet to be released at SimC (I guess it's 406-13). Can some devs run an iLvl 359 comparison? Hope this here one works out: chardev 8 - Cataclysm Beta

My impression is that with high level gear UP will outperform FP due to the more haste so waiting time will be lower than now.
According to the simcraft release 12 scale factor, I reforged all to haste and hit till spell cap (mastery is still our worse stat it seems) and I had better results.

Hello everyone,

That's quite some time since I posted something on ej, probably more than a year, but I just recently began to play wow again with my death knight.
Simulationcraft is a fantastic tool in theory, but to be honest, it really lacks some basic ingame number validation. I perfectly understand that the devs do not have the time to do it, so it would be great if the community could handle it in their place.

Here's a little sample feedback, knowing that I always build simcraft from the latest svn with the ui.

Issue : Obliterate damage is completely off of what it's really doing in game (only tested for frost dps).

Way obliterate damage is calculated in game :
(((AP/14*3.3 + weapon_damage) * 1.6 + 1040) * (1+(annihilation*0.15 + MotFW*0.04 + GoOB*0.2))) * (1+number of disease*0.125)
fully talented with OB glyph, obliterate damage without disease = (normalized_weapon_damage * 1.6 + 1040) * 1.77

a sample of the checks I did :
AP = 9935
Weapon = large club 4-6 dmg 3.2 speed
Obliterate damage witout any disease ((9935/14*3.3+5)*1.6+1040)*1.77 * 0.7012 ~= 5951
0.7012 is the damage reduction via armor on a lvl 85 dummy target
In game average number = 5954
deviation < 0.05%

Way the bugged obliterate tooltip is calculated in game, fully talented with OB glyph :
160*1.77 = 283% + 1040*1.77*1.77 = 3258,216 -> Yes there is a "double dip" in the tooltip
IG : 283% + 3257

Way obliterate is currently calculated in the simulator without disease :
((AP/14*3.3 + weapon_damage) * 1.6 + 1040) * (1+annihilation*0.15) * (1+MotFW*0.04) * (1+GoOB*0.2)
fully talented with OB glyph, obliterate damage = (normalized_weapon_damage * 1.6 + 1040) * 1,9488

So the simulated OB does 10.1% more damage than what it should, which is quite a lot when the goal is to have an accurate simulation to be able to determine the optimal abilities priority.
I checked that it really did 10.1% more damage by putting my gear in the simulator, with thoses lines in addition :

Notes :
- no Merciless combat talent to have perfectly stable damage
- no str/ap proc items
- only obliterates without disease

average simulated obliterate hit : 6525
the stats tab shows 9900 AP. -> ((9900/14*3.3+5)*1.6+1040)*1.9488*0.7012 ~= 6534


Issue : Blood Strike seems to do less damage than what it should (only tested for frost dps).

This issue is a little strange, and is probably linked to a blizzard bug.
in game blood strike without disease = 2010
in game blood strike with 1 diseases = 2387 = 1,1875 * 2010
in game blood strike with 2 diseases = 2875 = 1,204 * 2387 = 1,430 * 2010
It should be +12.5% for each additional disease, but it's clearly not the case here.
I didn't take the time to determine what's the (bugged ?) in game formula, but whatever it be, simcraft does not have the same one, and from the tests I did, underestimate the blood strike damage as expected.


On another note, here's a line to add in the action list to gain a little more dps :
actions+=/blood_tap,if=buff.killing_machine.react&((frost>=1&unholy=0&death=0)|(frost=0&unholy> =1&death=0)|(frost=0&unholy=0&death=1))
before :
actions+=/obliterate,if=buff.killing_machine.react
it justs blood tap only if there is a KM proc, to do an obliterate crit instead of a frost strike crit. The gain is nearly inexistant on the live, but present on the PTR. I added it because it's probably the way the majority of the players use their bloodtap.

And here is a feature request :
Make cooldown._rune_.remains works, that would be great to know if waiting some runes can be a dps gain, like the wait command in the paladin dps rotation with crusader strike.

That's all for now, because it takes quite some time to check things talent point per talent point in game

Thanks for the feedback, I'll get Obliterate changed shortly. None of us play a DK, so testing DK mechanics at 85 is challenging.

For your Blood Strike testing, it'd be useful to know what your spec was, so we could see if there's some weird interaction going on with another talent, mainly MotFW, or if diseases are affecting something really weird like the AP coefficient, instead of using a modifier of 18.8 for one disease and 21.5 for two diseases.

[e] The Obliterate change is in SVN. I'll do BS once we're sure on how the formula is being calculated.

Were you taking brittle bones into account? If your single disease test was with frost fever there would be the additional 4% damage taken, as well as with double diseases.

I removed Blood Tap from the PTR action list, as I didn't think it'd work anymore, since Frost no longer has Blood runes. If it it still works to activate a death rune ASAP on the PTR then I'll add it back into the action list. If it no longer works, then we'll make the appropriate changes to the sim.

The current action list were built around the priority listed in the OP to try and match other values in the OP. If your new actions prove to be a DPS increase, I think we're safe now with the amount of bugs we've fixed in the past few weeks to switch to using them. I'll test your action list changes and update them accordingly.

Regarding Blood Strike, post any and all data you can if you want help figuring out the formula. AP, Talents, if using FF always gives an X modifier and BP gives a Y modifier, etc.