Dual Wielding DKs: Demystifying the Myth.

[Krenian]

Correct me if I'm not remembering right, but KM procs no longer rely on crit, correct?

Essentially it is now far more worthwhile as a tank since you don't need dps stats to get the effect.

And the patch has plenty of relevance to DK tanking, since we're seeing a LOT of survival changes, with the baseline increase and the special move reduction. There are a pile of glyph changes. The up shot of this is that after the patch there can and will be a lot of swing room on things we wouldn't generally consider right now, particularly as far as not trying to load up on survival talents...

I've been messing around with stuff on the ptr and it's just plain more fun. I'm doing just fine tanking now and this is about to make me sturdier?? Bring on the patch!

Go figure that I write up a thread and forget about it. It's people like Satorri here that remind me that I should keep an eye on stuff even if they're great people talking about it.

KM's change is going to fix a few things. For one, we won't have the issue of having low crit. This is actually more beneficial to a DW DK than one would presume due to the fact that now it just on every hit. This puts a focus on hit which our gear does kinda lack. This should also give us a boost in damage for DW DKs due to the fact we're not so low in crit to ever see KM proc.

KM is now on a PPM, which means pretty much it'll proc a certain amount of times per minute. Pretty much the following is true:

You will have KM proc so long as you hit the target. Miss and your proc rate will be lower. but hitting more often won't give you a higher PPM due to it being maxed out to a specific PPM. I believe this PPM maximum as it stands now is 5 or 10. The number's been floating around and I've been working on school stuff right now.

Therefore a tank will have a FAR BETTER chance of getting a KM than it was in it's current 3.0.3 form which is an increase in TPS for us Which is good, of course.

So far, from what I've seen from the patch notes, there has been some slight changes to our cooldowns on how well we use our our key skills (IBF, BA..) However a lot of the changes have actually complimented dual wielding tanking moreso which may be something to sincerely look at even more.

I'll try to look into this thread more and answer what I can. And i'm happy to note that the thread hasn't dropped into a derogatory namecalling and idea bashing. It's nice to know a community can disagree on ideas and yet still discuss it as rational, mature adults. Keep it up folks.

[Satorri]

Aye, and some tanking gear will actually have hit rating. Tank gear never has straight crit value (though I'm hoping the next tier of tanking gear has agility mixed in with plenty of Str/Stam, it's too beautiful a stat for a tank).

Usually the way PPM's work out is that there is a functional cap, where provided you have the right means to reach it (haste, hit %, etc) you can get up to but not past a certain number of procs in a time frame. That being said, the caps almost always favor dual wielding (more hits to fill that potential). This seems to be the case here as well.

[Krenian]

Which will probably mean dual wielding will benefit the most out of this change. Which most people don't realize at all. Which makes me shake my head at the lack of common sense people have.

Oh well, people are too blind by math. ><

[Satorri]

Some people hide behind math, and some people run from it.

At the end of the day all the theoretical play we can do here takes a back seat to playing the game. The nuances of a talent point or a gem here and there is always overshadowed by the skill of a player.

[A Man In Black]

Therefore a tank will have a FAR BETTER chance of getting a KM than it was in it's current 3.0.3 form which is an increase in TPS for us Which is good, of course.

Us being all frost tanks? An PPM internal cooldown means that increasing your attack rate (with hit, expertise, faster weapons, whatever) sees sharply diminishing returns after a certain threshold, unless the internal cooldown is really short (4 seconds or less). The new KM may be better for all tanks, but the delta of KM benefit between DW and 2H has decreased.

It's going to depend on what the cooldown is; it seems to be something like 10-12 seconds.

I'll do some scratch math on this later.

[Satorri]

Right.

It may start off better, but it has a logistical cap, so you can only get so much mileage out of the talent alone (though the stats will continue to help you as usual so far as they did before).

[Futeko]

Howling Blast has now a 5-sec CD.
It totally killed my urge to test DW tanking after the patch ... I seriously hope they'll find a way to keep HB off CD. I feel like us tanks are paying for the poor balance between DPS spec.

[A Man In Black]

Right.

It may start off better, but it has a logistical cap, so you can only get so much mileage out of the talent alone (though the stats will continue to help you as usual so far as they did before).

Dual-wielding doesn't get any extra benefit out of hit or expertise that 2hers do not, until the 2her caps hit to yellow attack cap. Both specs are getting +X% white damage and +X% yellow damage and +X% to spell damage (in the case of hit only of course). As far as KM is concerned, 2hers get more benefit from hit/expertise because all stats that decrease time between KM procs (hit, expertise, faster weapon speed) essentially share diminishing returns. It's a simple curve approaching but never reaching average time equalling a perfect proc as soon as cooldown drops.

For tanks, this is a bigger buff for 2h than DW. And rightly so; KM was previously considered mediocre-at-best filler for a 2h frost tanking spec. It was really really bad.

tl;dr: KM's changes don't really affect the discussion much at all.

[Lunga]

As a beginner tank (still level 75), this thread has been very interesting to me.

Let me try to interpret things differently - correct me if I'm wrong I read the thread quickly and everything is new to me:

1) The idea that "dual weilding dks should be shot" because of parry-haste catastrophies (caricature of an actual comment at Crescent of Brooding Fury - Item - World of Warcraft) has been debunked.

Why? First if increasing parry-haste attacks is never acceptable, then it would be logical to conclude that "unholy DK tanks should be shot" (on the ground that they would have more parry-haste events than 2H frost DK - arguably - if I understood properly).

And then, the math of AMiB, summed up differently, means that "even if you have a 2H, you'll still have 75% of the parry-haste events of a DW DK". Hence IF you accept that a parry-haste event is unacceptable then it follows that: "DK tanks who are not expertise capped should be shot".
Expertise would have to be capped like defense is. Funnily, once you are capped, it doesn't matter if you 2H or DW from this point of view, so this contradict the initial statement. Also, since it's hard to cap expertise, it would follow that: "almost all DK tanks should be shot".


3) Once you accept that parry-haste is survivable (avoidance, cooldowns or healer reaction) and that you will have some of them whether you are 2H or DW, you can consider that DW is doable and is like a talent that would say:
[.] Icy Curse of Good Looking Knighthood: increase parry-haste events occurences by 32% but: makes you look more cool ; may increase your threat generation ; and allow you to use stats on two differents items instead.

Would you take it?
At first sight I would not, but really, how bad is it?

First, it seems the jury is still out as far as threat generation goes.

Second, it's better to receive parry-hasted attacks than critical strikes, so as pointed out, if you are lacking defense, then it makes perfect sense to go DW just to reach defense cap. (I'm not sure what critical strikes rate you are trading for which parry-hasted attacks rate though).

Third, assuming the cost is only increasing damage (due to parry-haste), they it was argued "one extra hit a minute from that two-second swing-timer boss is like losing 2.8% avoidance from your gear". Not sure is an accurate figure when taken out of context, but assume it is.
Now there was a long debate about min-maxing, but the fact is that we all do min-maxing to some extend and all stop min-maxing at some point. For instance if there was an extra enchant +200 Strength available for 1 gold, almost every DK would take it ; but if there was an enchant +1 Strength for 5000 gold, few people would take it. Hence everyone has threshold for min-maxing (which is different depending on people, guild, content, and so on). The question is that is the cost of DW within the threshold of many people?

What I would like to see is to quantify is how much this 2.8% avoidance represents in easily understandable terms -- for instance an estimate of how many pieces of normal Naxx gear you would have to upgrade to heroic Naxx gear to compensate (or blue gear to normal Naxx). Because if it is one piece, then the debate is closed (no consequence) ; if it is 10 pieces also it is closed (sorry but DW is out for me) ; if it is ~3 pieces, then it is the grey area.

Does anyone have an idea?


PS: more personnal opinion: it has been argued that you could compensate DW by stacking expertise or/and hit. that's quite true and interesting to quantify the impact of DW. But in practice if you wouldn't seek extra expertise/hit for a 2H build, then most likely you wouldn't seek it when DW ; because, most likely the marginal value of each stat would be comparable in both cases (uneducated guess). You don't need to further go out of your way to compensate imperfectly the fact that you went out of your way (with DW).

[GravityDK]

For those interested, on dk.info I have summarised DW tanking in a stickied post here. Any comments on the accuracy or otherwise of my summary would be appreciated.

That summary is below. (Forgive the colours, dk has a black background and it was a bit too time-consuming to change them for the below)

[btw, good recap in previous post Lunga]

This thread should consolidate our knowledge on dual-wield tanking, and ideally any new threads on the topic should come in here. DPS DW is discussed this other thread.

I'll update this post as we agree on more facts. Updated for patch 3.08.

NOTE: we are not promoting nor advocating dual-wield tanking. This thread's purpose is to discuss it and show it can be viable under the right conditions (and dangerous otherwise).


The summary
Some credible analysis and maths has been done which has shown that:
* in terms of parry-gib risk, DW tanking is equal to two-handed tanking if you have 21 expertise skill.

If you have less than 21 expertise:
* Dual-wield tanking exposes you to more parrygibs than 2H tanking
* Therefore you increase risk of taking spike damage
* and should be avoided.

Casual DK tanks can do whatever they want, because their guild won't care about a wipe or two more. They should just be aware of the risk they're taking.


ANALYSIS
The risks and disadvantages
- you offer more chances for the boss to parry you and for him to get a hasted next attack (60% faster iirc). This behaviour is confirmed to still be in the game.
- your death strike heals you for less than it would from a 2Her.

The advantages
- talents which lever from auto-attacks become more valuable (eg. killing machine) although some in turn increase your risk of being parried (eg. Blood caked blade) and are not attractive.
- you can skip Annhiliation talent because you will use Howling Blast as FU ability

Other factors
- rune strike cannot be parried, so if many of your attacks end up being rune strikes rather than normal hits, you reduce your 'parryable' attacks by that proportion
- because you are attacking faster, you get to use a higher proportion of your rune strike opportunities than a 2Her tank. A 2H speed of 3.4 makes it possible for you to dodge a boss' attack twice within that window, but because RS works on next melee swing, you will only get to use one of those two opportunities. In constrast, a DW tank will get to use more of their RS opportunities with their 2.5 attack speed.
- patch 3.08 introduced a 25 Defence skill plus 2% health runeforge for 2Hers only.

Expertise, to-hit and reducing those parrygibs
If you stack expertise, you reduce the chance the boss can parry you.
Bosses have a 14.25% parry chance and reducing that to 1% would take 467 expertise rating (= 57 expertise). That's massive and it's not sensible to even try to achieve that when stacking stamina would be more effective instead.

For level 83 the soft cap is 6.5% dodge or 26 expertise (214 rating), and the hard cap is 14.25% parry or 57 expertise (468 rating). Source.

It has been calculated that you need 21 expertise skill in order to bring your parrygib chance to equivalence with a 2Her tank.

Spell Hit and Melee Hit scale differently with Hit Rating. You need 26.2 hit rating for 1% chance to hit with spells and 32.7 hit rating for 1% hit to melee attacks.

[size=8pt]Expertise:
* 1 Expertise Skill reduces the chance your attacks will be dodged/parried by 0.25%
* 32.79 Rating is equal to 1% less chance your attacks will be dodged/parried.
* 8.20 Rating is equal to 0.25% less chance your attacks will be dodged/parried and thus is equal to 1 Expertise Skill.[/size]

To-hit
Dual-wielding, the offhand has a 24% chance to miss rather than the usual 9% for two-handed weapons. So to-hit as a stat increases your dps nicely. Note, your special attacks (eg. frost strike) disregard the offhand miss rate, your special melee attacks are 9% and spells are 17% even if you are DW tanking. To-hit does not effect your survivability, just your TPS.

There is some debate at the moment and analysis in this thread that perhaps it's now 8%. It's a minor point, effects DPSers more.


OK I want to DW tank anyhow! Spec? Talents?
You will largely be a Frost tank.
Ideal starting weapons: Infantry Assault Blade and Titansteel Bonecrusher.

You want two slow weapons, otherwise you increase the chance of parrygibs beyond what 21 expertise protects you from. Other choices are here, there aren't many.

Frost has the one-handed talent, so every DW spec would take that up for TPS and to-hit. For DPS, the 0/32/39 is often considered the highest spec, but a tank has different concerns. We'll look at any tree that's had some analysis.

Death coil is a great RP dump for all trees because it cannot be parried either, but note it suffers from being treated like a spell for to-hit, but benefits from scaling on attack power (and not from weapon damage, like frost strike as an alternative). We speculate DC will do more damage than FS as an RP dump, but need more maths on whether eg frost talents swing it.

Sample build 1 and build 2.

Frost
Frost strike can't be dodged, blocked or parried. Treated like melee for to-hit. Good.
Howling Blast is brilliant. It can't be parried and it replaces obliterate in your rotation, therefore reducing the chances of parrygib. Mandatory.
Killing machine, more dps without increased risk of parry. Good.

Unholy
Scourge strike can be parried, but so can obliterate which it replaces. No advantage.
Necrosis increases your white damage with no additional risk. Good.
Sample unholy build.


MATHS
Analysis maths to show that DW DK parrygib requires +16 expertise
This linked thread is the central argument where DW tanking has been shown to be viable. It's good reading.

Here is the thread

Here are links to the key posts within that thread:
Satorri's first analysis seemed pretty solid.
It gets heated here posting. Quite a good debate ensues.
Lyd then puts the maths together which is foundation of our 21 expertise baseline.
Tony also restates the purpose of the thread here, which is not to advocate DW tanking just to show that it's viable.


WWS
There is also an effort to collect data on raids using WWS, and an account set up for this purpose.
Here are the details.
So you can upload your combatlogs there and share the information.


Useful factual references
Threat values for DK abilities

Updated Jan 4:
- refreshed to 3.08 content, made more positive to DW tanking

[A Man In Black]

Why? First if increasing parry-haste attacks is never acceptable, then it would be logical to conclude that "unholy DK tanks should be shot" (on the ground that they would have more parry-haste events than 2H frost DK - arguably - if I understood properly).

Unholy DKs have advantages that make up for their greater parry haste exposure. DW frost DKs do not (or at least, none that have been shown).

Basically you've missed the idea that it's okay to have greater parry haste exposure as long as you're getting something out of it. Not everyone parry-caps expertise because expertise is a weak mitigation stat. Not everyone specs 2h frost because unholy (and possibly blood, post-patch) have other advantages.

The strawman that "Parry-haste is completely unacceptable" has been thoroughly debunked, but nobody was making that argument in the first place.

[.] Icy Curse of Good Looking Knighthood: increase parry-haste events occurences by 32% but: makes you look more cool ; may increase your threat generation ; and allow you to use stats on two differents items instead.

Yeah, that's pretty much DW tanking right there. Plus, it costs three talent points.

Second, it's better to receive parry-hasted attacks than critical strikes, so as pointed out, if you are lacking defense, then it makes perfect sense to go DW just to reach defense cap. (I'm not sure what critical strikes rate you are trading for which parry-hasted attacks rate though).

Not an issue post-patch.

Long analogy about min/maxing and money

DWing isn't cheaper; as a practical matter, it's more expensive, as you'll need two drops instead of one. In this case, you're doing more work to get less results.

What I would like to see is to quantify is how much this 2.8% avoidance represents in easily understandable terms -- for instance an estimate of how many pieces of normal Naxx gear you would have to upgrade to heroic Naxx gear to compensate (or blue gear to normal Naxx). Because if it is one piece, then the debate is closed (no consequence) ; if it is 10 pieces also it is closed (sorry but DW is out for me) ; if it is ~3 pieces, then it is the grey area.

It's approximately 44 dodge rating (before DR). It's nearly three talent points. It's more than two and a half gem slots left empty. It's like replacing Deflection Band with Dragonflight Great-Ring.

PS: more personnal opinion: it has been argued that you could compensate DW by stacking expertise or/and hit. that's quite true and interesting to quantify the impact of DW. But in practice if you wouldn't seek extra expertise/hit for a 2H build, then most likely you wouldn't seek it when DW ; because, most likely the marginal value of each stat would be comparable in both cases (uneducated guess). You don't need to further go out of your way to compensate imperfectly the fact that you went out of your way (with DW).

You are absolutely correct.

[Krenian]

I'm going to bump this seeing how it has been asked a couple of times and I believe that it's a rather solid discussion about the pros and cons of Dual Wielding. The idea is still sound to this day however I'm awaiting the changes in Frost to see if it may impact either Dual Wielding tanking or DPS. We'll have to wait and see.

[Texz]

Just thought i'd put my 2 cents into this thread. Still fairly new to tanking in general and im loving my frost spec DW on my dk. So far i've OT'd both heroic and normal naxx and havent had to much trouble. I'll get a link to spec and stats sometime later today. At first i used the DW setup just to cap out my def. but later it kinda grew on me. I have more avoidance using dw and my exp. is max'd w/o having to gem for it. Would recommend Acclimation in ur frost tree tony... i know starting otu heroics when that would proc loken's nova would only hit me for bout 1100... havent tried it lately tho so might b less for me now. Anyhow great info from everyone and thinks for breaking stuff down... I'm no math wiz either and it helps to see it put in a lil more English terms


PS... Only reason im OT atm is our MT has 2 years more tanking experience so i let him lead and take up the slack where i can... plus easier for my aoes to pullin the adds and he jst focus' on one target. Lead tank is war btw...

[Rennadrel]

Would you not want hit rating for dual wielding or is there no insane hit cap requirement for DK's like there is for rogues with two weapons where it takes so much work to get to hit cap.

[Satorri]

The hit cap for auto-attacks is the same for all dual wielders, and it's high, usually higher than most dps will get to, let alone tanks.

The hit cap for special moves is the same for all dual wielders, and there is no penalty to that (it's the same as if you were wielding a single weapon).

For all intents and purposes, in the melee world, those carrying a 2-hander or a shield and sword/mace/axe will not have use for hit rating beyond the 8% cap, ever. DK's and dual wielders on the other hand will have use for hit rating beyond that, though after the 8% mark it becomes a little less valuable.

Personally, hit rating is one of the last things I'll consciously pursue in my tanking gear, only after I feel like my health, avoidance, and expertise are all at an acceptable level. As a tank, you'll get double the hit value out of Expertise rather than hit until you get to 26 expertise on the sheet (32.79 hit rating reduces your miss chance by 1%, 32.79 expertise rating reduces the targets parry AND dodge chance by 1%). The only mechanism that hit helps for you is threat, so it is a matter of squeezing the last bits of threat out of your gear once your survival is at least sufficient.

[lyd]

i'm just going to report in on this one:

here's the armory for my toon:

The World of Warcraft Armory

(stats are outside frost presence, obviously)

24 expertise, roughly 550 effective defense with the relic proc up, avoidance well over 50% self-buffed with blade barrier up.

I have tanked every boss in ten-man naxx, saph, sarth, and ot'd patch on 10 and 25 man nax. a couple of my pieces are somewhat below-average blues, and throughout all this, i have never once had a problem with tanking due to dual-wielding. no healer has ever complained that i was difficult to heal. i am smart with my mitigation cooldowns and i know how to time my rotations correctly. note the fast off-hand and less-than-impressive slow main-hand.

there are DK tanks on my server that DW fast/fast whom are soft-capped and have tanked all 25-man content successfully and repeatedly. the expertise gear is out there and works great. i know that some people are worried that the theory behind DW makes 2h always a superior choice, but it really is a matter of preference. if you don't want to get the expertise gear, stick with 2h. if you're like me and you find an underdog/alternate playstyle challenging and fun, then go for it. once you hit the soft expertise cap, any amount of extra damage you take from parry-haste will be so small over the long run that the chance of slowing progression as a result of your gear style choice is practically nil.

[Satorri]

Absolutely.

I think there's a double return as your gear improves. With 20+ expertise and over 50% avoidance you just don't see the drawback of extra parriable swings.

[Ramalama]

According to my own combat logs over several weeks, I'm being attacked (not hit) ~4-7% more often by a single target due to parry hasting while dual wielding as opposed to using a 2H. This agrees with my personal theorycraft calculations as well.

Not only is that a relatively insignificant difference (in my opinion), but it may actually be the case that I'm taking less overall damage while dual wielding.

My passive avoidance is around 70-75%. So increasing my avoidance by ~1-2% means I'm taking ~4-7% less damage. That's right in line with the defensive bonuses that I gain by dual wielding two tanking weapons. People have talked about having to "compensate" for dual wielding by gearing for additional expertise, but it appears there's nothing even to have to compensate for.

So as far as average damage taken goes, dual wield may even come out on top after factoring in the extra tanking stats. Either way, the differences are so slight as to be imperceptible given the intrinsic random nature of incoming damage while tanking.

To put it another way, the difference between a non-Night Elf tank and a Night Elf tank is larger than the difference between a DW tank and a 2H tank. I don't think anyone could argue that choosing to play a non-Night Elf tank is an unreasonable decision.

[Satorri]

Can you post that data, I'd love to see it. No one has been able to so far, and I think it's important for people to see.

[Kazeyonoma]

I'm interested in this as well.