I think five things are wrong with the napkin math:
1) Assumption of only 20% white damage is a low estimate. I think it's closer to 24%, which will increase haste's value 20% over your estimation.
2) Not including haste's impact on deep wounds damage is a huge mistake. About 27% of haste's value comes from additional DW procs. 1/(1-.27) shows that including this would increase your estimate of haste's value by 37%.
3) Haste's loss of value due to slam is being overestimated. Slam is only used about once every 8.75s by my sim's estimate. This might seem like a longer delay than you'd expect, but Slam isn't used during execute range, which raises its average time between uses quite a bit. (.5/8.75) = 5.7% loss of value due to slam cast times, a far cry from 15%, which is just absurd and implies you're using Slam every 3.3s, or about every other GCD the entire fight.
4) It fails to take into account increased execute damage. Faster swings will certainly increase damage in execute range through increased rage generation/shorter delays between hits and is about 10% of haste's value(11% increase over the estimate)
5) Strength has 0 effect on proc uptimes, while haste will increase Greatness/MoT/T8 2pc/Zerk uptimes by a small amount each. This is probably its weakest impact, resulting in about a 1% increase in value.
Haste is still bad compared to strength, but it's certainly not 1/4 of its value. It comes out to .5-.55 SEP by my simulations, which seems perfectly reasonable.
The odd thing is that haste given a 5m fight has a rather strange impact on your dps. Taken in big chunks gives some results, but little numbers often don't provide any real dps until you stack enough to make another hit before the boss is dead. Nor in a 5 min timeframe you wouldnt notice any rage generation increase (or higher proc uptime) unless happens:
a) you pass from 3:1 to 2:1 wing per GCD (around 3s if you include slam delay)
b) you stack enough haste to execute every gcd+lag for the whole <20% phase.
I find haste to be sometimes 3/5rd and sometimes even less than 1/4rth of STR (since all haste you stack until you can effectively hit once more in a fight is completely wasted). It's average value is probably less than 1/4rth but unless we're talking about 8k AP I doubt it's just half of STR.
The haste value is then very different from STR since it's a sort of sinusoidal damage increase.
STR has a more reliable impact on damage regardless of the length of the fight. It is anyway very difficult to calculate its real contribution in therms of damage %. You have to examine the 2 curves of scaling and compare them at discreet points.
In general the effects of haste tend to gein a good SEP value on very long simulations and very variable SEP values on short fights.
Slam usage in simulators tends to be a bit lower than in raid due to human and lag factor. The opposite appears to happen to HS since a simulator can decide to HS directly when the swing happens with 0s reaction time.
I may be wrong, but I don't see in short fights an average value of haste higher than 30% of STR, with only few isolated points where you get a spike in SEP due to 1 more hit landed.
I will try to draw some graphs of how haste SEP increases with AP in a given short fight and in the long range.
This probably can summarize better why I believe haste is not so good as it seems when compared to STR or ARP or AP that in the short therms have a more stable value.
The odd thing is that haste given a 5m fight has a rather strange impact on your dps. Taken in big chunks gives some results, but little numbers often don't provide any real dps until you stack enough to make another hit before the boss is dead. Nor in a 5 min timeframe you wouldnt notice any rage generation increase (or higher proc uptime) unless happens:
a) you pass from 3:1 to 2:1 wing per GCD (around 3s if you include slam delay)
b) you stack enough haste to execute every gcd+lag for the whole <20% phase.
I find haste to be sometimes 3/5rd and sometimes even less than 1/4rth of STR (since all haste you stack until you can effectively hit once more in a fight is completely wasted). It's average value is probably less than 1/4rth but unless we're talking about 8k AP I doubt it's just half of STR.
The haste value is then very different from STR since it's a sort of sinusoidal damage increase.
STR has a more reliable impact on damage regardless of the length of the fight. It is anyway very difficult to calculate its real contribution in therms of damage %. You have to examine the 2 curves of scaling and compare them at discreet points.
In general the effects of haste tend to gein a good SEP value on very long simulations and very variable SEP values on short fights.
Slam usage in simulators tends to be a bit lower than in raid due to human and lag factor. The opposite appears to happen to HS since a simulator can decide to HS directly when the swing happens with 0s reaction time.
I may be wrong, but I don't see in short fights an average value of haste higher than 30% of STR, with only few isolated points where you get a spike in SEP due to 1 more hit landed.
I will try to draw some graphs of how haste SEP increases with AP in a given short fight and in the long range.
This probably can summarize better why I believe haste is not so good as it seems when compared to STR or ARP or AP that in the short therms have a more stable value.
Having an unstable value does not diminish its real value. You could just as easily apply that same line of reasoning to crit and say it's not useful because its SEP can differ wildly from fight to fight since there's absolutely no guarantee whatsoever that you'll get any more crits than before. Heck, crit is even more unstable than haste since it merely increases your probability of seeing a return while haste's returns are very predictable. That's why an average return is more meaningful than the tiny sample size of a single 5 min fight.
Also, I have no idea what you're talking about with the 3:1/2:1 GCD/swing thing. The number of swings per GCD is not important at all until you reach a swingtime lower than GCD+latency, at which point(which is 106% haste @ 150ms with a 3.4spd wep, basically will never happen gearing arms appropriately) haste slightly diminishes in value due to no longer increasing the frequency you can get executes off at. Up until then, there are no swings/GCD breakpoints that matter.
Lastly, stacking enough to reach 1 extra hit is not a high value at all. A 5 min fight with BF/WF/Swift Retribution/3.4 spd wep = 28 haste for an extra hit(30 haste after slam delay penalty). That's also completely ridiculous to attempt to calculate, since those exact thresholds will vary wildly from fight to fight depending on exact fight times. An average number of gained swings is a more reliable approximation.
In the end, it won't really matter what haste's value is since arpen/str/crit are so much better that it'll basically be impossible for any item with haste on it to be competitive regardless of whether it was 0 SEP or .5
I think five things are wrong with the napkin math:
1) Assumption of only 20% white damage is a low estimate. I think it's closer to 24%, which will increase haste's value 20% over your estimation.
On a boss fight, especially a stationary one, if your white damage is 24% that just means you are not filling in all your global cooldowns properly which could be due to rage starvation or just general laziness/slowness hitting your buttons buttons. I have never had my white damage that high whether with crappy blues early on or the close to BiS i have now.
2) Not including haste's impact on deep wounds damage is a huge mistake. About 27% of haste's value comes from additional DW procs. 1/(1-.27) shows that including this would increase your estimate of haste's value by 37%.
Deeps wounds should be up 100% of the time with or without any haste. If you see it fall off now you have bad RNG and one extra white hit every half minute isn't going to increase your chances of keeping DWs up much against the other 25 hits (or whatever it is) you got in that time. When 3.1 comes out you have a overpower every six seconds which pretty guarantees DWs always being up even if you don't have 50% raid buffed crit.
3) Haste's loss of value due to slam is being overestimated. Slam is only used about once every 8.75s by my sim's estimate. This might seem like a longer delay than you'd expect, but Slam isn't used during execute range, which raises its average time between uses quite a bit. (.5/8.75) = 5.7% loss of value due to slam cast times, a far cry from 15%, which is just absurd and implies you're using Slam every 3.3s, or about every other GCD the entire fight.
I agree its loss of value is overestimated in this case (doesn't mean haste still is not garbage). Slams however can be very well used in execute range. If MS/OP/rend is not ready and you do not have a SD proc you could use slam. Using execute too early can leave you waiting for your next white hit. Your rotation should stay almost the same in execute range but you can time in extra executes if you know a white hit is going to come before your next GCD is ready.
4) It fails to take into account increased execute damage. Faster swings will certainly increase damage in execute range through increased rage generation/shorter delays between hits and is about 10% of haste's value(11% increase over the estimate)
As i somewhat said above, execute spam is not the best dps.
5) Strength has 0 effect on proc uptimes, while haste will increase Greatness/MoT/T8 2pc/Zerk uptimes by a small amount each. This is probably its weakest impact, resulting in about a 1% increase in value.
Strength however can have an effect on how much damage a proc adds (procs that add crit or armor penetration rating for example). With internal cooldowns and just thinking about how many hits we average per second, haste's effect on procs is negligible.
Haste is still bad compared to strength, but it's certainly not 1/4 of its value. It comes out to .5-.55 SEP by my simulations, which seems perfectly reasonable.
Just practice in raid, simulations just tell us what kind of damage they think we should be doing. They can average out how many procs we will get on average but for an arms warrior it is more important what order our abilities proc. A sim cannot predict this and it has a major impact on our dps. I personally hate sims; they can't really help me be a better raider or dpser because most things need to be figured out in raid as every fight is different. I really don't care how much damage I can do on a dummy.
Deeps wounds should be up 100% of the time with or without any haste. If you see it fall off now you have bad RNG and one extra white hit every half minute isn't going to increase your chances of keeping DWs up much against the other 25 hits (or whatever it is) you got in that time. When 3.1 comes out you have a overpower every six seconds which pretty guarantees DWs always being up even if you don't have 50% raid buffed crit.
Uptime of DW was not what he was talking about. But the number of crits (which define DW damage).
Though I do not see how haste does benefit DW damage that much. The number of crits may be higher, but the DW base dmg is less. Shouldn't this be a wash?
Uptime of DW was not what he was talking about. But the number of crits (which define DW damage).
Though I do not see how haste does benefit DW damage that much. The number of crits may be higher, but the DW base dmg is less. Shouldn't this be a wash?
The DW part of damage haste can provide is something we should factor in haste relative value, I don't think Rallik meant our DW will do 37% more damage.
I guess the DW contribution from haste must be something like:
hitsGainedFromHaste*critPercent*0.48WD splitted into any DW tick or added as total DW damage.
So the DW base damage should be more since it's stacking and not only increasing on base damage:
1 more crit over let's say per minute will add 0.48WD per minute supposing a 100% DW uptime it's an average of 0.8% DPS for DW. Anyway this value gets more significant the longer the fight.
Originally Posted by Buentoe
I personally hate sims; they can't really help me be a better raider or dpser because most things need to be figured out in raid as every fight is different. I really don't care how much damage I can do on a dummy.
Simulating human factor is probably the most difficult thing to do so we just have to rely on personal experience (since human factor is different from player to player and depending on lag too).
This doesn't mean a simulator can't help you gear or understand the mechanics.
There is something I mentioned and I maybe didn't express well about haste "steps".
Passing from 1 hit after 3rd gcd to 1 hit after 2nd gcd makes a big difference for a human, nearly none (or very few) for a simulator. It lets you predict your ability to keep the gcd flow stable especially when you have to make decisions that can leave you starved.
In general what I don't like of haste (so consider this very personal ) is that compared to STR or ArP has a very different behaviour in scaling especially on <5min fights.
It's similar to crit as pointed already, but contrary to crit doesn't rely on RNG.
While in the average crit can even be worse than haste, I know I equip it as a "bet". Haste is much better if your white+hs sample (not only%) is higher and with a 2h this can only happen in long fights.
I generally have to consider haste in therms of "what I give up" from other stats, and I often find this loss is either AP or ArP or Crit, all better stats. This can be the reason why my perception is that haste gives a so low increase.
Long time reader, first time poster. I am one of the 2 Main Tanks in my guild, and I've been looking for a dps Offspec for 3.1, I've always liked warrior dps mechanics anyways. I've been messing around on the PTR and I think I may have a new spec.
Don't laugh.
27/31/8 +5 Talent Calculator - World of Warcraft
With my extra points in Weapon Master, Strength of Arms and the last point in Imp Execute, but you can opt for Intensify Rage, Imp Berserker Stance and whatnot if it shows higher returns, that's what I'm here to find out.
Warriors have one the highest, if not THE highest return on crit in the game. Impale + Deep Wounds is a seriously Potent combo. To counter-balance this, we have low crit gain rate from crit rating and agility. The most efficient way to get crit up for warriors is through talents, which is why Cruelty looks better than it already is. Throw in Incite + Axe Spec, and we're getting some pretty high crit numbers. Incite + fast weapon (1h suffices) + Heroic Strike and we start to see the blood flow.
My gear is decent, things I've picked up from raids, a lot of items somewhere in the 3rd BiS neighbourhood (except my rather embarassing weapons), I hit cap and expertise cap myself based on spec, I know warrior mechanics. Spec'd as Arms on the PTR, I peak at 2.7k dps and average 2.5k dps. Fury is similar at 2.6k peak and 2.4k average. With my new spec, I get betweek 2k and 2.5k depending on my setup and rotation. That's scarily close to the top 2 dps specs we have now, considering I'm using Last Laugh and The Key (level 80 3.6 1h axe for those that haven't heard of it). All tests were conducted solo on a lone level 80 target dummy with the same gear (minus weapons and maybe slight tweaks for caps) and under 3 minute controls.
You can use 2 rotations with this spec with similar success: Battle and Berserker Stance basically. Battle goes for Bloodthirst > Rend > OP, with Heroic Strike as much as possible obviously, while Berserker similarly goes for Bloodthirst > Whirlwind. I get higher dps with Battle Stance atm, by 150-300 dps.
My dps percentages come back almost exactly the same every time, 25% melee, 25% heroic strike, 25% deep wounds, and the rest is split between the rotation being heavily skewed to Bloodthirst (10-15%). During my tests, I found that which axe I had Main Handed matter relatively little, about 100 dps difference in favour of The Key Main Hand. This leads me to believe that Slow Main Hand is better since Last Laugh has over 40 dps of an advantage over The Key.
Yes, I realise that there are no dps 1h axes in the game. Last Laugh is the highest dps 1h axe now AND in 3.1, save
PvP weapons. However, there has already been talk of The Furious 2h Axe potentially being Best in Slot should Yogg-Saron nor Algalon drop a 2h Axe of their own. So maybe the PvP 1h axes will be good enough. Maybe Yogg-Saron or Algalon will drop our first dps 1h axe. (Here's hoping)
There are a lot of things to consider here: Slow Main Hand vs Fast Main Hand. Does the increased frequency of Heroic Strikes outweigh the higher damage on Rends and Overpowers? For Fury, I'd guess that Fast Main Hand is better since Whirlwind is such a low percentage of damage in the rotation ( less than 10%) but for Battle, the mixture is muddled. What about threat generation? That's a lot of heroic strikes going out.
Another strong point of the build is stat scaling. It has very strong scaling, something that makes Titan's Grip Fury so potent, and, IMHO, is holding Arms back. Strength obviously scales well, through not as quite well as Titan's Grip. Crit again is obviously good, and we don't need thresholds for it. Armor Penetration is the new stat to shoot for with all specs, and this one is no different. Haste is very strong, miles ahead of Arms and I think ahead of Titan's Grip, and that's saying something considering haste is pretty good for dual wielding 2hers. Hit past cap, while not good, isn't as bad as Arms, and again comparable to Titan's Grip. Expertise is good in the sense that it's flexible, you can adjust it based on your current gear by shifting points from Weapon Mastery/Strength of Arms. The build should also scale very well with raid buffs.
If anyone else wants to try the spec out, I'd love to hear from you. Maybe better gear will bring different results. And please, flame me and shoot me down, I want to know what I'm doing wrong and try to rectify it. (I know that wacky specs sometimes can't be fixed, but I'm willing to try.) Maybe I should post this on the Fury dps forums too?
Thank you for reading and keep up the excellent research and work.
Brief and simple:
You might be able to pull 2.5k dps with this spec and the cookie cutter spec on a dummy but when you get around 5k dps in raids going deep arms will be much more beneficial. Overpower and rend are the strongest dps abilities any specced warriors has (most damage per global cooldown on average). Your rend, overpower, and deep wounds damage will be reduced with a 1her. Flurry is far better for a fury warrior that has a very large part of his damage as HS or white hits. On average an arms warrior will do about 20% white damage so flurry does not scale as well. You also miss the burst AoE damage from sweeping strikes and bladestorm. Deathwish will not do more damage than wrecking crew and bloodthirst will do about the same damage as MS. More HS will just eat up rage faster and your rage gen im pretty sure would not be higher than a deep arms or fury spec. In an unlimited rage situation this spec could be decent if you could hold down the HS button, but then again i would rather be able to hold down the HS button with a 2her.
I see your point about scaling in raids, and I agree that Titan's Grip would definately scale better in raids. However, I think some of your points are a little off.
First off, I'd like to restate that this spec Dual Wields One Handed Axes.
Deep wounds damage is irrelevant of weapon type or speed except instant attacks. This spec delivers most of its damage from non-instant attacks, that is to say, auto-attacks and Heroic Strikes, which make up 50% of its damage, much more than the 20% or so of an arms warrior who may or may not even use Heroic Strike. You said, "Flurry is far better for a fury warrior that has a very large part of his damage as HS or white hits." which fits perfectly into what I'm trying to accomplish here.
Secondly, Mortal Strike will do FAR less damage than Bloodthirst, even with a 2h, never mind 1h. The Burst AoE damage you're dead on about, I'll admit, although in the defense of the spec, it does have the increased Thunder Clap and Cleave, which may help a little.
This spec generates a lot of rage, and burns it pretty efficiently. Overpower and Rend are both cheap, Bloodthirst is the most expensive, and Heroic Strike with 1h'ers (fast especially) is extremely efficient. Glyph of Heroic Strike really shines in this build as well. Also, why would you prefer to Heroic Strike with a 2her over a 1her? The bonuses it gives are the same (improved crit returns and no glancing) except that the extra 500 or so damage gets applied more often with the faster 1her.
I appreciate the quick reply and feedback, and who knows, I might be missing something and look like a moron, but I urge you to try it out. It plays suprisingly better than it looks on paper.
EDIT: Forgot to point out that you're also right about Rend and OP doing less damage, but there's nothing else to do but wait for Bloodthirst. I did try using only Bloodthirst and dumping all excess rage into Heroic Strike, but it didn't do quite as well. Any damage is better than no damage, and the damage isn't THAT terrible.
I see your point about scaling in raids, and I agree that Titan's Grip would definately scale better in raids. However, I think some of your points are a little off.
First off, I'd like to restate that this spec Dual Wields One Handed Axes.
Deep wounds damage is irrelevant of weapon type or speed except instant attacks. This spec delivers most of its damage from non-instant attacks, that is to say, auto-attacks and Heroic Strikes, which make up 50% of its damage, much more than the 20% or so of an arms warrior who may or may not even use Heroic Strike. You said, "Flurry is far better for a fury warrior that has a very large part of his damage as HS or white hits." which fits perfectly into what I'm trying to accomplish here.
Secondly, Mortal Strike will do FAR less damage than Bloodthirst, even with a 2h, never mind 1h. The Burst AoE damage you're dead on about, I'll admit, although in the defense of the spec, it does have the increased Thunder Clap and Cleave, which may help a little.
This spec generates a lot of rage, and burns it pretty efficiently. Overpower and Rend are both cheap, Bloodthirst is the most expensive, and Heroic Strike with 1h'ers (fast especially) is extremely efficient. Glyph of Heroic Strike really shines in this build as well. Also, why would you prefer to Heroic Strike with a 2her over a 1her? The bonuses it gives are the same (improved crit returns and no glancing) except that the extra 500 or so damage gets applied more often with the faster 1her.
I appreciate the quick reply and feedback, and who knows, I might be missing something and look like a moron, but I urge you to try it out. It plays suprisingly better than it looks on paper.
EDIT: Forgot to point out that you're also right about Rend and OP doing less damage, but there's nothing else to do but wait for Bloodthirst. I did try using only Bloodthirst and dumping all excess rage into Heroic Strike, but it didn't do quite as well. Any damage is better than no damage, and the damage isn't THAT terrible.
I'll be back after work to check on the thread.
I tried few DW arms spec on PTR and I can tell you that it is not that great at first glance.
The best DW spec I tried was full arms, cause of SD and BF (that 's 6% perma haste).
Even Bladestorm is a great increase considering WW is probably your highest damage move.
Wreacking crew is extremely better than any increase you get from BT over MS and DW (especially with no intensify rage).
Now consider you give up much with a DW arm spec (that is anyway comparable or even better than a fury DW with no tg): DW damage is normalized that means very fast weapons will deal a lot less deep wounds damage. You can balance this with more procs but you need something like 60% faster speed with a slow 1h.
HS will scale better with haste as this spec would probably favour it by quite a lot, OP and MS are normalized to 2.4 so you will lose a bit from the weapon damage difference and 37.5%sh of AP contribution.
You have no weapon damage modifier (6% becomes 6.42% with all other dmg increase buff debuffs), your special:white ratio will decrease reducing the contribution of impale.
Rend damage will be drastically reduced but you can probably keep it up 100% without wasting a single op proc time.
The real issue with the new OP is that even at 2.4s with 20% damage and 1s gcd it is like 25%sh better than WW so you have to dance quite much before being able to do quite good DPS.
Keeping bladestorm on CD, glyphing for Bladestorm, Hs, and WW.
Depending on how much crit you have a slow-fast or a fast-fast setup can be preferable.
Theorically if you have very high crit deep wounds is still a big chunk of your damage and going for a fast MH can give detrimental results. Anyway atm there are too few MH options with slow speed and if you try Last Laugh against SoH axe the big difference will hide the results.
Hit rating has a surprisingly higher value than fury builds since it increases exe procs and can convert more gcds into executes.
DW fury builds are more or less like TG builds but scales a lot less. You probably need far too much AP to compensate the weapon damage difference.
However 2h specs are definetly superior to both DW builds and scale much better so I guess DW is not an option yet.
In fact due to powerful nature of ArP (and socketing ArP everywhere), BiS Ulduar gear only shows around 3600 AP unbuffed for me now. Using Landsouls spreasheet to model procs (its for fury i know, but it can approximate it pretty well), It shows 6200 average AP with raid buffs.
Hmm that's not bad, I have about a bit over 3600 now in basically full BiS (wtb SNR) with str gems. What set are you getting as BiS so far? A lot of the pieces are a bit awkward, been hard to come up with a good balance without overshooting hit and undershooting exp while still going for the 4 piece T8.5.
Did the full arms spec you try have Incite? I'm convinced that Heroic Strike is the key to DW builds atm, which is why I would even consider a fast Main Hand. If you focus your efforts on buffing something that has great support, while not neglecting other areas, you achieve great results. Heroic Strike is something of an Ace in the Hole for warriors. That is to say, no matter your gear level, you WILL resort to Heroic Strike once you get flooded with rage/GCDs, and it will always provide positive results for proper use. If we maximise it, I think it will show on the dps charts. The only thing stopping Heroic Strike from being completely OP is that its damage bonus component doesn't scale with your stats. But we can increase its crit chance by 15%, decrease rage cost 2 ways, and it scales with EVERY warrior dps stat.
I'm concerned that your Deep Arms build swapping stances burns a lot of the potential Heroic Strike rage, but then again, I haven't tried it. I'm sure that with Flurry, Windfury Totem, and Haste on gear, you can get an almost permanent 60% haste on your auto attacks. That's scary for rage generation and potential Heroic Strike uses.
Playing around with the talent calculator, I can see how a Deep Arms DW build could be potent. Again, I don't know if you went for Incite or opted further down Fury, but it keeps a lot of the same ideas. This build however puts an even bigger question on Fast vs Slow Main Hand: Increased Deep Wounds, Mortal Strike, Rend, OP, and possibly Whirlwind or increased chance for Sudden Death and better Heroic Strikes?
I'm going to try out a Deep Arms build with Slow/Fast and see how it goes. Thanks for the feedback.
I don't think that a mace spec offhand would work since it only applies to attacks made with the mace. (I'm pretty sure.) I'm curious to see if sword spec offhand would work though, maybe with one of the Ulduar fast offhand swords.
I just got back form testing Deep Arms Dual Wield + Incite, and I must say, I'm impressed!
I opted for the straight Battle Stance priority rotation of Execute > Rend > Overpower > Mortal Strike instead of dancing to Whirlwind on cooldown since I found that it lowered my dps considerably, although, that may be due to my inexperience with that particular rotation. Using this Rotation, I managed to TIE my personal Titan's Grip numbers. I couldn't believe it. On my best attempt, I managed 2.6k dps, although I think I got lucky with the SD procs. Even so, I averaged about 2.5k every attempt, and never dropped below 2.4k, matching my TG parameters almost exactly. Remember, this is all with The Key Main Handed (unenchanted to boot) and Last Laugh offhanded.
These numbers are still about 100 dps below regular 2h Arms for me personally, but I STRONGLY feel that this type of spec, just as my original build, will scale much better with our stats and raid buffs. Especially haste, which 2h Arms makes little use of, DW Arms makes great use of it.
Imagine for a moment that Yogg-Saron/Algalon drop a slow 1h dps axe, coupled with Remorse or Rune-Etched Nightblade and the appropriate specs, we could be looking at a rival for our top 2 dps specs.
I have been playing fury throughout whole wotlk so im not quite sure. But with overpower being on a 1 sec cooldown, isn't it better NOT choosing expertise? Im not sure how important gcd is for arms, being fury and not able to press anything sometimes.
In previous posts and spreadsheets i see plenty of unnecessary expertise gear, hell I'd even prefer taking hit over expertise even after hit capped. Assuming you got no expertise at all, are hit capped and attack for 2500 non crit with 50% crit chance;
You lose:
Rage buildup = 28(non crit) 54(crit) or an avarage of 41 rage + 5rage overpower cost
A normal hit = 2500non crit - 5000crit damage or 3750 avarage dmg
Possible deep wound damage = 2400 over 6 sec or an avarage of 1200dmg with 50% crit chance
A gcd
Hit + deep wounds = 4950 damage
You gain:
A certain 120% weapon damage crit = 6000dmg
An additional certain deep wound damage = 2880 over 6 sec
OP + deep wounds = 8880 damage
To break it down:
You spend 46 rage and a gcd(1.5sec) on an additional 3930 damage
Now, theres a 6.5% chance of this happening and the weapon speed is (assuming) 3.4sec, by NOT choosing expertise would increase your overall dps by 50. This is excluding the other stat you would normally trade expertise for i.e crit/arp/haste.
These numbers are still about 100 dps below regular 2h Arms for me personally, but I STRONGLY feel that this type of spec, just as my original build, will scale much better with our stats and raid buffs. Especially haste, which 2h Arms makes little use of, DW Arms makes great use of it.
Imagine for a moment that Yogg-Saron/Algalon drop a slow 1h dps axe, coupled with Remorse or Rune-Etched Nightblade and the appropriate specs, we could be looking at a rival for our top 2 dps specs.
You have to factor in the loss from not having a weapon% spec.
Flurry uptime is higher with 2x2h than 1x2h and I'm not sure you can keep it up so much.
Yes the builds I tested for DW were both with incite or with imp exe and anyway the problem is that these builds scales much slower than 2h ones.
The real bonus of HS is on very fast weapons, and with such fast speeds (like LL so in the range of 1.6s base) the bonus of HS per second is compensated with slow weapons with more bleed damage.
However the more AP you get the less the bonus of HS is significant.
What I don't see is a very wide range of good 1handers especially axe dps weapons. Mace can be the best options since they are shared with rogues and shamans and there could be good ones. Also since your bleed damage will be lower the ArP contribution will be very high.
The problem is that summing everything up you'd need a lot of AP to make DW viable since you must compensate the loss of Slam, MS, OP, Rend and DWounds damage as well as a static 6.4% bonus.
Incite is very expensive since it's 8 points (first 5 useless) to get a boost of 15% on crit on a single spell that is around 26-30% of your damage. Not bad but very tied to outer sources of damage.
To "dance" well you need macros and nearly perfect timings for HS-swaps so you don't cripple too many due to rage loss. Especially with a fast MH you can dance quite often just after a special+HS with no rage loss.
It would be intresting to test such builds in a raid environment with proper gear, but in the end I am thinking they aren't planning for us to go DW 1h anymore. 2h damage will be higher and unless you find a situational advantage of using 2x1h it won't be viable.
I have been playing fury throughout whole wotlk so im not quite sure. But with overpower being on a 1 sec cooldown, isn't it better NOT choosing expertise? Im not sure how important gcd is for arms, being fury and not able to press anything sometimes.
In previous posts and spreadsheets i see plenty of unnecessary expertise gear, hell I'd even prefer taking hit over expertise even after hit capped. Assuming you got no expertise at all, are hit capped and attack for 2500 non crit with 50% crit chance;
You lose:
Rage buildup = 28(non crit) 54(crit) or an avarage of 41 rage + 5rage overpower cost
A normal hit = 2500non crit - 5000crit damage or 3750 avarage dmg
Possible deep wound damage = 2400 over 6 sec or an avarage of 1200dmg with 50% crit chance
A gcd
Hit + deep wounds = 4950 damage
You gain:
A certain 120% weapon damage crit = 6000dmg
An additional certain deep wound damage = 2880 over 6 sec
OP + deep wounds = 8880 damage
To break it down:
You spend 46 rage and a gcd(1.5sec) on an additional 3930 damage
Now, theres a 6.5% chance of this happening and the weapon speed is (assuming) 3.4sec, by NOT choosing expertise would increase your overall dps by 50. This is excluding the other stat you would normally trade expertise for i.e crit/arp/haste.
Am i out of my mind and missing something here?
Due to OP mechanics Expertise could have a lower SEP but if you plan to have dodges to get more OP procs you always have to remember you trade that missed damage with OP and more often 1 gcd, that is a big impact on your rotation.
Since your special:white ration is 1.3+:1 you have 30% more chances to miss a MS, Rend or slam than a white hit.
An average white hit at 50% crit is
25%*75%WD(glancing) + 50%*(208%WD + 48%WD) + 18.5%WD = 18.75% + 128%WD + 18.5%WD = 165.25%WD
An average OP is
WD*120%*(228% + 48%) = 331.2%WD
An average special is
WD*46.75%+(228%WD + 48%WD)*46.75% = 297.19%WD + various multipliers and static damage
In case you miss a white hit you gain most benefits: OP damage - white dodged - the spell you replace with OP
a) 331.2 - 165.25 - (331.2-297.19) = 131.94%WD
While when you miss a special: OP damage - missed special - replaced spell difference
b)331.2 - 297.19 - (331.2-297.19) = 0 - multipliers/static values of the spell
In this second case you should also think that you lose DPS since you are doing an OP in 2.5s.
So instead of dealing 297.19/1.5 = 198.13%WDPS
you deal 100/2.5 = 40%WDPS
in 2.5s it is 290%WD loss.
As a sidenote You get slightly more damage since DW is not mitigated, but I'm pretty sure only a spreadsheet can tell us this
Conventional wisdom on this thread seems to be TfB -> MS -> SD procs -> Slam with everything else on cooldown -> HS if next swing could put you over the rage cap.
Since I get bored at work and the forums I read have started moving slowly in anticipation of Ulduar, I've started reading (for better or worse) the official warrior forum. The theorycrafting there, from people who generally don't come off as stupid, seems to suggest TfB -> SD procs -> Slam -> MS as a rage dump/while moving. Their reasoning is that Slam is much more damage/rage efficient (especially with the 2pT7 bonus), and that MS does terrible damage for a 30 rage ability.
This seems to fly in the face of everything I read here, so I'm curious what the math guys here thought.
Conventional wisdom on this thread seems to be TfB -> MS -> SD procs -> Slam with everything else on cooldown -> HS if next swing could put you over the rage cap.
Since I get bored at work and the forums I read have started moving slowly in anticipation of Ulduar, I've started reading (for better or worse) the official warrior forum. The theorycrafting there, from people who generally don't come off as stupid, seems to suggest TfB -> SD procs -> Slam -> MS as a rage dump/while moving. Their reasoning is that Slam is much more damage/rage efficient (especially with the 2pT7 bonus), and that MS does terrible damage for a 30 rage ability.
This seems to fly in the face of everything I read here, so I'm curious what the math guys here thought.
Edit - These are obviously 3.0 priorities. Both places seem to suggest that TfB procs will fall down the list with the new TfB.
I don't want to slap you:P but I really don't see anything supporting those speculations if not "this hits for ~200 more but costs 15 more rage".
It isnt clear if they are talking about sustained dps or burst, the OP just seems to rely on procs and going for highest DPR instead of DPS.
The examples from the OP are really uninformed, there's no mention to how damage stacks, nor any example gives an idea of scaling (except one poster mentioning Slam scales better).
I personally don't believe that kind of theories unsupported by evidence or math, especially when they are almost all based on wrong assumptions (MS damage applied only to bonus damage, no mention of opportunity costs etc).
At the end they are trying to fix a rotation and a priority list, and surely the OP didnt read a single post of this thread if he jumps to those conclusions like slam>MS cause of DPR
Back in TBC, after they super-nerfed all the bosses, I broke off from Arena a bit to try some of the raids I had never been to. I used a Deep Arms DW build back then when Sudden Death was, correct me if my memory fails me, 30% chance on crit. Using all S2 PvP gear and Breeching Comet Main Hand, Quickening Sword of the Prince Offhand, with matching specs, I was able to outdps every member of the second best raiding guild on my old server. Execute was far and away my number 1 damaging ability, followed by Deep Wounds and Melee. They were good players with good gear too. Granted that Deep Wounds was still double dipping, and SD was still OP, the concept was there.
If at 50% crit rate (which I'm sure I was only at or below 40% then) the old Sudden Death becomes a 15% chance on hit, that's not a far cry away from our 9% today plus the new raid haste effects and extra haste on gear.
I'm not sure what you're saying with Flurry. Flurry only gets consumed by auto-attacks/heroic strikes, and is rather easy to keep up while Dual Wielding with 40% crit not counting heroic strikes' extra 15% crit chance. The same is true for TG, which keeps it up nearly 100% of the time. Arms doesn't get Flurry, so I'm a little confused by your statement. All I can gather is that it doesn't really matter, as specs with Flurry should be able to keep it up 100% of the time without too much trouble.
All that aside, you make good points regarding Dual Wield Scaling versus 2hamd scaling. Yes it does scale slower with AP since instant attacks are a smaller part of Dual Wield damage, which I admit, I underestimated, and that's a pretty big hurdle to overcome since AP is our strongest stat (well, aside from ArP, but I think the jury is still giving its final verdict on that). The 6.4% two-hand modifier further hurts this fact.
You do have to take into consideration that DW scales the same if not BETTER with the remainder of the stats. Armour Pen should favour DW a bit since less of its damage comes from Deep Wounds/Rend, but the gain is negligible in light of the 6.4% static bonus. Crit and Expertise should be similar for both specs. Hit is better for Dual Wield since it increases your chance to auto-attack hit past yellow cap which in turn increases the frequency of SD procs, not a huge improvement, probably negligible due to 6.4% static bonus again, and its not really something to gear for, just a ncie bonus if you happen to be over cap.
Haste is the big player here IMHO. It scales so well for Dual Wield on many levels. First off, let me remind you that static Dual Wield auto-attack dps is higher than 2h dps, so haste starts a little ahead. Secondly, it increases the rate at which we can apply Heroic Strike, which 2h Arms doesn't do very often. Third, it increases the rate at which we proc Sudden Death, which increases MUCH faster than it does for 2h, since it is applying to 2 weapons which can each proc it. And fourth, Arms gets a decrease due to having it clash with Slam mechanics. Again, not a big decrease, but enough to make a difference.
I would love to model the effects of haste in a raid environment for this spec, but the severe lack of weapons for this spec is hurting it greatly. As a side note, does haste still provide better return per point the more of it you have, like our current ArP? Because that alone could convince me to try an ArP cap/haste hybrid and see what happens, although, the lack of AP in that build would counteract a bit on Execute Damage, but I think the ArP would offset this.
The real problem I can see is the weapons available to us. We'll never be able to test something like this (not just this spec, but anything similar to it) until we have us some suitable painmakers. I guess we'll just have to wait and see if the undiscovered parts of Ulduar will be kind to us.
Thanks again for your feedback, I feel like I'm learning here and even if we're not making progress, we're at least finding out why since as it stands now, it LOOKS as if Dual Wield Arms could compete, but if it can't, it's nice to know what holds it back.
Well another reason DW arms worked in TBC was because Sudden Death wasn't capped at 30 rage like it is now. All this talk about dual wielding got me thinking, there is one talent that scales with 1h weapons in the prot tree. Of course there is a lot of crap talents to get there but who knows..
Ah yes, I forgot about SD not capping at 30 rage, but you also have to consider that it also didn't return 10 rage. While the tradeoff hardly seems worth it, you'd be suprised at how many times you execute below 30 rage, especially if you're dumping rage into a suped-up Heroic Strike, which was not the case back then either, but I digress.
One-Handed Weapon Specialization is indeed very tempting, but you have to give up so much for it. Either you give up Deep Wounds by going into Fury, or you miss the good stuff in Arms but not being able to reach them, and besides, Wrecking Crew functions as 1h spec for us. Shame you can't get 'em both.
Now, if devastate was more viable, we could be getting somewhere, but they neutered that dog when it was still a puppy, so I don't think it'll work. Besides, if Blizzard doesn't want us Dual Wield dps'ing anymore, (which I don't think they entirely do, but I see how they're pushing us away) imagine what they'd do to a prot dps build.
Small side note: For Dual Wield Arms glyphs, you said you used Execution, Heroic Strike and Bladestorm? I used Rend over Bladestorm when I got my results, I felt like it freed up a couple more cooldowns for potential SD procs. I think Rend might be better than Bladestorm in raids, but again, this is just theorycrafting and probably not a big difference anyways.
You guys are pretty funny talking about DW arms. This isn't 3.0 anymore. It doesn't work with the on-hit limitation of SD! The only benefit you gain is a little more SD and a little more melee damage at the cost of so much more.
Your OP hits for a lot less
Your MS hits for a lot less
Your rend hits for a lot less
Dual weapon spec isn't good (sword spec is generally terrible, and offhand sword only hits off-hand damage. Mace in offhand is okay and applies the penetration to all attacks, but mace spec is only equivalent to ~180 ArP rating)
You replace slam with heroic strike (not that great of a swap)
If you get Incite then you miss out on Imp Exe and/or deep arms talents
Haste isn't good (it's best with fury and it's only about .7 STR)
There's too much opportunity cost..........
TG isn't great on dummies compared to other builds, stop comparing TG dummy DPS to other things. Any arms variation is great "poor man" spec due to the strong base abilities it offers compared to Fury.
You have reassured me that Dual Wield Arms may be viable.
All the cons you listed were ones I was already aware of, so I don't have to take anything new into my considerations.
I know that TG dps on dummies is not an indication of its raid dps, nor was I claiming it was, all I was saying was that DW Arms competes with it on dummies, as a baseline. This is why I'm so interested on all this scaling talk. I'm pretty convinced that TG will still scale much better than any dps spec a warrior can have, but if something comes close, its worth a look at. I'm also convinced that for this very fact, TG will be the top dps spec in 3.1, maybe not right off the bat, but almost certainly nearing Ulduar BiS gear.
The instant attacks in the build are not where its damage comes from (save Execute), they're there out of needing something to do (some dps is better than none) and that they can proc SD. They're damage isnt THAT bad either, Overpower can crit for well over 3k with The Key. This build deals most of its damage, and gains most of its benefits from straight melee, which is a key advantage it has over 2h Arms.
Dual Weapon specs is something I haven't explored yet. I was under the impression that mace offhand only applied to attacks made with the mace.
"Your attacks with maces ignore up to 15% of your opponent's armor."
If it applies to all damage, it could be a viable option. Sword spec would be almost purely for additional SD procs, which is also isn't a huge gain, but it's something to consider.
I think the key to success here is haste, which is where you and I differ in opinion. If fury has a .7 STR value on haste, I'm confident that DW arms has a higher value. Yes, you lose benefit from speeding up 2 x 1h rather than 2 x 2h, but you get increased benefit from Heroic Strike and SD procs. Also keep in mind that Fury has the highest STR valuation too.
In summary, When I'm decked out in Ulduar BiS, I'm certain that TG will beat everything else out, as it should IMO. But if I wander in there with my gear now, pick up a couple pieces and stumble upon one of the new 1h axes, I would not be suprised if at that point DW Arms rivals the other specs.
The melee/HS damage of a Fury spec with 1h would probably outperform DW arms. Flurry>BF, IBS>>dual weapon specs, precision>weapon mastery, BT>MS, WW>OP, etc etc. Execute and incite don't stand up to fury's benefits. How are you so confident that haste is better for DW arms than Fury? That's a pretty general claim, and you base the entire spec around SD. SD isn't that good. What makes you think it's so good that is scales so well with all stats especially haste? It scales pretty terribly in my opinion. Haste only benefits SD via autoattacks. SD's proc rate is halfly based on your instants. Without slam, you aren't even filling all of your GCD's anymore.
DW arms SD spec would only be good if SD worked off of crit rate in combination with incite and HS spam. You could basically sit in arms stance and spam hs, overpower, and execute nonstop and fill most GCD's with those if that was the case. Can I stress more that execute scales poorly?
The spec isn't based on SD, its a combination of SD and melee/HS, which both scale with haste, which is why I think haste will scale well.
In all honestly, I have no idea how the spec will perform in a raid environment or with better gear. But as it stands now, it does a lot more damage than I would have expected, and it seems to me that it has potential for growth in a raid and with better gear (especially weapons).
I'm completely open to suggestion, in fact, I'm gonna see if I can cook something up with that 1h Fury build you were talking about.
04/07/09, 1:25 PM
) is that compared to STR or ArP has a very different behaviour in scaling especially on <5min fights. 
Similar Threads
