I've specifically asked in this thread for PM's or posts regarding builds for the OP, to date nobody has sent me a spec submission for the OP that comes anywhere close to the standards we strive for.

I'll say it again, If you have a spec you feel deserves to be on the OP, please either send me a PM, or post it with evidence. I'm not going to update the OP with untested or builds that have been tested on the targeting dummy for obvious reasons. In the absence of solid raid data the simulators are our best tool, we need to use it to make our case for why one build is superior to another.

We are closer than we were when we started, but at this point nobody has stepped out and said "This build, this priority/rotation, here is the simulator data to prove it." If someone thinks they are confident enough to state this, let me know. Until then, we should continue to work on individual pieces of the puzzle.

-Side note: Based on all available data it looks like OB heavy blood presence rotations are going to be superior, has anyone run the numbers assuming 100% crit on both hands with a FS via a KM proc? Nobody knows for sure if its going to be fixed, but I think its logical to at least test if it would be a game changer if it was a bug.

The numbers I have run still show obliterate heavy rotations come out ahead in the case you mention, though I am not convinced of the spec and glyphs completely yet.

However, there is something to it in another sense which I have mentioned before. If KM only crits on the MH then your priority for a KM proc should be a rime-procced HB>FS, however if they fix the bug then once again FS will always be appropriate for a KM proc.

My numbers suggest that KM always should be used for Rime procs if possible, mainly due to FS having a far higher crit chance (test run without KM/DC put FS at 52% crit and HB at 27% - note: that's including the T8 bonus and a relatively high amount of agility). HB should on average gain 4615 damage from using a Rime proc, while the benefit on an FS would be 2851 (using same gear as all my tests, math done on the abilities with KM/DC excluded, assuming KM/DC are meant to affect both main/off-hand FS).

@Spec post request: Been a bit lazy, got most of the post lined up in my head and hopefully I will get around to writing something (main focus would be on 3/51/17 and variants, including Glyph of Disease theory - could include some comparison to unholy specs, but I haven't been as in-depth in my research of those, and hence there might be some things I've missed that would improve the numbers there).

I've seen a lot of variations on the frost / unholy spec, so far. Most of them come down to deciding where to put the last few points.

The specs that float around are:
3/53/15
0/53/18
3/51/17
2/51/18


The talents in question are:
- Merciless Combat (correct me if im wrong but this is where you take points from in a 51 frost build right ?)
- Subversion
- Blood Caked Blades

I'd like to say which is better, but i have no simulator and all i can go by are raw imaginative numbers that dont help anything if 2 talents are close. I can take an educated guess at some though.

Merciless Combat gives 12% more damage, 1/3th of the fight on abilities that do about 60% of your total dps (this is my guess, i have no clue what the raw numbers are that).
That would mean 7.2% dps increase for 1/3th of the fight or 2.4% dps for 2 points = 1.2% dps per point.

BCB, i can't really say anything about that. Although someone in this thread mentioned BCB being 5.6% of the total dps for him. That would mean 1.8% per point (sounds reasonable).
That would already dismiss the 3/53/15 build.

Subversion. The 9% extra crit on blood strike is worth about 7% more damage if you have 30% crit.
The 9% crit on obliterate is worth about 6% more damage, since obliterate already has a higher crit rate (mostly due to rime).
Very very raw self made up numbers, say obliterate will be 30% of our dps and blood strike 10%. That would mean a total dps increase of ( 0.3 * 0.06 + 0.1 * 0.07 ) = 0.025 = 2.5%
That would be 0.83% dps per point.

That would make subversion the worst. But my numbers are totally made up. Would be nice if someone with a simulator or with some actual PTR parses ( NON-DUMMY ) would adjust them.

That would promote the 0/54/18 build though.

-------

Another item we really need to get straightened is:
- Is blood subspec definitely inferior ?

My guts say yes as you trade Necrosis and BCB (apparently worth a solid 10%+ dps) for Bladed Armor and 5% crit (which is somewhat diminished due to frosts already high critrate)

EDIT: Just noted that the minimal amount of points in frost is 52 unless you're pulling out 1 point on an important talent.
Assuming you take IIT (we can discuss builds without IIT and glyph of diseases, but a lot of raiders are just expected to bring the raid buff), are you stuck with 52 points or can you take a point out of a talent ? (Most seem to vital to miss to me).

The builds would change slightly but the talents in question are still the same.

That's fairly accurate based on my 3.1.3 parses, BCB does 5-6% of my total damage on any given fight. One thing to remember is with the removal of normalization two things happen. First, fast weapons get a nerf in BCB damage, which honestly is no big deal since ToT builds are going to be Slow/Slow. The real gem is that Slow weapons are going to be receiving a buff, I can easily see BCB doing 6% or more of the total damage in 3.2, making it at or above 2% dps per point.


1h normalization is set at 2.4, so any weapon with a speed slower than that will produce more damage via BcB in 3.2.

@Foxx in general:

The issues you pointed out are (in my opinion) exactly what is holding folks back from doing quality spec posts. Most people can look at the frost tree and easily pick out the core talents, its the subspec that is the problem.

What we need is to find the per point dps value of each talent we can possibly get in the subspec, IE once all core frost talents are filled out. Its looking like they are going to be close enough that only math is going to be able to show the superior subspec choice.

To me this seems to be the only logical place to move from here, I'll start working on this myself after work, but anyone with strong math and simulator skills should really chime in.

I mentioned this before and people agreed but then it died down. If we want to try and find the best build, we need to have some set stats that each build should be tested with. This will prevent any differentiating data based on stats. Some possibly Stats to use are:

Stat set 1: Stat set 2: Stat set 3: Some standard for stats would be good. When posting a spec, it would also be beneficial if Stat weights were posted as well as the DPS/EPAP. How you do this is when it calculates the EPAP of AP, you take the difference between the DPS of the Dry Run and the Attack Power Run, aka Attack Power Run - Dry Run = ????. This should give you the DPS per amount of AP you have the EP calc run. If you don't change any of the EP Calc settings, it should be normally 50. You then take the DPS difference and divide that by the EP change being used, i.e. ????/50 = DPS/EPAP. This will help when determining the Scaling of a build.

REMINDER - These are just suggested guidelines, the stat idea hasn't been set in stone, but the EPAP + DPS/EPAP is very helpful.

As for viable builds, I'll start doing some sims for 3/51/15+2 builds, most likely something like 3/53/15 for testing.

Its not quite that simple.

First you need to compare 100% crit HB vs 100% crit FS mianhand +Base crit rate off hand, and come up with which is more damage.

If you are using a 4 OB, 4ish FS, Blood Strike Glyph rotation, then you only want to use Rime procs when you have KM. If KM HB is more damage then KM Frost Strike, then you should use HB every time Rime and KM are up at the same time. If KM FS is more damage, then you only want to use HB when you have Rime, KM and if you have low* RP.

If you are using 5 OB 4ish FS, 1ish HB with Glyph of Disease, then you cant really afford to be very cute with Rime procs, use em when you get em. In that case, you should only delay using a Rime HB if you dont have KM and you do have a high* amount of RP built up.

*What constitutes high and low RP is an interesting question, that is a good deal harder to figure out.

Wow that sounds really low. Are you sure its counting the offhand as well ?

With my 2H unholy build BCB usually is about 3.5% of my dps.

Basically BCB (correct me if im wrong here) is a 50% weapon damage strike every other 3 auto attacks (average).
So basically its 1/6th of your auto attack as extra damage.

If its less than 2% your white dps is less than 12%. Sounds a bit low. (Or is there an internal CD? I remember something of that kind.)

If your sim is right though, we might just as well skip BCB because 0.7% dps per point is a waste

@Foxx: My numbers were a bit outdated (might have been faulty notes as well, possibly just 2/3 BCB), but I rechecked the code and new tests show BCB at about 2.45% of the total damage with 3p.
@Odii: Yeah, I've only looked into the details a little bit, but there should quite often be times when you don't risk overcapping RP by using Rime, and generally you'll have time to wait a few GCDs before using the Rime proc if you don't have KM up at the time (looking at the Disease rotation at least). I'll see if I can get some better numbers for RP thresholds tho-

I have already included a parse from Live that shows contributions from BcB to be 5.3 percent on a non-gimmick single target Patchwerk-esque fight while DW.

World of Logs - Real Time Raid Analysis

Your testing appears flawed, how are you trying to measure BcB contributions? You should not be using dummy testing for anything other than rotations, especially when you are posting data to this thread. You confuse the audience.

@Octopi: I'm using simulations, updated for the new BCB mechanics (non-normalised), as far as I can see the code is all correct (I replaced the function call that referred to normalised MH/OH damage with a call to the function for non-normalised, the one used for autoattacks - note that other multipliers, such as the one for diseases, are still being applied)

Some things I noted from a quick look at your parse: You're showing 31.3% autoattack damage (mine is closer to 20%, so yours is about 50% higher), and you also have Crypt Fever (an additional disease, 50% more BCB damage) That would account for about double (should even be slightly more) the BCB damage. Basically - our tests show (more or less) consistent results when all factors are taken into account.

My question is in the 3/51/17 builds, or the variations where we are trying to find 2 points to cut from the build, is it possible to drop 2 points from Black Ice? Seems like BcB is still running about 1.2-1.8% DPS per point, Merciless Combat is a 12% increase for 35% of a fight on about 70% of your damage[IT+FS+HB+OB], so 1.47% per point, and then each point in black ice is 2% increase in (I lost most of my parse screenshots when i reinstalled stupid PTR) [FS + FF + BP + IT], so to match merciless combat or bcb per point those 4 abilities combined need to be 73.5% of your overall damage? Or am I doing the math totally wrong here.

@xtremers: While Black Ice isn't an incredibly strong talent, it still seems to be stronger than Killing Machine (might be able to push the value of the latter up slightly by more intelligent usage, but I doubt it'll be enough). Edit: Thanks Kyruski, another thing to correct for the simulations then.

@Foxx: I'm using a (self-)modified version of Kahorie's, I think it's slightly more accurate/up to date, but I haven't checked if he did any unannounced bugfixes.
Anyway, my autoattack numbers seem to be generally quite low for a DW build (~20% white damage), but compared to Kahorie's simulator I fixed a bug that would apply TS bonus damage to autoattacks, which would make a fairly large diffference compared to the numbers others are getting.

Actually, Necrosis was changed to be a flat 4/8/12/16/20% of your White damage that cannot be modified by talents or abilities.

Your parse is flawed. You were not using a heavy frost build and white melee makes up 30% of your damage. Obviously BCB is going to be a higher percentage of your damage if you are using a build that favors white melee hits. You arnt even using heavy hitting attacks like Obliterate that would suck up more of the percentage of damage for that fight.

Show me a parse of any of ?/51/? with a heavy obliterate rotation and get BCB that high and then you will have something. For me BCB always has a lower percentage of damage than necrosis and necrosis normally comes in at 4.1-3%.

Edit: He was posting his heavy unholy BCB damage in response to discussion about frost BCB damage. These are conflicting ideas. His post proves nothing about how well BCB preforms for a heavy frost build.

The parse isn't "flawed" at all, its normal for 0/17/54. Remember everyone, this is a 3.2. DW thread, not specifically a frost DW thread.

The important thing to realize is we actually have two competitive builds. Obviously, the "Core" frost build is going to be good, however, 0/17/54 DC spam is still holding up quite well on the simulator. Its clear that the deep frost build is the shiny new car, most people are focusing on it, and rightly so but we need to remember that it isn't the only build that needs to be scrutinized.

Make sure when discussing DPS value per point, you keep it clear which deep build you are refering to, frost or unholy.

This is the Dual Wield thread. Not the only do the frost thread. Or only the Unholy thread. Everything dual wield that's not retarded should be discussed.

@ Travaggie

This will remain a thread to discuss the best possible builds for DW. Without resorting to arguing, I'll simply say that several of your notions regarding unholy DW are categorically false. The OP of this thread will be updated in due time to reflect any DW build that can prove itself to be competitive, regardless of which tree the build is heavily invested in.

The idea that we should disregard unholy DW even if it does more dps than frost DW for any reason is absurd. There are just as many guilds that rely on the deathknight to bring Ebon Plaguebringer as Imp Icy touch. I don't know about you, but I feel the OP should have as much diversity as possible, while still only including quality, competitive PVE specs.

All I said about DW unholy is that is is inferior in pure DPS to 2H Unholy, which is true. My point is that it makes no sense to compare Frost DW to Unholy DW in the first place for practical raiding purposes if the end goal is to find the most OPTIMAL and VIABLE build, considering they bring different buffs. There is no Frost DW thread and there is a DW Unholy thread, so naturally people will and are using this thread to speculate the DW Frost builds and to assume people are always comparing or even taking into consideration the Unholy suboptimal counterpart is just dumb. So as I said, if you're only goal is to DW regardless of viability or optimal raid DPS, then continue to debate this argument.

Edit: To clarify, I don't see an issue in discussing the two builds seperately, although I personally don't see the reason behind DW Unholy to begin with. The post I was responding to was COMPARING as well as assuming that people would know he was referring to Unholy DW without siting it, which there just is no way to compare the two logically, based on what I wrote above.

On a related/slightly unrelated topic, I'm curious to see if anyone has come up with AEP values, or if anyone actually knows how to do this.

I haven't really tried to sit down and figure it out.

I'm going to go 15/53/3, personally. Don't care if it's the maximum DPS. I enjoy how it plays, and the DPS it puts out is more than substantial for me. So, how do I figure out what the AEP values are for it? And, if someone already has, can you post them for that spec?

The Current DW Unholy thread is run by me and it is for live now. Once 3.1 goes live, I will be asking a mod to close it since quite a bit is changing in 3.2. There has been very little to no theorycrafting in the other thread. The reason why there is no Frost DW Thread is because there was never a real need for one since the original Dual Wield Thread was like this, it was for all Dual Wield specs.

@Ashen: There is a function in the simulator for EPAP Calculation.

People are saying that DW unholy is inferior to 2h unholy in 3.2, as if it's a well-known fact. The 2h unholy build is taking a SS nerf (0/17/54 doesn't use it), and the new UB should benefit DW unholy more than 2h, since it throws out considerably more DCs. I don't believe anyone has done any real math or sims comparing the two, which would be a lot more useful than just saying we shouldn't discuss DW unholy.