@Foxx: Thinking in terms of GCD rotations rather then strict rune rotations is exactly why I had an issue with calling FS filler. If you are given a choice of sacrificing an underperforming rune for no damage or no use, in exchange for getting another FS or Rime HB in, you should. Both the Glyph of Disease and Glyph of Howling Blast theories are based on this thought process.

@ Astallion: To properly simulate an HB rotation, what you really want/need is a system for having a priority element withing a rotation. So, for example:

(1) <OB></OB>

(2)<Priority>
<Rime RPless=35></Rime>
<FS>
</Priority>

(3)<OB></OB>

(4)<Priority>
<Rime></Rime>
<FF refresh=10></FF>
<BS></BS>
</Priority>

(5)<BS></BS>
(6)<FS></FS>


(7)<OB></OB>

(8)<Priority>
<Rime RPless=35></Rime>
<FS>
</Priority>

(9)<OB></OB>

(10)<Priority>
<Rime></Rime>
<FF refresh=10></FF>
<BS></BS>
</Priority>

(11)<PS></PS>
(12)<FS></FS>

Where the RPless is a conditional if RP is less then the number, and the refresh sets the FF refresh to any number below 10 seconds. Works like a Rotation, but uses priority based logic for certain elements. To make it more clear, I tagged each step in the rotation with a number.

@Everyone discussing the HB glyph: I urge you to reread my posts from earlier in the thread, as everything I did was painstakingly thought out. There is nothing about that proposed rotation that is not designed to maximize damage while not getting too complicated to make execution impossible. If you have any specific questions on something because it isnt properly explained, I would love to address them if they are addressed to me.

@Odii: I was thinking of doing almost exactly the same thing when I first started researching Glyph of Disease (the Rotation/Priority system) - I ran into some problems when trying to figure out how to handle varying latencies and missed attacks (leading to differing amounts of GCDs per Rune Refresh) though, but I may look into coding a system like that, I think it could have more uses.

@Doc: Yeah, I've seen some very odd looking set EP values from people using the latest versions (I think) of the sim. Try just switching set bonuses on and off one at a time and calculate EP values from those.

Those two sets of gear seem to have relatively low amounts of armor penetration considering it's the second best stat after hit and expertise soft caps are reached, or at least using the stat weights I've seen so far in this thread. I thought we'd be looking at something a little bit closer to a blood BiS set. Am I missing something?

I have tried and tried and I cannot replicate 7.2k+ DPS frost builds. I've changed priorities, use rotations, changed specs, and everytime I come up with a total that is short of what is being posted here.

What am I missing here? I figure Foxx2405 was using the wrong DPS for his weapons since he thought it was 200 DPS weapons. A casual glance at the post history here shows StatSet2 has had 180 DPS weapons from the start which was my memory as well. But I can't figure out Doc's numbers at all. Doc, could you show us the priority and stats you used so we can duplicate your work? As of right now, I can't duplicate the high end Frost build results.

@dr_AllCOM3: As for calling DW Unholy dead, that has yet to be determined. If you don't like Unholy DW, that's your call. But until I see some independent verification of your numbers, I am dubious of your results. Somehow producing 500+ DPS over what anyone else has been achieving is quite a feat and I'm curious about how you did it using the same statset that we have been using. If you can show us some data so we can duplicate your results, then indeed DW Unholy is dead for now.

Dead, since when? Last i herd it was competing with the most complicated pure dps frost builds.

Sorry about the long quote earlier. I do have to agree. I have spent long hours on the ptr testing many different deep unholy DW specs versus deep frost dw specs and the deep frost keeps coming up on top, namely the spec I linked earlier (0/53/18). My tests have shown that this spec is competitive with 3.2 Blood 2H dps on single target fights, and, of course, better on multiple targets. I haven't tested 2H Unholy, mostly because I don't think it is any good with higher level gear.

Also, dr_AllCOM3, are you ever planning to update your spreadsheet again, maybe with a DW spec that DOESN'T involve getting 2H Weapon specialization. I PM'd you about it and never got a response.

I am a bit confused with Priority systems, i mean i understand the general idea, but how would you work something like that on live?

Frost Fever
Blood Plague
KM-Rime -> So dont use rime unless KM procs, or if KM procs, only use obliterates until a rime procs?
Obliterate
Blood Strike
Frost Strike -> Wont this allow to much Runic Power to build up?

Something like this seems very complex to monitor in the actual live server, putting aside simulators.

So, would that translate to a rotation on Live, Or is that just the best option so far?

Basically, the priority means you use the first thing on the list that is "available" - if a disease isn't up, apply it, FF before BP if neither are up. Then, if KM and Rime are both up, use HB, otherwise move down the list and use the abilities you have runes for - Obliterate taking priority over Blood Strike. If nothing of the above works, use FS. And basically you would run through this list for every GCD, but obviously you can predict when things will/might happen. Roughly you'd get the rotation:
IT-PS-OB-BS-BS-FS-FS
OB-OB-FS-OB-FS
And weaving in HB if both procs are up, but not delaying IT/PS to do that.

Here's 0/18/53 with Awareness Glyph using most recent simulator build:

Ability Total % Landed Hit% Crit% Miss% Average
Obliterate 2134681289 15 392662 57 42 0 5436
Plague Strike 178171910 1 98174 51 48 0 1814
Icy Touch 282915059 2 98170 66 32 0 2881
Blood Strike 405902175 2 196335 56 43 0 2067
Death Coil 1700332892 12 374112 66 32 0 4544
UB 340117700 2 374112 100 0 0 909
Frost Fever 762547891 5 599373 100 0 0 1272
Blood Plague 762232134 5 598592 100 0 0 1273
Necrosis 660317382 4 1668930 100 0 0 395
Blood Caked Blade 468293718 3 500022 100 0 0 936
Wandering Plague 642856754 4 505097 99 0 0 1272
Main Hand 2096493771 14 834697 41 42 15 2511
Off Hand 1206528354 8 834233 41 42 15 1446
Ghoul 1808887786 12 2006451 86 13 0 901
Gargoyle 639135197 4 184223 86 13 0 3469
DPS 7116
Total Damage 14089414012 in 550 h
Threat Per Second 4743
Generated in 1040s
Template :OB001853.xml(C:\Users\Ben\Desktop\DKSimulator0.9.7\Templates\OB001853.xml)
Priority :Oblit001853.xml(C:\Users\Ben\Desktop\DKSimulator0.9.7\Priority\Oblit001853.xml)
Presence :Blood
Sigil of :Awareness
RuneForge :FallenCrusader / FallenCrusader
Pet Calculation :True

You can see the difference. Awareness is just a better glyph for 0/18/53. I still have not been able to replicate a Frost build that was over 7.2k DPS. As far as I can tell, Frost builds and Unholy builds are neck and neck and it's just a matter of choice at this point. The 4pcT9 set bonus heavily favors a Unholy build but Frost scales better so it should be interesting to see what wins the horse race in the end.

Interesting thing about the 4pcT9 set bonus is the way Wandering Plague works. Wandering Plague just copies disease damage by 100% so if a disease gets a critical tick when the Plague checks, Wandering Plague will use the critical damage amount. Better scaling for Wandering Plague is a good thing.

Edit: I used 200 ms latency and Astalion, please use the real simulator if you want to verify my spec. Your simulator has always had different numbers than the standard simulator.

Ability	Total	%	Landed	Hit%	Crit%	Miss%	Average
Obliterate	5306150000	36	959746	35	64	0	5528
Plague Strike	7811329	0	5796	54	45	0	1347
Icy Touch	8825345	0	2931	41	58	0	3011
Frost Strike	2817352905	19	778028	53	46	0	3621
Howling Blast	1024614527	7	104215	22	77	0	9831
Frost Fever	572093996	3	655668	100	0	0	872
Blood Plague	519484992	3	654447	100	0	0	793
Necrosis	599353231	4	1687157	100	0	0	355
Blood Caked Blade	243724198	1	336922	100	0	0	723
Main Hand	1901883934	13	843645	45	39	14	2254
Off Hand	1095596197	7	843512	45	39	14	1298
Ghoul	338414619	2	464194	87	12	0	729
Raz	17152970	0	1715297	100	0	0	10
DPS 7299
Threat Per Second 4309

Using Glyph of Disease, 3/51/17 setstat2, 200ms, 550h simulation. Replacing both Blood Strikes with Pestilence which would make keeping diseases up far more manageable, with this priority (added the <Pestilence> priority myself, it casts Pestilence when non-Death Blood Runes are up, no matter the circumstances - same way as Blood Strike is normally used):
Edit: Also, if I'm not mistaken, your test used 100ms? - Only way I can match your ability usage numbers and DPS anyway, and doing that with my frost spec I'm getting another 90 dps or so above what I posted.

Long time reader, first time poster.
I just thought about the GoDisease use. As someone mentioned before pestilence keeps updating disease that was first applied. In the beginning of a fight you get very high chances of proccing greatness, fallen crusader and maybe another trinket effect together, so if you reapply diseased with procs right away you get the strongest diseases you can get and may keep it up throughout the whole fight with pestilence. You may also use UA at this point and maybe a heroic potion to make effect even stronger and use ERW if you dont have unholy and frost runes ready and reset your rotation. I dont know if it's worth it, but looks like a good idea to me.

@Drakojin: Testing on dummies suggested that at least some of the buffs (only tested with some proc and the TS effect) stick to the disease after reapplications, which should imply that they all do.

TS is more a debuff than a proc. I can't recall the specific data for WL and SP dots, but debuffs work another way.

IF some procs (like FC) stick to diseases, then it is nice for now. But i can't imagine it to stay for long. Especially for unholy builds.

@Drakojin: TS doesn't work like a normal debuff though - if you for example only apply FF once, it will not get the TS bonus, it needs to be applied again when FF is already up to get the extra damage (same goes for RoR and BP, and is easily tested)

@Astalion: How on God's green and verdant Earth did you manage to get 5.5k Obliterates while DW? Or 3.6k Frost Strikes for that matter?

Ability Total % Landed Hit% Crit% Miss% Average
Obliterate 4451395484 31 415402 36 63 0 10715
Plague Strike 159603987 1 58066 53 46 0 2748
Icy Touch 195375613 1 59847 46 52 0 3264
Frost Strike 2668334762 18 387049 53 46 0 6894
Howling Blast 697732557 4 80054 27 71 0 8715
Blood Strike 573200851 4 142275 50 49 0 4028
Frost Fever 484593168 3 597748 100 0 0 810
Blood Plague 482477314 3 595051 100 0 0 810
Necrosis 626758568 4 1728975 100 0 0 362
Blood Caked Blade 251781634 1 346007 100 0 0 727
Main Hand 1989156742 14 863884 44 39 15 2302
Off Hand 1145571609 8 865091 44 40 15 1324
Ghoul 321386977 2 430208 86 13 0 747
DPS 7095
Total Damage 14047369266 in 550 h
Threat Per Second 4191
Generated in 1014s
Template W-03-51-17.xml(C:\Users\Ben\Desktop\DKSimulator0.9.7\Templates\DW-03-51-17.xml)
Priority :Frost.xml(C:\Users\Ben\Desktop\DKSimulator0.9.7\Priority\Frost.xml)
Presence :Blood
Sigil of :Awareness
RuneForge :FallenCrusader / FallenCrusader
Pet Calculation :True

Look at the huge differences between the average numbers here and the ones you show. I understand that Glyph of Disease isn't simulating as well as it should. But you really broke something or changed something. My numbers are way different than yours.

Edit: At last, I finally got a Frost Simulation over 7.2k.

Ability Total % Landed Hit% Crit% Miss% Average
Obliterate 4427169686 31 417128 36 63 0 10613
Plague Strike 152696151 1 56442 54 45 0 2705
Icy Touch 205757537 1 58272 46 52 0 3530
Frost Strike 2904462252 20 387264 52 47 0 7499
Howling Blast 760163130 5 80536 28 71 0 9438
Blood Strike 557115352 3 140252 51 48 0 3972
Frost Fever 522094452 3 597564 100 0 0 873
Blood Plague 473052445 3 594632 100 0 0 795
Necrosis 616679741 4 1729246 100 0 0 356
Blood Caked Blade 247713017 1 345535 100 0 0 716
Main Hand 1958029427 13 864509 44 39 15 2264
Off Hand 1126151100 7 864737 44 39 15 1302
Ghoul 316229671 2 432832 87 12 0 730
Raz 7780581 0 864509 100 0 0 9
DPS 7210
Total Damage 14275094542 in 550 h
Threat Per Second 4261
Generated in 1123s
Template W-03-51-17.xml(C:\Users\Ben\Desktop\DKSimulator0.9.7\Templates\DW-03-51-17.xml)
Priority :Frost.xml(C:\Users\Ben\Desktop\DKSimulator0.9.7\Priority\Frost.xml)
Presence :Blood
Sigil of :Awareness
RuneForge :Razorice / FallenCrusader
Pet Calculation :True

Astalion's priority and switching Blood Strike glyph to Disease did the trick. Used dual FC by mistake in the first one.

@Doc: DW Unholy is not dead, its actually fairly competative -

Ability Total % Landed Hit% Crit% Miss% Average
Plague Strike 330132511 2 211329 52 47 0 1562
Icy Touch 1696259460 12 684195 66 32 0 2479
Blood Strike 320903616 2 181890 58 41 0 1764
Death Coil 2922977285 20 535282 66 32 0 5460
UB 584700596 4 535282 100 0 0 1092
Frost Fever 298148258 2 272267 100 0 0 1095
Blood Plague 596262121 4 543942 100 0 0 1096
Necrosis 629747196 4 1863788 100 0 0 337
Blood Caked Blade 451544087 3 559472 100 0 0 807
Wandering Plague 356544200 2 325357 99 0 0 1095
Main Hand 1997572103 14 931440 42 41 15 2144
Off Hand 1150928073 8 932348 42 41 15 1234
Ghoul 1936104525 13 2148080 86 13 0 901
Gargoyle 775610563 5 223822 86 12 0 3465
DPS 7095
Total Damage 14047434594 in 550 h
Threat Per Second 4636
Generated in 915s
Template W-00-17-54.xml(C:\Documents and Settings\Keith\Desktop\DKSimulator0.9.6\DKSimulator0.9.6\Templates\DW-00-17-54.xml)
Priority ualUnholy.xml(C:\Documents and Settings\Keith\Desktop\DKSimulator0.9.6\DKSimulator0.9.6\Priority\DualUnholy.xml)
Presence :Unholy
Sigil of :VengefulHeart
RuneForge :FallenCrusader / FallenCrusader
Pet Calculation :True

Build: 0/17/54 Dual Wield

This is without set bonuses and at 200 ms latency. The test also uses statset2 which falls short of the spell hit cap, a disadvantage for a DC spam spec. 4pcT9 benefits Unholy the most due to Wandering Plague and Crypt Fever. It is also worth noting that DW Unholy's best specs provide a raid buff (Ebon Plague) while DW frost's best skips imp icy talons. Further sincet he sim assumes a patchwerk style fight it fails to factor in the movement bonus of unholy presence. Nor does it take into account that a DW Frost spec is highly unlikely to keep their ghoul alive for the full duration on quite a few fights. Finally the DW Frost spec has far more room for human error with the proc watching that needs to be done.

DW Unholy is still quite strong in both the DC Spam and Oblit builds. I'm tempted to test the SS build to see where that sits right now.

Edit: The old SS build is definately dead, but I rather expected that.


@Astralion: You're pulling more DPS than I've seen managed with 3/51/17. Were you by chance using set bonsues?

@Orlgin: I keep getting 0/53/18 to 7310 with both BS glyph and IT glyph, I'm new to the 3.2-sim so I might be off with something..? I also get the 0/18/53 build to ~7100 so unholy is far from dead and this gives you the opportunity to choose spec depending on what buff you need to provide.

@Kithus: The old spec is definitely dead but Odii got some odd experimental spec up to ~7000 it seems (although i can't sim it higher than 6750, I believe he used a 244 dpser which is higher ilvl than 180 dpsers). I also agree that frost demands more of the player to maximize your dps and it won't always be that easy to get the perfect KMRimes etc. in actual encounters.

I get 6523dps with your setup and 7329dps with mine, using the same items (Str heavy).
Your priority is more like the Unholy spam build. I tested that a few days ago. You need a very good ping (~50ms) to make it even competitive. Judging from my raid experience, super fast spam build don't work very good on live. Unholy has taken a hefty nerf in 3.2 anyway. Less aoe dmg due to UB and PvP related nerfs on SS. The best Unholy spec are the ones that don't use SS.

Edit:
4/5 T8 seems to be the best. The optimizer doesn't seem to take the set weightings into account. Add a hardmode item and stuff with Str+Exp/Hit/Arp, then you're good to go.

@doc

Can you clarify which unholy DW build you are getting ~6500 with? Multiple people have reported both DW builds in the OP to be simming roughly on par with frost builds. Are you sure your numbers aren't off somehow? Are you simming a build that uses SS? You also speak of Unholy getting a nerf, the only major nerf to my knowledge is SS, which neither build uses. The new unholy blight is an AOE loss, but it is also a huge single target buff. Again, not sure why we are considering this a nerf when single target dps is by far the most important factor. I would say that the addition of UB, and the change to gargoyle are significant buffs to Unholy.

If you are simming 0/17/54, and getting a result of 6500 with the new UB then something is off for sure. Without boring everyone with too much personal anecdotal evidence, I'll just say that this spec is capable of surpassing this figure on live without the addition of UB.

I used the one I quoted, it's in the first sentence. I'm quite sure my numbers aren't off.

Would you mind posting your sim report? Either your numbers are off, or Kithus's are. The two of you are saying that the exact same build, exact same everything is showing a 570 dps difference between simulator results. Needless to say, this is far beyond the realm of RNG differences.

Your numbers might be correct, but it would be nice to compare them to his and see where the problem is. Both simulator results can't be correct.

Char

Frost

Ability	Total	%	Landed	Hit%	Crit%	Miss%	Average
Obliterate	677665652	25	66640	44	55	0	10169
Plague Strike	46701368	1	16886	52	47	0	2765
Icy Touch	59290703	2	16922	42	57	0	3503
Frost Strike	587837177	22	76475	47	52	0	7686
Howling Blast	154740436	5	13943	0	100	0	11098
Blood Strike	125623689	4	33376	59	40	0	3763
Frost Fever	94817624	3	112026	100	0	0	846
Blood Plague	86088777	3	112196	100	0	0	767
Necrosis	113483706	4	322493	100	0	0	351
Blood Caked Blade	70514641	2	96599	100	0	0	729
Main Hand	360461838	13	158363	44	40	14	2276
Off Hand	206950925	7	164130	44	41	14	1260
Ghoul	52240983	1	82710	87	12	0	631
Raz	1641300	0	164130	100	0	0	10

DPS 7328
Total Damage 2638058819 in 100 h
Threat Per Second 5780
Generated in 104s
Template :Frost 0-52-19.xml
Priority :Frost.xml
Presence :Blood
Sigil of :Awareness
RuneForge :FallenCrusader / Razorice
Pet Calculation :True

Unholy DW

Ability	Total	%	Landed	Hit%	Crit%	Miss%	Average
Plague Strike	106882952	4	66604	50	49	0	1604
Icy Touch	300986698	12	125579	67	32	0	2396
Blood Strike	62720172	2	34515	56	43	0	1817
Death Coil	438685184	17	82529	67	32	0	5315
UB	87706059	3	82529	100	0	0	1062
Frost Fever	60791377	2	57592	100	0	0	1055
Blood Plague	92655830	3	87717	100	0	0	1056
Necrosis	114618932	4	345385	100	0	0	331
Blood Caked Blade	85506267	3	103581	100	0	0	825
Wandering Plague	63202153	2	59847	100	0	0	1056
Main Hand	363821042	14	169475	42	42	14	2146
Off Hand	209076979	8	175910	42	42	14	1188
Ghoul	308612863	12	399375	87	12	0	772
Gargoyle	151466040	6	45540	86	13	0	3326

DPS 6796
Total Damage 2446732548 in 100 h
Threat Per Second 4475
Generated in 128s
Template :Unholy DW.xml
Priority ualUnholy.xml
Presence :Unholy
Sigil of :VengefulHeart
RuneForge :FallenCrusader / FallenCrusader
Pet Calculation :True

Here's the requested data (150ms). I've even improved the DualUnholy priority a bit. A set with 4/5 T8 benefits Frost even more.