DW Builds 3.2 -Revenge of the Offhand - Page 14

Valtiel · 07/15/09, 12:06 PM

Originally Posted by Foxx2405

Awesome, thanks for the test.

I'll definitely try to get slow/slow and so should anyone. But the first thing i noticed when the ToT was announced (and thus the rebirth of DW builds) was that people started saying:
"Speed > Stats > DPS"

Without any decent tests or simulator to back up that statement.

I mean its obvious that slower weapons are a lot better, but so is higher dps. I remember back in 3.0 that the dps upgrade you'd get from switching from anything to Betrayer (2H axe from KT) would result in a massive dps increase simply because of the high dps.

1 hander weapons are a completely different world.

Betrayer was such a good upgrade because you were moving from 3.5 speed or 3.6 speed weapons with lower dps to a 3.5 weapon with higher dps. Even if it was 0.1 sec faster, it had more instant damage due to the dps conversion

So, a 222 dps weapon with a swingspeed of 3.5 may actually have more strike damage than a 213 dps 3.6 speed one. If anything Betrayer's weakness was the AP instead of Str.

Now looking at 1h weapons, we go from 2.7/2.6 speed weapons to 1.6/1.5 ones. The instant damage difference is massive and in this case it very well may be that the (slightly) lower dps weapon with a 2.6 speed is better that the (slightly) higher dps weapon with a 1.5 speed because the little gain on strikes the 2.6 speed weapon nets you is quite higher than the actual white damage dps difference.

Sakuratei · 07/15/09, 12:10 PM

Originally Posted by Foxx2405

Awesome, thanks for the test.

I'll definitely try to get slow/slow and so should anyone. But the first thing i noticed when the ToT was announced (and thus the rebirth of DW builds) was that people started saying:
"Speed > Stats > DPS"

Without any decent tests or simulator to back up that statement.

I mean its obvious that slower weapons are a lot better, but so is higher dps. I remember back in 3.0 that the dps upgrade you'd get from switching from anything to Betrayer (2H axe from KT) would result in a massive dps increase simply because of the high dps.

There is a breaking point where Speed>DPS doesn't count anymore, but that's something we have to discover through testing once we can make builds and test them properly. The main reason why BoH was so much better than anything else was because all the KT loot was a tier higher than the rest of Naxxramas. BoH had roughly 20 more dps than all the other twohanders in the instance. Also, I can't recall any weapon in Naxxramas having 3.6 speed, so the speed difference would be none, or 0.1. I doubt you'll find any onehanders that are 2.4.

So yeah, the Slow>fast rule doesn't make Malice (2.5) crap for example, it's still a decent onehander. But you shouldn't use 1.x weapons under any circumstance.

Valtiel · 07/15/09, 12:24 PM

Originally Posted by Sakuratei

Also, I can't recall any weapon in Naxxramas having 3.6 speed, so the speed difference would be none, or 0.1. I doubt you'll find any onehanders that are 2.4.

[The Jawbone] was the only one i think.

Astalion · 07/15/09, 12:55 PM

Regarding weapon speed: According to the EP calcs I ran way back when I first started looking into deep frost specs, speed is important enough to put Sorthalis (would seem to be the #1 fast weapon for us) slightly behind Titansteel Bonecrusher for mainhand, and just behind Silent Crusader for offhand. LootRank listing for reference. Since then, changes to both spec and simulator would probably cause autoattack dps to be slightly higher, meaning weapon dps would become slightly more important, but slow weapon should still be far better - I'll run some EP calcs again to get some better values though. Note: gear level shouldn't have a huge impact on the importance of weapon speed vs weapon dps - haste would increase the value of weapon dps, while increasing weapon dps itself would increase the value of weapon speed, other stats should affect both more or less equally (there's a few things such as ARP not affecting FS meaning it favors weapon dps, but the effect should be minimal).

Edit: Fair enough, wasn't aware of the normalization, nevermind that part of my post then.
As for advantages with fast weapons - BCB would not cause any more damage with a faster weapon, it should always come out to a set percentage of your autoattack damage, no matter the speed (unless I missed some internal cooldown?)

• Darkside · 07/15/09, 12:59 PM

Originally Posted by Astalion

As for advantages with fast weapons - BCB would not cause any more damage with a faster weapon, it should always come out to a set percentage of your autoattack damage, no matter the speed (unless I missed some internal cooldown?)

Wrong; BCB is a normalized attack (at least it was the last time I checked) meaning that youl will end up getting slightly more damage from it while using a fast weapon. However, this increase in negligible, especially when compared with the damage loss you will suffer on your strikes.

mongoose221 · 07/15/09, 2:00 PM

I believe in 3.1 that BCB had a 3-second internal cooldown applied to it, which would greatly reduce a % increase in number of procs a fast weapon gets vs. a slow weapon, kind of like how shamans windfury has an internal cooldown.

Foxx2405 · 07/15/09, 2:20 PM

Originally Posted by aethereal

The case in question isn't the 20 dps but the base damage difference between the 2 weapons. For example, the top end damage difference between razorscale talon and remorse is 552-349 = 203. Even if you divide that value by half for offhand (ignoring NoCS), that's ~101 before modifiers per strike that is affected by ToT.

The reason why BoH was such a dps increase over everything else in naxx follows the same reasoning, for it has the highest base weapon damage.

Perhaps "weapon speed" should be replaced with a better term like base weapon damage to avoid confusion.

A 101 damage increase is tantamount to almost 600 ap ( 2.4 * 600 /14) in terms of pure weapon damage. I do recognise, however, that strikes use only a % of weapon damage for calculations while other stats such as ap affect all aspects of a DK's dps. Ultimately it all depends on the proportion of weapon damage in the said build's damage breakdown.

Hmm in hindsight using the betrayer wasn't a good example.

Weapon speed and base weapon damage are not the same though. Sure same dps weapons with a lower speed have higher base damage, but same speed weapons with higher dps have higher base damage too.
Might be smart to indeed not use the word weapon speed if you talk about weapon damage, there is quite a difference.

Ok well when you're talking base weapon damage the fast (1.5ish) weapons indeed can be striped off since their average base damage usually lies a lot lower.
I did some calculations (i got them saved if anyone really wants to see them) and you gain something in the area of 100-150 dps using a slow offhander instead of a fast offhander.

Kyruski · 07/15/09, 2:22 PM

Originally Posted by mongoose221

I believe in 3.1 that BCB had a 3-second internal cooldown applied to it, which would greatly reduce a % increase in number of procs a fast weapon gets vs. a slow weapon, kind of like how shamans windfury has an internal cooldown.

This was in the patch notes for 3.1 but then Blizzard reverted the change.

Fargom · 07/15/09, 3:02 PM

Darkside's version of blood caked blade is exactly correct, this log can be viewed as proof.

World of Logs - Real Time Raid Analysis

Navigate to Fargom's damage on Ignis, and note that there are two separate entries listed for Blood-caked strike. WOL records each hand separately, we can clearly see that the fast hand is producing .2% more damage via BCB than the main. (Also, as a side note we can see that BCB itself is producing 5.6% damage for 3 talent points, quite good.)

Darkside is correct in saying that fast weapons produce more damage, but the difference is fairly small.

-If someone wants to really put this speed issue to bed, run a simulator for a common frost build, find the percentage of white damage vs strike damage, then find the break point. IE is there a logical point where using a 178 dps fast weapon is better than a 163 or 158 dps slow offhand.

*Sorry for not being able to drill down into the WOL report and link the specific page, I'm at work and using some "creative use of Internet mechanics" to get around our draconian security software.

Afabar · 07/15/09, 4:03 PM

I have been toying with the simulator to find out if speed > dps on DW.

For all the test, I have use
Strength = 1654
Hit Rating = 280
Crit Rating = 750
ExpertiseRating = 170

Using 00-54-17
- 400 haste / 100ArP
- Using on MH a 180dps@2.6 and a 180dps@1.6 => 6275 DPS
- Using on MH a 180dps@2.6 and a 150dps@2.6 => 6276 DPS

Using 00-54-17
- 100 haste / 400ArP
- Using on MH a 180dps@2.6 and a 180dps@1.6 => 6418 DPS
- Using on MH a 180dps@2.6 and a 150dps@2.6 => 6415 DPS

Using 10-54-7
- 400 haste / 100ArP
- Using on MH a 180dps@2.6 and a 180dps@1.6 => 6382 DPS
- Using on MH a 180dps@2.6 and a 130dps@2.6 => 6381 DPS

Using 10-54-7
- 100 haste / 400ArP
- Using on MH a 180dps@2.6 and a 180dps@1.6 => 6549 DPS
- Using on MH a 180dps@2.6 and a 130dps@2.6 => 6560 DPS

Conclusion
For a 00-54-17 a Naxx25 Slow weapon equals a fast Uduar25 fast weapon.
For a 10-54-07 a iLevel 200 blue Slow weapon equals a fast Uduar25 fast weapon.

entity · 07/15/09, 4:17 PM

Has any math been done to show that BCB does significantly more damage than the added crit bonus from Subversion?

Edit: Reason: more in-depth clarification
Sorry as I was posting this I received a phone call that required my full attention.

The reason that I ask because the more recent posters on this board have been centralizing thier builds with a focus on damage from OB and HB over the damage from FS. this makes perfect sense because of the nerf incoming to FS.

I dont happen to have any sort of sim program so I can't do the math for myself but I was wondering if someone would take the time to do a point by point comparison of these two talents for a deep frost dw build.

for the build with Subversion this http://talent.mmo-champion.com/?deat...&version=10072 build can be used.

and for the build with BCB the same build but with the three points moved over to BCB can be used.

With a test like this done we would be able to see the comparative value of these two talents.

Astalion · 07/15/09, 4:30 PM

Originally Posted by entity

Has any math been done to show that BCB does significantly more damage than the added crit bonus from Subversion?

My tests showed Subversion adding more damage (don't have the exact numbers), but you can kinda get both (2/3 BCB) anyway,

entity · 07/15/09, 4:41 PM

Originally Posted by Astalion

My tests showed Subversion adding more damage (don't have the exact numbers), but you can kinda get both (2/3 BCB) anyway,

What build would you be referring to?

Foxx2405 · 07/15/09, 4:53 PM

Nice test...

I guess that pretty much clears that up.

Average damage is the most important stat in a weapon. (I prefer talking about weapon damage than speed, as that is a combination of high dps and slow speed)

Guess a safe guideline would be: "A slow (2.4+) weapon of a certain tier is as good as a fast (1.4+) weapon 1 tier higher".

Kyruski · 07/15/09, 5:04 PM

Originally Posted by Foxx2405

Nice test...

I guess that pretty much clears that up.

Average damage is the most important stat in a weapon. (I prefer talking about weapon damage than speed, as that is a combination of high dps and slow speed)

Guess a safe guideline would be: "A slow (2.4+) weapon of a certain tier is as good as a fast (1.4+) weapon 1 tier higher".

Actually it would be 2/(3) since there would be 213 for Naxx25, 219 for Uld10, 226 for Naxx25 KT, 232 for Uld10 HM and Uld25. But something you have to really look at are the stats also. [Caress of Insanity] and [Serilas, Blood Blade of Invar One-Arm] are supposedly getting more EPAP than the iLvl 239 2.6spd 188.1dps PVP weapons: [Furious Gladiator's Truncheon] because of the stats that it brings. This is something important to note.

EDIT: But You could consider the PVP weapons to be a separate condition because of their main use.

Lazengann · 07/15/09, 5:32 PM

Originally Posted by Foxx2405

I haven't done calculations (I could but im at work atm). But be aware you dont undervalue the 4xT9.

Dots atm can't crit at all, and with 40% spellcrit (which i assume is reasonable for T9 content ?), it already increases DoT damage by 40%.
Checking Phantazums recount a few posts up and dots did about 6.4% damage, increasing that by 40% would mean roughly 2.6% more dps.

8% extra frost strike crit is awesome, but keep in mind that due to KM it already has a very high average crit chance. Say it has a 60% crit chance already, increasing that by 8% would mean:

168% / 160% * 100 = 105% damage. A 5% frost strike damage increase.

If frost strike is 25% of your damage (again using Phantazums recount just because its there), that means a 1.25% dps increase.

These are mostly made up (but i assume somewhat realistic) numbers that just imply that "at first sight" can be decieving.

Then I guess keeping the 2 piece T8 wouldn't work as well compared to 4p T9.

Astalion · 07/15/09, 5:37 PM

Originally Posted by entity

What build would you be referring to?

This, or similar ones. A variant if you want to provide IIT is moving 2 points in your spec from KM to BCB, gave me better results at least (and err, BS or Pest Glyph instead of Ghoul would be better I believe).

AtheistGod · 07/15/09, 5:47 PM

Originally Posted by Foxx2405

I haven't done calculations (I could but im at work atm). But be aware you dont undervalue the 4xT9.

Dots atm can't crit at all, and with 40% spellcrit (which i assume is reasonable for T9 content ?), it already increases DoT damage by 40%.
Checking Phantazums recount a few posts up and dots did about 6.4% damage, increasing that by 40% would mean roughly 2.6% more dps.

8% extra frost strike crit is awesome, but keep in mind that due to KM it already has a very high average crit chance. Say it has a 60% crit chance already, increasing that by 8% would mean:

168% / 160% * 100 = 105% damage. A 5% frost strike damage increase.

If frost strike is 25% of your damage (again using Phantazums recount just because its there), that means a 1.25% dps increase.

These are mostly made up (but i assume somewhat realistic) numbers that just imply that "at first sight" can be decieving.

Also, to further your argument, + crit modifiers weaken Killing Machine. Say 20% of your FS are made with KM up. So of the remaining you have a 50% crit chance. now you have a 58% crit chance 80% of the time and 100% crit chance 20% of the time. You will have a 66.4% crit chance, Only 80% of the benefit from the 8% crit bonus actually helped.

Foxx2405 · 07/15/09, 7:20 PM

Originally Posted by Kyruski

Actually it would be 2/(3) since there would be 213 for Naxx25, 219 for Uld10, 226 for Naxx25 KT, 232 for Uld10 HM and Uld25. But something you have to really look at are the stats also. [Caress of Insanity] and [Serilas, Blood Blade of Invar One-Arm] are supposedly getting more EPAP than the iLvl 239 2.6spd 188.1dps PVP weapons: [Furious Gladiator's Truncheon] because of the stats that it brings. This is something important to note.

EDIT: But You could consider the PVP weapons to be a separate condition because of their main use.

Ah yea, i was counting 25 naxx -> 25 ulduar as 1 tier (T7.5 to T8.5).

Indeed stats make an impact as well,
On that subject I haven't been following this whole threat, just the first page and the last few pages. But is there a sort of stat value yet? (I know its kinda hard without even being 100% about the spec, but roughly would do).

Its obvious that strength still reigns, but im more wondering about the other stats:
Expertise
Crit
Hit (after hitcap)
Haste
Armor Penetration

- Do we prefer to be expertise capped or take whatever we can but not hunt it down ? (BiS list suggested the latter iirc)

- With the 245% crit modifier crit is pretty good for frost, but with the high innate crit (due to talents mainly) crit has a reduced value. (See my post about the 2 set T8 bonus)

- We want to be special hit capped for sure. How valuable is spell hit cap, since from what i read an Obliterate heavy build will be best, that rotation only really has IT and Rime HB as spells. Do we want to gem cap it if we lack hit from gear ?
Again don't be to quick to judge on this on "first sight", it may seem stupid not to cap spells but the chance to miss say an icy touch with 1-3% spell MISS is rather small. Once in 2 boss fights maybe, thats stat that could have increased every attack on those 2 boss fights by the strength you'd gain.

Also do we want to avoid going over the spell hit cap, even for DW hit rating takes a huge dip after the spellhit cap, because our white dps is (relatively to other dual wielders) small.

- How valueable is haste, i assume its not that good.

- What about armor penetration, in an Obliterate heavy build i could see that as a very interesting stat.

My guess is (assuming you at least get the special hitcap):

Strength > Armor Penetration > Expertise > Hit (pre spell cap) > Crit > Haste >= Hit (post spell cap)

(This could be way off though)

Fargom · 07/15/09, 7:34 PM

Statweights are a direct function of talent builds, unless I am mistaking it is pointless to try to come up with a general statsweight that is not attached to a spec.

I view APE values and BIS lists as the finishing touch of a "spec post." Once someone has established and simulated a given spec, it is then time to see how well the APE / BIS lists fit with the real world gear. Often times, there are multiple builds so close to each other in raw DPS, that their various statweights VS. available items is the determining factor.

Moving forward I'd like to urge everyone to hold off on statwieghts and continue to work on quality (IE non targeting dummy) spec posts. I'd be happy to update the OP with any spec that can stand up to peer review. Post your spec including evidence and or sim data, then let the community tear it apart. If it passes the test, it deserves to go on the OP. Anyone disagree with this method?

Foxx2405 · 07/16/09, 3:24 AM

Originally Posted by Fargom

Statweights are a direct function of talent builds, unless I am mistaking it is pointless to try to come up with a general statsweight that is not attached to a spec.

I view APE values and BIS lists as the finishing touch of a "spec post." Once someone has established and simulated a given spec, it is then time to see how well the APE / BIS lists fit with the real world gear. Often times, there are multiple builds so close to each other in raw DPS, that their various statweights VS. available items is the determining factor.

Moving forward I'd like to urge everyone to hold off on statwieghts and continue to work on quality (IE non targeting dummy) spec posts. I'd be happy to update the OP with any spec that can stand up to peer review. Post your spec including evidence and or sim data, then let the community tear it apart. If it passes the test, it deserves to go on the OP. Anyone disagree with this method?

Fair enough. Then again i wasn't asking for APE numbers, i was asking for a general idea.

I'll see if i can get my hands on a simulator and test out some stuff over the weekend.

Astalion · 07/16/09, 5:03 AM

Originally Posted by Astalion

LootRank listing

Posted this a few times, it is kinda outdated, but should give you a rough idea of what stats we want as deep frost (this is before spellhit cap, after that the value of hit was around 0.77). Started some new EP calcs during the night, but I got up to see Vista had rebooted for some update and removed all results, gg. Will give it another shot tonight (planning to include that for a more thorough post about "my" frost spec anyway)

Amiko · 07/16/09, 4:26 PM

A few posts back the disease glyph came up again, and I'm really not happy with the theory craft and what not that's gone down about it.

I'm using it quite effectively, in blood presence, prioritizing obliterate over all else.

Here's the build:
The World of Warcraft Armory

the rotation basically goes:
OB, OB, Pest, Dump

Death rune mastery is important, -- you can pestilence anywhere in the cycle and and that pestilence will turn a blood rune on CD into a death rune when it comes up so the obliterates actually come really quick.

A lot of the theory craft i read seems to comparing 1 obliterate to a it/ps disease refresh, but if you have both death rune mastery and BotN you're not just avoiding the it/ps refresh, but you're also avoiding the super weak blood strikes as well -- there's no bloodstrikes in the rotation beyond the initial setup at all at all -- just a lot of obliterates.

Here's a WWS:
Wow Web Stats

I'm usually a tank so only the hodir attempts are relevant ( not the best fight for this sort of think i know -- Hoping to get a deconstructor parse for reference next week ). It's important because there's a standard unholy dk there for reference.

Given how the SoA seems to be working on the PTR right now ( full application to the off hand strike ) I'm thinking that a very similiar build:

17/54/0 - Talent Calculator - World of Warcraft

dummy to dummy tests ( between live and ptr ) seem to indicate ( using Malice/Razorscale Talon with FC/CG ) show the dual wield coming ahead by a good 10% with the blit/blit/pest rotation in BP.

Since the thinking here seems to be moving towards more oblit heavy rotations i was curious if any else was working with or testing similar builds/rotations.

Astalion · 07/16/09, 4:44 PM

Originally Posted by Amiko

A lot of the theory craft i read seems to comparing 1 obliterate to a it/ps disease refresh, but if you have both death rune mastery and BotN you're not just avoiding the it/ps refresh, but you're also avoiding the super weak blood strikes as well -- there's no bloodstrikes in the rotation beyond the initial setup at all at all -- just a lot of obliterates.

First off - are you using only 5 of the runes? (If so, you would always have a blood rune up for pestilence anyway, making DRM completely useless?) Blood Strike will be better than not using the other blood rune at all (more Obliterates from turning that blood rune into a death rune as well) or using it for another pest (you most likely will cut some unneccesary ticks off the disease as well as not dealing any damage with the rune). Also, I wouldn't consider a "trade" of "minimal BS damage" to 0 damage (from pestilence) to be the good part of the trade really. From your reasoning you seem to either somehow be extremely GCD capped or counting Blood Strike as negative damage?

That said, using Glyph of Disease does look promising from the (sort of rough I guess, but the best I could think of) math I did a few pages back.

Amiko · 07/16/09, 5:16 PM

@Astalion:

No you're using all six runes -- that 6th rune should always be a death rune so what you're doing is shortening the length of the rotation and _not_ assigning specific runes to specific attacks ( since almost everything you do makes a death rune). That 6th run will be consumed by an obliterate as soon as any other rune comes up -- the "extra" gcd that comes up by not blood striking becomes part of the rune dump or an opportunity to use a rime proc. that blood rune that comes back goes back out as a pestilence, yielding a death rune to get spit back out as an obliterate.

An early or late pestilence to deal with movement issues will often hit a death run that should be a frost or unholy rune and is totally not a problem because it'll still convert a refreshing blood rune to a death rune and and the F/U rune will get consumed by an incoming obliterate -- Basically it becomes a priority system more then a rotation, with obliterate > frost strike > pestilence, but you have to pestilence 1+ times per disease duration.