Feorthas: I'm having trouble figuring out how haste could (significantly) reduce the actual damage output from Killing Machine. Could you explain this assertion?

Possibilities:

- Haste does not affect the proc per hit chance, such that if a 4.0 weapon procs at 50%, 100% haste will provide 2.0/50% and double the procs per minute.

- Haste affects the proc per hit chance, such that if a 4.0 weapon procs at 50%, 100% haste reduces this to 2.0/25%. Both scenarios produce exactly the same average procs per minute. This normalization applies only to white attacks; yellow attacks continue to proc at 50%.

- Haste affects the proc per hit chance, such that if a 4.0 weapon procs at 50%, 100% haste reduces this to 2.0/25%. Both scenarios produce exactly the same average procs per minute. However, this also reduces the chance to proc for yellow attacks by half. Roughly, this reduces overall proc rate by 1/3 for 2H and 1/4 for DW.


The first scenario applies to Maelstrom Weapon and Killing Machine on the PTR build where the offhand could proc KM, as well as many proc-based enchants: definitely classic WoW ones, and anecdotally at least some WotLK ones (fallen crusader).

The second scenario applies to TBC proc-per minute effects such as Mongoose.

The third scenario, as far as I know, applies to absolutely nothing currently known. Furthermore, as it provides a significantly greater damage decrease for 2H Frost (the spec that most needs buffs) than it does for DW non-FS (the spec that most needs nerfs), it seems extremely short-sighted for this to be intentionally implemented. Add in the fact that Blizzard has already shown with the elegant change of KM to PPM that they are interested in buffing 2H Frost and nerfing dual wield simultaneously, I have extreme trouble believing that this scenario is the case, barring bugs.

The same as dr_AllCOM3 believes, this is my understanding too.
Through all of BC, haste increased PPM for all melee classes, I do not see why this would be any different.

This is exactly what I'm modeling... except that the last talent tooltip for new (3.0.8) KM doesn't include Yellow attacks as applicable for procs.

The problem isn't that you shouldn't expect the same number of procs in a minute, it's that the likelihood of a proc happening between skill X and skill Y went down.

Example:

Say we're doing IT BS OB OB FS DMP.

Now, with the theoretical 3.0 spd weapon, we have a 25% chance for a KM proc per white hit. We know that the 'distance' between IT and FS is 6 (seconds), therefore we will swing our weapon an average of 6/3 (=2) times. The expected procrate of KM during that time will be:

1-(1-25%)^2 = 43.75%

Now, we ramp up to a 2.0 spd weapon, we're at 16.6667% chance for a KM proc per hit. The distance between the abilities we care about is still 6 and we now swing 6/2 (=3) times. Expected proc rate during that time will be:

1-(1-16.667%)^3 = 42.13%

This is a 3% reduction in KM procs with a 33% increase (3.0 -> 2.0) in haste, which is very close to what the full IIT line yields so you'll probably have about that much in pretty solid / Tier 7(.5) gear if you pick up IIT (which is also a solid increase in white damage output; ballpark of 33.8% auto attack damage). Anyway, when speccing for DPS, I tend to throw my filler into IIT first, then Toughness because everybody loves Windfury and I don't raid with an Enhancement Shaman right now.

Anyway, a 3% reduction in KM is more annoying than anything else, but I still don't like having two talents working at cross-purposes whereas KM benefited from IIT previously.

Edit:

And yes, the faster autoattacks increase the likelihood of a KM proc between FS and DMP but not by nearly enough to offset the full 3%. In fact, in the more KM-Optimal IT BS OB FS OB DMP, you see a close to 3.5% decrease with 25% haste (and, yes, swapping to IT BS OB FS OB DMP from IT BS OB OB FS DMP will recover nearly 100% of the latter's lost damage; it's about 3.5% better across the board, amusingly enough).

And, while we're at it, IT BS FS OB FS OB DMP (after 3.0.8 + FS Glyph & 4T7) has 5.8% better KM damage than IT BS OB OB FS FS DMP. Yes, we're talking a whole 12.5 single target raid buffed dps here (my gear, ymmv, etc. etc.) but still, I like optimization that doesn't require much effort .

This was not true with Windfury in TBC, and we didn't see it on Mongoose either. I do not believe that your statement is true, but testing could prove me wrong.

First time looking into this thread, long time lurker on DW/Unholy threads, might try out this 21/50.


A quick question, is the new glyph of obliterate(20% more damage to your obliterate) better than the current? (20% more weapon damage 50% less bonus from diseases.)

Lurk more thoroughly the oblit glyph has only had its tooltip updated to accurately reflect how it has always worked. Even if this were not the case, I'm not sure how a 'no' answer would be possible to that question...

That is how PPM has always worked, to my knowledge. This is pretty easy to test - take an untalented death knight and have them melee the test dummy for a while doing white damage only with Crusader on their weapon. It should proc once a minute on average. Now take the same death knight and spec for improved icy talons. They should proc 20% more often.

If you'd prefer, do the same thing with a shaman and Windfury Totem. Should see the same results. Obviously you don't want to use a modern ability since almost all of them have internal cooldowns, so use the old world Crusader enchant.

6% Haste yielded approx 1 PPM over 10m

26% Haste yielded approx 1 PPM over 10m (and then proceeded to proc right AT 10 minutes within 5s of the previous proc, showing that there is little to no ICD)

So, basically, PPM includes haste.

FWIW, a 28.65% Critrate yields a ~2 PPM (@ 26% Haste) for KM so a 5 PPM KM is a big big upgrade.

Hello guys, I am quite new here, but after reading about the 27/44 build and the 17/54, I played with the talent calculator and I was thinking about something like this: 23/48/0

This way, we'd get V3W maxed [+6% STR and stam, which means around 60 str and 78 with Fallen Crusader, for 156 AP along with 6 expertise] and 3/5 Tundra stalker [6% dmg on spells and abilities, +3 expertise].

I'm not very used to the whole theorycrafting, but I think it might help where the 27/44 seems to have a lack.

As a frost spec, why the heck do you have Death Rune mastery? As for your spec, it could use some work, you sacrifice 10% AP for 6% more damage with only spells/abilities. AP scales with spells / abilities as well as white damage. It's also a raid buff, rather then just a self buff.

It's discussed a few pages back. 21/50/0 is the highest dps 2h frost buid by, according to Feorthas, 7 dps over 23/48/0. Most people take 2/3 imp rune tap because Scent of Blood is truly terrible and go with Deathchill+UA rather than Icy Reach. I personally give up more dps and go the offtank spec route because we have 4 DW DKs in guild and I don't need IITal for raid buffs.