Dual Wielding DKs: Demystifying the Myth.

[Satorri]

Before you accuse the numbers of being tweaked, go back and ready the way I came to the numbers.

Nips, the final numbers I had posted for damage increased were based on special rotations for each weapon style and spec (the numbers that included both auto-attack swings and specials). The reason Frost dual wielding came out lower is because the frost dual wielder will use IT, HB, and FS as most of their rotation, even including the OB/DS's filled in when Rime procs left them with fewer parryable attacks than Unholy single target where they're using BS, ScS, PS for most of their swings, all of which are parryable. There are debatable assumptions, but that's really not one of them.

Fonager, if you're tanking as a Frost 2-hander, the question is, at what point do you decide it's better to do HB than Oblit for the sake of limiting parry-haste. Personally, I often favor HB anyway, even if Oblit may hit a bit harder (which isn't always the case, that depends on spec, gear, state of debuffs, and target properties like armor). Also, with my frost rotation I deliberately use IT a lot, and only PS to keep the disease up. Extra IT's, which hit plenty hard on their own, also give more chances for free HBs. Generally speaking, Frost has a lot more mechanisms to avoid parry liability regardless of weapon choice.

The fact that we can use 2-handers drops our parry liability in general WAY down, and the only thing that brings us back up on par with other classes is our specials. The numbers reflected above also take into account RS. For those who aren't familiar, RS is not an in-addition next swing move, it replaces our next melee swing, and is unparryable (or dodgeable, or blockable). The major reason that dual wielding "could" come out on par or better with a smart frost rotation, is because you can make use of more Rune Strikes, reducing your auto-attack liability a fair amount so it's not THAT much worse than having a 2-hander.

The numbers aren't slanted, and they weren't back-engineered to favor dual wielding.

[Fonagor]

Fonager, if you're tanking as a Frost 2-hander, the question is, at what point do you decide it's better to do HB than Oblit for the sake of limiting parry-haste. Personally, I often favor HB anyway, even if Oblit may hit a bit harder (which isn't always the case, that depends on spec, gear, state of debuffs, and target properties like armor). Also, with my frost rotation I deliberately use IT a lot, and only PS to keep the disease up. Extra IT's, which hit plenty hard on their own, also give more chances for free HBs. Generally speaking, Frost has a lot more mechanisms to avoid parry liability regardless of weapon choice.

I expected it Should have phrased it differently but it was simply a too early "morning" post to bring in all the thoughts
But at least in dummy training way back than it resulted for me with 2h frost to be the better TPS variant. May have changed - I don't know because I'm now addicted to the Dual style

What I didn't understand properly - you use IT several times WITHIN let say a "perfect" rotation or was it just according to reality (movement, kicking, stun etc.)?

[nips]

No offence but you guys are comparing apples and oranges here. A unholy or blood tank will chose to use a 2 hander period. So comparing them to a dw frost spec is flawed at best. I know its harsh, but its the truth.

The only spec you can choose to use a 2h or dw is in frost. So the only way you can compare dw vs 2h is if you compare a 2h frost build with a dw frost build. Everything else becomes as irrelevant as comparing how much a paladin gets parried vs how much a dk gets parried.

[Krenian]

No offence but you guys are comparing apples and oranges here. A unholy or blood tank will chose to use a 2 hander period. So comparing them to a dw frost spec is flawed at best. I know its harsh, but its the truth.

The only spec you can choose to use a 2h or dw is in frost. So the only way you can compare dw vs 2h is if you compare a 2h frost build with a dw frost build. Everything else becomes as irrelevant as comparing how much a paladin gets parried vs how much a dk gets parried.

Any sane DK will realize that Blood and Unholy heavily rely on either disease or weapon damage to do a lot of the damage. In that case, a two hander will be the first choice for Blood and Unholy. Let's face it, Scourge Strike and Heart Strike are weapon damage skills.

However it is duly noted here that he has a point: You would not dual wield either in blood or unholy because the damage the trees do rely heavily on weapon damage and as such, dual wielding is at a severe disadvantage.

Furthermore, I am more than aware that the Frost tree is sincerely the Main Tank tree for DKs and as such, believe that this conversation really should pertain to single target situations.

However, I think what Sattori was trying to show when they did that example is to show that at one point, Dk dual wielding could potentially go below even most common two handed specs such as Unholy tanking. It's a comparison only of numbers to show that there is merit to the theory listed above. However, I do understand where you're coming from in the fact that Unholy tanking and Frost tanking are truly two different things, as well as the fact that really, frost is the only tree plausible enough to dual wield.

But again, i might get corrected by Sattori for it, but I believe the example given was to show the chances compared to other specs. It was a focus on the avoidance, and not simply a way to show that unholy is worse for parry situations.

Anywho, hope that clears some stuff up.

[Farothin]

Not sure how active this still is but I'm curious about the percentages listed on page one there. In my own looking at parry haste damage I got very different numbers. Using the assumption that a DK will use 50% of their GCDs on a parriable attack and a warrior will use 7/8 (revenge is parriable) a DK with a pair of 2.4 speed weapons will see slightly fewer parry hasted attacks than a typical warrior with a 1.6 speed weapon.

[Satorri]

No offence but you guys are comparing apples and oranges here. A unholy or blood tank will chose to use a 2 hander period. So comparing them to a dw frost spec is flawed at best. I know its harsh, but its the truth.

That is not necessarily true. I've seen dual wielding work perfectly fine dpsing with both Blood and Unholy as a primary tree. In every case it is a matter of how you use it, how you tailor your spec and your ability priorities.

That being said, Tony is right. I'd generally see Frost as the easiest spec to dual wield with because it has a high percentage of damage from non-weapon based moves, and many of them are also not parryable. That does not mean it is the only spec, just would be my first pick. I compared it to my current spec out of curiosity.

The contention without actually doing a simulation is that a dual wielder will always have a higher parry haste vulnerability. I wanted to see how distinct that difference really was. If you go back to my original breakdown the comparison I am most interested in is a common 2-hand Unholy tanking setup next to a Frost DW tank. There are many more factors that weren't accounted for in terms of avoidance, variability of gearing for different specs, etc, but the major assumption that flavored the calculation was that a dual wielder can use more Rune Strikes.

Based on that, and allowing for a generous variability in actual outcome, the conclusion you can come to is that they are both in the same ballpark, not that the dual wielding tank is horribly worse.

I'm not an advocate the dual wielding is better, or that all specs and gear will work equally well with any choice of weapons. I merely want to stress that there is a way in which it can be done that is not so much more dangerous than using a 2-hander which is the accepted (without real testing) only way to tank as a DK.



Read the name of the thread, the goal is to demystify, to ellucidate and dispel misconceptions, not preach one method over another.

[nips]

Oh well guess i was putting more into it than what was intended. To my defence i was realy looking for a dw frost vs a 2h frost comparing when reading your first post satorri so when it was only a dw frost vs a 2h unholy comparing i might have been a bit harsh

My point still stands though and i hope you can make a 2h frost vs a dw frost comparing in the future since i recon it's what most people will be desiding between.

As for calling the numbers tweaked... unitended or not they are indeed biased for dw. Using a 50% avoidance model benefits dw and using zero expertice benefits a frost build compared to a more parryable unholy build. Also the way you round numbers in your first calculations makes a 15% base error that benefits dw which would be okish, but when the final numbers is not rounded it becomes very wrong. Yes i know i'm looking for errors here, but as stated earlier i was putting to much thought into your post then what you inteded it to be

[lyd]

Oh well guess i was putting more into it than what was intended. To my defence i was realy looking for a dw frost vs a 2h frost comparing when reading your first post satorri so when it was only a dw frost vs a 2h unholy comparing i might have been a bit harsh

My point still stands though and i hope you can make a 2h frost vs a dw frost comparing in the future since i recon it's what most people will be desiding between.

As for calling the numbers tweaked... unitended or not they are indeed biased for dw. Using a 50% avoidance model benefits dw and using zero expertice benefits a frost build compared to a more parryable unholy build. Also the way you round numbers in your first calculations makes a 15% base error that benefits dw which would be okish, but when the final numbers is not rounded it becomes very wrong. Yes i know i'm looking for errors here, but as stated earlier i was putting to much thought into your post then what you inteded it to be

i can't go as in-depth on the math as others here can, but as for a comparison of 2h v. DW frost builds, there are a couple obvious pros/cons i can point out to you:

Killing Machine: 50% after an auto-attack critical strike to give your next IT, HB, or FS a 100% critical hit chance. this -highly- favors DW. i DW with this talent and the ae threat/dps i can generate with a 100% crit howling blasts is astounding. also, in 3.08, HB will no long have an internal cooldown. this means that with death chill, you could generate ridiculous spike dps/threat with back-to-back HB crits, not to mention spammability functioning off of Death runes and Rime procs.

Nerves of Cold Steel: a 2h-build can forego this talent, giving 2h frost tanks 3 additional talent points to play around with. of all the benefits and drawbacks of both styles, i think this is the most obvious favoring 2h.


those two are Frost-spec specific. hope that gives you a little bit of additional insight. i am DW frost right now and i love it. no complaints from healers yet

[Krenian]

Although I can see the advantages of testing out dual wield Blood (Threat a la galore) and Unholy (AoE insanity)...I find it very very hard to even think that it would be successful on this single point: The Key abilities learned are very reliant on weapon damage.

Both Heart Strike and Scourge Strike are potent Weapon Damage skills that are severely hindered by the fact you're using a one handed weapon with a 2.4 normalized. With a two hander, that number jumps to 3.5 I believe and thus, will always do more damage with the two hander. Plus, a lot of the skills in each tree benefit a two hander over two weapons.

I am not saying it's not possible that there has been some successes, but you will not see me argue with anyone about it being even closely better than using a two hander in those builds. They were made for a two hander and they should stay with a two hander. Rime is that much needed to make dual wielding a good threat generation due to it's pure damage in numbers and not in weapon damage.

Keep in mind that every time you use a weapon damage ability, that scores much further ahead with a two hander than a one hander ever will score.

The only thing I could actually think about is maybe the blood spec being dual wield on the sole purpose of going deep enough for Rime, and then going through Blood for the self healing to help your casters out. Either way, you still have to go at least 31 points into Rime, and with that in mind, I don't even think you can go deep enough to get Bloodworms which is what I could potentially see as a good way to go. (Keep in mind that it's a percentage chance that the worms would spawn per weapon hit)

http://talent.mmo-champion.com/?deat...h=000000000000

This would be a spec to be interested in trying. It focusees more on giving a lot of frontline threat, however suffers greatly as the only damage reducing ability in our repitoire is now IBF and Lichbourne. And one's on a 3 minute CD.

I left two points to the discretion of the players. You can put them in Imp Rune tap to heal more, BLood to have your raid heal every time they hit, or 80 resist in the frost tree, or even boost your strength through unholy. You can also potentially put one point in Hysteria and have a happy dpser get that.

[Lightmgl]

Honestly I can't contribute to the dual wield tank theorycraft just the DPS which I think is going to be absolutely nuts with the howling blast. If you can get 4 HBs off per 10 second rotation (Optimal, might not be realistic) that alone is a good 3.5k-4k dps. If Frost is already competitive DPS aside from buffs as I stated before even a gain of 500 or 600 dps would be enough to make it equivalent or better. Its the same reason we make rogues pick up the +2% physical damage over warriors atm cause the Fury warriors lose alot more than the rogues lose. If the DK gains enough over his alternate spec to counter the loss of hysteria or the movement time of unholy aura from the change it would make DW the better choice.

As for tanking? With the current state and ease of the game none of it matters cause you don't lose enough mitigation dual wielding to get yourself killed. The only things that would get you killed are bad healers, bad gear, or a bad tank. Sure your threat may be worse/better and sure you may lose a couple of % of average mitigation over the course of the fight but its never going to be enough to kill you over the 2h DK if all else is equal.

As I've complained about before with DK scaling our 25 man geared DK runs 65% buffed avoidance (Night Elf) and I really feel he could fight the raid bosses partially naked and still tank just fine. Theres no way him putting on a pair of swords would get him killed or even make it hard enough to require extra healing. I suppose the mages might be a bit unhappy if they pulled aggro though.

[Satorri]

Really, the stats I used above were assuming a void of expertise all around. I had better than 50% avoidance and 22 expertise (worth 5.5% dodge/parry neglect) in mostly 80 blues.

I think the heart and soul of it is that you don't have to be unparryable for dual wielding to work, and if you show a favoring to expertise on your gear you won't be looking at a massive loss of avoidance, I would wager you won't be losing more than 5-6% if that. The other part to remember is that if, for your given scenario/spec choice, using a 2-hander had double the damage increase (bearing in mind that's a 7% increase to a 14% increase which is still not "zomg instagib you idiot!"), that expertise would similarly have double the effect in reducing that liability since it reduces parry chance by %'s.

Personally I've been very dismayed at the amount of expertise I've been shedding from my tank set getting epic upgrades and I have a 2-hander. I'm less worried about parry haste (though I know it's as significant a factor) as much as the loss of hitting swings. I want to get my hands on one of those delicious 2-handers with gobs of expertise (I <3 the trash drop from Naxx25).

[Tenin]

Killing Machine: 50% after an auto-attack critical strike to give your next IT, HB, or FS a 100% critical hit chance. this -highly- favors DW. i DW with this talent and the ae threat/dps i can generate with a 100% crit howling blasts is astounding. also, in 3.08, HB will no long have an internal cooldown. this means that with death chill, you could generate ridiculous spike dps/threat with back-to-back HB crits, not to mention spammability functioning off of Death runes and Rime procs.

Killing Machine might be fine as a dps talent but is a waste of points in a tanking build. Tanks generally have a horrible base crit rating so your auto attack won't crit often(10% range here). Spending 5 talent points for a small chance for a 50% chance doesn't make much sense. Then take into consideration that your replacing you MH auto attacks with Rune Strike and Killing Machine becomes garbage for a tanking build.

My issue with a Frost DW tank is that it has no clear advantage over a 2h Frost build. A 2h frost tank build is nearly immune to parry-gib having so few attacks that are parryable. By going from 2h build to a DW wield build you're knowingly opening yourself up to boss parry. For what? The only advantage I've seen for DW tanking is becomeing crit immune easier, which is becoming a non-issue soon. Tanks don't open themselves up to take more damage, no matter how small, for no gain. That's why DW tanks take heat, and will continue to take heat, from the rest of the tanking community.

As far as Expertise as both a mitigation and a threat stat goes, as a Frost DK you shouldn't need any expertise. So few of your attacks should be parriable/dodgeable that the expertise you get from talents and the occasional tank piece should be more than enough. Instead by duel wielding your making Expertise into an important stat. Why?

Also, keep in mind that Strength is also a mitigation and a threat stat for Death Knights. Difference is that the parry rating gained from Strength works on all incoming attacks not just those generated by being parried. Expertise also only benefits those attacks that are parriable while the AP gained from Strength affects every attack a DK makes. I don't claim to know how the math works out but the question needs to be asked, which is better to stack?

[lyd]

Killing Machine might be fine as a dps talent but is a waste of points in a tanking build. Tanks generally have a horrible base crit rating so your auto attack won't crit often(10% range here). Spending 5 talent points for a small chance for a 50% chance doesn't make much sense. Then take into consideration that your replacing you MH auto attacks with Rune Strike and Killing Machine becomes garbage for a tanking build.

try it before you knock it. i assure you that it procs more than enough to make it worthwhile. i have had great success with this build, both as a tank and as dps, but if you don't feel comfortable DW tanking, then don't bother, as it requires the haste from icy talons/imp. icy talons and a DW setup to proc with any frequency. it is definitely not worth taking in a 2h build.

[GravityDK]

Hey Satorri, I can't fault the maths (I tried ,) but the layout on your post #2 here is a bit hard to read.

I think this is an important thread, and I think that's pretty solid analysis, but would you mind maybe just formatting it a bit more clearly? Going from your workings to your final comparitive conclusions is a little tough.

For example this point is really key but it's not super-clear how you got there cos its' kind of buried "End result? For purposes of survival, you're looking at DK's in the middle of a +/- 2% damage increase on bosses due to parry haste.".

eg. add in bold 'workings', 'summing all those up', and 'conclusions' kind of headers.

Secondly, could you look at Berthold's analyis here, which comes up with different conclusion (his finding was an increased wipe probability on DW DKs) and see if you can work out what premise he had that was different in order for his result to be so widely apart from yours? I'd be curious on your thoughts.

I'm still trying to get my DW DK thread on deathknight.info to reflect solid facts and knowledge and this thread here is turning out to be gold.

[lyd]

As I've complained about before with DK scaling our 25 man geared DK runs 65% buffed avoidance (Night Elf) and I really feel he could fight the raid bosses partially naked and still tank just fine. Theres no way him putting on a pair of swords would get him killed or even make it hard enough to require extra healing.

well, the only time they would require extra-healing is when a parry-hasted attack occurs. in an ideal situation, DW tanks will only generate parry opportunities from weapon-based yellow attacks and off-hand attacks, and with the changes to HB, most DW tanks won't even bother using obliterate, removing a parryable attack from their rotations that many 2h tanks will probably still use. this in of itself closes the gap even further.

i chatted with a friend over the weekend about my plans to make a DW DK tank, and he shared my view: with the removal of crushing blows, the likelihood of an honest-to-god parry gib happening, even as a DW tank, even if more likely than a 2h tank, is still extremely unlikely. no crushing blows, uncrittable, and over 50% avoidance on most DK raid tanks means that the chances of this happening are small, to the point that it should not even be a controversial issue. here is how we came to this conclusion:

basically, the average tank will have 6 expertise fom talents, meaning -1.5% parry from the get-go. so realistically, most tanks are working with 12.75%, not 14.25. so already, there is only a 12.75% chance to even proc a parry-hasted attack. this will only ever be a normal hit: tanks can no longer be crushed, and they are assumedly uncrittable. so for a parry-gib in wotlk to be set-up, you would realistically need two parry-hasted attacks back-to-back from an extremely hard-hitting boss. the chances of this happening are (.1275*.1275) = 1.625%. if we assume a generously low 50% avoidance on a raid-level DK tank, we can assume that he will, on average, dodge or parry at least one of those parry-hasted attacks, even when it occurs. so even if a boss gets two parry-hasted attacks back-to-back, he will, on average, only hit with one of them. this means that for a tank with 50% avoidance and no additional expertise, there is only a 0.81% chance of a situation occurring where a DK tank is hit by both parry-hasted attacks and in danger of an parry-gib. in order for this to equal a guranteed insta-wipe, we would have to assume alot: we have to assume that this occurs at a point where one of the DK's mitigation cooldowns is not active, particularly IBF. we would have to assume that the DK is under the effects of no external damage reduction effects, such as earthshield, discipline priest buffs, inspiration, a damage shield, etc. Further, we would have to assume that there is 100% chance that the 2 parry-hasted attacks in a row will actually kill the tank, even if he is @100% health, which in many instances is not the case. so all that considered, the chances of 2 parry-hasted attacks occurring back-to-back on a DK tank when he would actually be in danger of dying from them is even lower than .81%.

let's account for just on additional variable and see what happens: assume that the tank hits IBF everytime it's up. as of 3.08, that's roughly 40+% damage reduction for 18 out of every 60 second rotation (if you're frost). this amounts to 30% of the time. so accounting only for IBF, the chance of those 2 parry-hasted attacks occurring at a point in time where it would realistically put the tank in danger of a parry gib is .81(.7) = .567%, and again, this assumes the block of text above is true, which it almost certainly will not be in its entirety. a raiding DK tank will likely have more than 50% raid-buffed avoidance; they will likely have more than 6 expertise: they will likely utilize their other mitigation talents; they will likely be under the effects of external damage reduction effects a good portion of the time; so the already low percentage goes down even further, to the point where it would take an enormously bad stroke of luck for any tank to die due solely to their vulnerability to this mechanic, even a DW DK tank. so the margin of success vs. failure due to DW tanking vs. 2h tanking given everything else being equal is so small as to require a stroke of very bad luck to occur, and one in which a 2h tank would likely succumb to as well.

as before, so again, it really just comes down to preference. the parry-gib boogeyman is really not worth discussing at this point given it's relative likelihood and the fairly unlikely circumstances that would have to be present for the -chance- for it to happen to even manifest itself. i don't foresee it being a dealbreaker for any DW tanks or their guilds in the future.

[Satorri]

Secondly, could you look at Berthold's analyis here, which comes up with different conclusion (his finding was an increased wipe probability on DW DKs) and see if you can work out what premise he had that was different in order for his result to be so widely apart from yours? I'd be curious on your thoughts.

So, the long and the short of his simulation is that it is hardly in depth, though it's a lot more regimented and automated than my synopsis. What I did above was off the cuff, very cursory and a lot of rounding. If there's enough interest I could do several scenarios and post the math with a lot more clarity. That post just turned into a lot larger of a run down than I had intended.

As for Berthold's simulator, the problem is it's a passable thing for a Warrior, but it neglects some very key features that DK's have and Warriors do not. Warriors use all physical/melee abilities. With the exception of Revenge all of which are subject to parrying, and their auto-attack does not change so it is a constant baseline for that (and a noticeable one since Warriors traditionally use fast fast tank swords ~1.5 sec swing speeds).

DK's have several complicating features that make it a lot harder to simulate casually. Rune Strike is unparriable, and replaces a melee swing (reducing the baseline vulnerability). It also causes DK avoidance to have a double or half-again impact (your avoidance saves you one hit and your next melee swing is not vulnerable to parry haste). On top of that, depending on the spec, weapon choice, and even just the player's in-combat play, you use a mix of parriable melee moves and non-parriable spells. In my admittedly cursory consideration I still noticed a very distinct part that these abilities play in the parry haste equation.

If you want a true representation you'd need to take a specific spec and a functional, realistic, rotation accounting for player variability for moves used. I maintain the best DK tanks are ones who adapt on the fly using priorities, not ones who try to adhere to a rotation.

So, like I said, if there's interest I can do out some better formatted and more diligent calculations for particular builds and weapon styles.

[Schyne]

Although I can see the advantages of testing out dual wield Blood (Threat a la galore) and Unholy (AoE insanity)...I find it very very hard to even think that it would be successful on this single point: The Key abilities learned are very reliant on weapon damage.

Both Heart Strike and Scourge Strike are potent Weapon Damage skills that are severely hindered by the fact you're using a one handed weapon with a 2.4 normalized. With a two hander, that number jumps to 3.5 I believe and thus, will always do more damage with the two hander. Plus, a lot of the skills in each tree benefit a two hander over two weapons.

I am not saying it's not possible that there has been some successes, but you will not see me argue with anyone about it being even closely better than using a two hander in those builds. They were made for a two hander and they should stay with a two hander. Rime is that much needed to make dual wielding a good threat generation due to it's pure damage in numbers and not in weapon damage.

Keep in mind that every time you use a weapon damage ability, that scores much further ahead with a two hander than a one hander ever will score.

The only thing I could actually think about is maybe the blood spec being dual wield on the sole purpose of going deep enough for Rime, and then going through Blood for the self healing to help your casters out. Either way, you still have to go at least 31 points into Rime, and with that in mind, I don't even think you can go deep enough to get Bloodworms which is what I could potentially see as a good way to go. (Keep in mind that it's a percentage chance that the worms would spawn per weapon hit)

http://talent.mmo-champion.com/?deat...h=000000000000

This would be a spec to be interested in trying. It focusees more on giving a lot of frontline threat, however suffers greatly as the only damage reducing ability in our repitoire is now IBF and Lichbourne. And one's on a 3 minute CD.

I left two points to the discretion of the players. You can put them in Imp Rune tap to heal more, BLood to have your raid heal every time they hit, or 80 resist in the frost tree, or even boost your strength through unholy. You can also potentially put one point in Hysteria and have a happy dpser get that.

Tony, in looking at your spec, and being a DW DK tank and DPS myself, I have to tell you that I almost NEVER use Obliterate with Dual weapons because it just doesn't do the damage, and it's a waste of runes unless it's on a Rime proc. What I would suggest is putting the 3 points from Annihilation into Black Ice, since most of your damage will still come from frost spells like Icy Touch and Howling Blast, then a large portion from Melee white damage.

In DPS spec, my white melee damage is 30% of my total. That may not look like a lot, but consider my hit is around 315, and I miss 15% of the time with melee swings - but I'm special hit capped, so none of my spells ever miss. If I had 400 hit, I would miss about 8% of the time, and would have a full 35% of my damage done as white damage. That's nothing to scoff at, especially when swinging two weapons.

In a tanking spec, my hit is around 150. I miss about 18% of my melee swings, and about 8% of my spells are misses. In unholy spec, with 9/9/53, once I pop Unholy Blight, no one can pull off me... I don't care if the warlock in the group is doing 5k DPS AoE on seeds... I have agro. I've also never had a problem holding threat in Blood spec tanking, though I've never attempted that with Dual Wielding. Frost is actually my nightmare tanking spec. Any time I've ever specced frost to tank, I suck at threat production, damage done, and more than likely, the group will wipe a few times and I'll go respec back to Unholy to tank the rest. I've tried frost tanking with both a 2h Sword and Dual 1h Swords.

If you check out my armory, you'll see that I'm swinging 2 Silent Crusaders, and they are fairly decent for both DPS and tanking, though I wish they had a single yellow socket on them so I could be Raid Defense capped when using them. I have yet to get the trinket from HoL or Naxx... never seen either drop yet.

Anyway.... I'm DW specced right now, and it's not only fun, but viable. I can offtank in DPS gear fairly well when the SHTF, and I consistently top every raid's DPS chart. When I tank, I lose about 1000-1250DPS overall, but I'm still making 3000TPS with two swords. Maybe if I end up respeccing to run HoL all day tomorrow, I'll post my tanking spec for you guys to see.

Excellent thread, I've taken a lot from it to build my own tanking set - which is severely lackluster right now - and I will continue to use it until I stop tanking stuff.

[A Man In Black]

This thread fails to answer the two key questions about DW tanking as a DK, and its math and modeling are poor.

The first post, for some reason, assumes that 2h tanks will be Unholy, which happens to have many more weapon strikes than Frost. Obviously a spec whose spammed attacks are parryable weapon strikes is going to have more parryable attacks than Frost, which spams unparryable magic attacks.

I would like to try the OP's napkin math again, with the assumptions laid bare and the math a bit more visible. I'll start by paring the analysis down to parryable attacks per minute. Both tanks are buffed with Imp WF/Imp Icy Talons, so they both have 20% increased attack speed.

A DW tanking DK is, if he's lucky, going to be wielding two Broken Promises, 2.5 speed weapons. (There aren't two worthwhile slow epic tanking weapons, so this is a best-case scenario.) That's 57.6 parryable attacks per minute from autoattacks. A 2h tanking DK at a comparable gear level is going to be wielding Inevitable Defeat, generating 21.2 parryable attacks per minute from autoattacks. (Both of these numbers incorporate attack speed buffs, but not the DK's own parryhaste from parrying the boss's attacks because this is complex enough, I figure.)

And then, since they're both Frost, their rotations are exactly the same, both before and after patch. 2h tanks will not use Obliterate if it is a significant risk to themselves through parry-gibbing. They will obviously use Howling Blast if it is advantageous for them to do so.

IT > PS > HB > BS > BS, HB > HB > HB

Both specs do 3 parryable strikes per 18 seconds (assuming optimal use of runes, but whatever, we're already in Target Dummy Who Hits You Back land.) So that's 30 parryable strikes from runes per minute, and obviously none from RP dumps.

So, now we're at:
DW: 87.6 parryable attacks per minute
2H: 51.2 parryable attacks per minute

Now, Rune Strike opportunities. Let's dispose of the assumption that boss misses are evenly distributed. (We're stuck with the awful assumption that the boss's rhythm will be perfect due to no parryhaste, the kind of lousy assumption that the only attack you can avoid is the boss's autoattacks, and the less horrible assumption that you'll have RP to immediately use every opportunity.)

DW frost has a cycle looking like this, with their 2.08333... MH swing speed:
0.0: Boss: 50% chance for a miss
0.0: DK: 50% chance for RS
2.0: Boss: 50% chance for a miss
2.1: DK: 50% chance for RS
4.0: Boss: 50% chance for a miss
4.2: DK: 50% chance for RS
[cycle repeats]
40.0: Boss 50% chance for a miss
41.99...: DK 50% chance for RS
42.0: Boss 50% chance for a miss
44.0: Boss 50% chance for a miss
44.1: DK 75% chance for RS

So, over 20 mainhand swings and 42 seconds, 19 swings have a 50% chance to be a RS, and 1 swing has a 75% chance to be a 75% chance to be a RS. (The start of every cycle except the first cycle, due to the pull, has this 75% chance to be a RS.) So, that's 10.25 RSes over every 42 seconds, or 14.6 RSes every minute.

Now, the 2H cycle, with their 2.833... attack speed.

0.0 Boss 50% chance for a miss
0.0 DK 50% chance for RS
2.0 Boss 50% chance for a miss
2.8 DK 50% chance for RS
4.0 Boss 50% chance for a miss
5.7 DK 50% chance for RS
6.0 Boss 50% chance for a miss
8.0 Boss 50% chance for a miss
8.5 DK 75% chance for RS
10.0 Boss 50% chance for a miss
11.3 DK 50% chance for RS
12.0 Boss 50% chance for a miss
14.0 Boss 50% chance for a miss
14.2 DK 75% chance for RS
16.0 Boss 50% chance for a miss
17.0 DK 50% chance for RS
18.0 Boss 50% chance for a miss
19.8 DK 50% chance for RS
20.0 Boss 50% chance for a miss
22.0 Boss 50% chance for a miss
22.7 DK 75% chance for RS
24.0 Boss 50% chance for a miss
25.5 DK 50% chance for RS
26.0 Boss 50% chance for a miss
28.0 Boss 50% chance for a miss
28.3 DK 75% chance for RS
30.0 Boss 50% chance for a miss
31.1 DK 50% chance for RS
32.0 Boss 50% chance for a miss
34.0 (exact repeat of beginning)

So, in 12 swings and 34 seconds, the 2h DK has 8 swings with a 50% chance to be a RS and 4 swings with a 75% chance. That is, on average, 7 Rune Strikes every 34 seconds, or 12.4 Rune Strikes per minute.

After Rune Strikes, we have:
DW: 57.0 parryable attacks per minute
2H: 38.8 parryable attacks per minute

So. The cost of dual wielding is 18.2 additional parryable attacks per minute, or a 31.9% increase in parryable attacks, assuming the same DK switches from an Inevitable Defeat to a brace of Broken Promises and assuming that Howling Blow has no cooldown in 3.0.8. This number scales upward and downward with expertise until you are fully capped against parry. No amount of expertise negates the percentage increase in parryable attacks until you have zero parryable attacks.

I will leave it to others to calculate the danger of a 31.9% increase in parryable attacks. What are you gaining for this cost? (A salient Wowhead item comparison.)

Does the increased avoidance/mitigation from two traditionally-itemized one-handed tanking weapons overcome the parry-haste disadvantage of the additional melee attacks, given the least-risky rotation and spec for both? Does the increased talent investment of DW frost into talents to prop up dual-wielding come at the cost of other useful talents?

Does dual-wielding offer a threat increase over 2h tanking? Does it offer a threat increase while wielding two tanking weapons, despite the itemization disadvantage? Is there a hit breakpoint of some kind?

(Other classes' attack speeds are moot, as they have tools that Death Knights do not; in addition, the OP's modeling of those classes is Not Very Good. Presumably, they are balanced on a certain amount of parry-haste increasing their natural damage intake.)

[GravityDK]

woohoo we've got a maths session happening! Looks right at first glance, AMiBlack.

[visualdeity]

AMiB, your math looks sound, but you make your numbers out to be bigger than they are. Slapping together some gear, gemming it, and enchanting it, using chardev (linky, if interested), I come up with 43 expertise. I'm unsure if the math on chardev is accurate (and have been too lazy to check its calculations on the ratings myself), so that could be wrong. I have gemmed blue sockets with stam, and red sockets with expertise, which I feel is generally a sensible gemming strategy for 2h or DW. I also am unsure if the 43 expertise figure calculates in the human racial for s-words, so let's throw that 3 expertise out (and hell, it gives us a nice round figure, which I always enjoy). I did not go out of my way to pick up expertise gear, I generally just took whatever seemed to me to be the best in the slot (although I did take full T7, because it looks damn cool. Girlcraft, I know, but sue me. )

This leaves us with -10% chance to be parried. Now, I've heard varying claims for boss parry rates, ranging from 12% to 16%. I'm going to go with 15%, as that's the one I seem to see cited most often. So, we have 5% chance to be parried. You have said (and until someone else disproves them, I'm willing to trust your numbers) that a dual wielding frost DK will generate 18.2 more parryable attacks per minute than a 2h frost DK. However, only 5% of those will be parried, so your actual parried attacks (assuming RNG fairness, which I know isn't guaranteed) goes up by .91 per minute. The biggest gap in my reasoning here is the fact that I'm trusting chardev to calculate the expertise correctly, but assuming that to be correct, I hardly feel we should get all excited about one more parry per minute. That's not going to kill anyone.

Also, I think that it is rather unfair of you to criticize Satorri for using an unholy DK as her 2h comparison, because let's face it: most DK tanks are unholy at the moment, due to Bone Shield being so good. This may change in 3.0.8, but it may not either. That's anyone's guess. At the moment, though, unholy is the status quo.

Finally, you say that we shouldn't compare to other tank classes, as presumably, they are balanced around their increased parry generation. This is faulty reasoning, however, as Ghostcrawler has directly said that they want death knights to be able to effectively tank in either style, and their hope is for there to be both dual wielding and 2h tanks. Now, you might say that the balance isn't currently in place to make that possible, and that may be the case, which is why we're discussing the matter. However, if we are willing to assume that DKs might not be balanced for the kind of parry generation DW provides, we can't assume the other classes are balanced for a similar level of parry generation. If we doubt one class, we need to doubt all.