Sorry I should have been clearer, I meant that I have serious rage issues in a raid setting, massive AOE fights like sapphiron and malygos excluded, same with loathed and thaddius. Part of it is obviously glancing blows but it seems that my white crit rate is always below where I should be (it's usualy around 30% on recount and WWS). This despite having almost 45% crit raid buffed.
On a different note, I recently picked up FotFF and...jesus, its amazing. But I got to thinking if it really was worth the 84 crit I dropped from Loatheb's Shadow. In terms of burst potential, a few well timed procs, beserking, Mirror of Truth, then activate Loatheb's shadow and pop bladestorm, it would in my opinion still be better for my DPS on both add fights and Short bursty fights. An example might be gluth, I already usually pull around 4.3k just by bladestorming off the bat, which almost perfectly syncs the cooldown with the adds, meaning that I just pick off the adds as they approach. That is after all where the strength of 54 points in Arms is, the ability to AoE more effectively in bursts. I haven't seen too much theorycraft along the lines of a Deathwish Arms build sicne there are a lot of talents in Arms that you would be missing to pick up an extra 4 in fury. At least 2 from imp Overpower, and maybe 1 from AM, I'm just going out on a limb though since I truly don't believe it would be any more effective than the current cookie cutter build.
Well...crit is RNG in a way, it's never a constant, so you can get terrible luck with it even if you have a high crit rate. Bosses also negate 4.8% of your crit, so that doesn't help.
The thing about mirror of truth that's it's a 10% chance to proc on crit. It might be better for a poleaxe spec, but compared to dual wielding classes and stuff, your uptime of the proc will most likely always be lower in comparison. It's not a very good trinket for me, since my uptime is generally only about 16%, with a 65% or so crit rate (including OP) on bosses. Loatheb's shadow is actually a better trinket AP wise, for me at least. I could actually also time Loatheb's shadow during heroism for added effect. FotFF is amazing though, it's definitely one of the best trinkets for any DPS that can use it, regardless of the crit you lose imo.
As for DW, you have to lose 2/5 wrecking crew to get it too. DW is about a 3.2% damage increase over it's duration, where as 2 points in wrecking crew is 4%. Of course, DW becomes better as the fight is shorter, say, a 2 min fight would make death wish a 5% increase. There's also factors like using DW during hero and trinket procs that will increase it's damage potential of course. I really don't think the damage you gain in the end is going to be very huge though, and losing bladestorm hurts on a lot of fights with adds or movement involved.
Here is my post from tonight's Naxx 25-Man WWS Loading...
As Negiva and that other person posted, My results are also preety good. My patchwerk kill would have been a lot higher, maybe even 5k DPS but the pull was a bit sketchy with our Enhancement Shammy not putting up his WF totem until someone gave him Battle Shout thinking we didnt have Imp. Might in the raid though we did and in addition Heroism was stupidly popped at 35 or 40% from some shaman which really didnt help too much either. All in all as you are gonna be able to see from this report I almost maintained 4k dps on every fight other than Anub Rhekan and overall I believe arms can be competitive for top dps spots easyly in some fights while not in others.
Lastly compared to Negiva's test with Mace which Are preety nice I shall admit, I went the standard path with Axe Spec and having almost 39% crit unbuffed. I do not gem for crit but AP and its a preety nice balance between my stats I believe. I could use some more t7 pieces for the set bonus and some more expertise together with a BoH and then I believe I could easyly brake 5k dps on Patchwerk baring not pulling hate
Romus : Try to produce a WWS or something so we can take a look. It sounds pretty weird that you`re seeing that low a crit on your white-hits, but I wont know anything for sure until I see a WWS.
About Deathwish-arms, I actually specced it right now just to check. Ended up doing aprox 1900 dps on the dummy. Death wish and wrecking crew does NOT stack, sadly.
While I was at it I specced 43 / 28 as well (o hai tbc 33 / 28) and I was really impressed with the damage-output.
Cost me little effort to keep up just about 2150 dps on the dummy. What I obviously did was bypass bladestorm and wrecking-crew + sudden death. It simplifies the cycle a great chunk, by far, and the dps was actually pretty damn close to a full arms-sd build but with way less mindfucking.
I would sadly assume that the lost aoe-dps would make the spec worse than a "full" arms-spec, but I noticed 3/5 flurry doing wonders for my rage-generation.
But there it is, pretty much. Still seems like 54 / 17 is the way to go. Feel free to screw around with it tho, but my 100 wasted gold will hopefully stop you doing it. Hehe.
Just checked out arms sword as well now, with Armageddon as mainhand. In 219 seconds fighting, I had 2 slam and 3 white-misses and still managed 2033 dps. I obviously gimped both my hit and expertise by going sword. It was mostly just my curiousity for swordspec vs axespec. I also have 110 ap on my deaths bite, and 85 on my armageddon.
I even felt kinda ragestarved at times with this spec, but the damage seem to be right up there, maybe better than axespec if I get my hit and expertise fixed. I was also sitting at 32% crit, but slam / whites critted for about 25% ... weird. Any thoughts on this? Im getting more and more confused...give me wingbone and Ill respec even more. Hehe.
Yet another update. Redid the sword-test with 40 hit-food just for the hell of it.
3907 ap, 236 hit (not capped) 32.82% crit and 22 expertise (5.5%, obviously not capped) but I did 2400 dps. I have NO idea how. Really ... I was not rage-starved at all. I suppose I had swordprocs on nice places to help me through, but I could keep abusing slam, overpower and rend. I did 3 executes on a 6 min cycle. Im pretty much as confused as I possibly CAN be at the moment. Chalk it down to silly luck, or actually ponder about the possibility that both swordspec and mace are better than axe at the moment? - And another one. Aprox 2200 dps with both some misses and dodges on my slam, melee, overpower and whirlwind.
That would really surprise me if swords ended up being that effective and having another somewhat RNG component to our DPS would be...fitting I guess. Though thats where racials would come in to play, as you said Ham, you're losing 5 expertise I believe switching to a sword, where as I would gain 3. So if we could come up with at least a decent comparison of all 3 specs, then factor in racials, the problems of AP>crit for mace, crit>AP for axe would be much simpler.
Ham: I wish I had any left over, we haven't raided in something like 4 weeks due to holidays so most of our WWSes have expired. It makes me sad.
[quote=Healranktwo;1039147] FotFF is amazing though, it's definitely one of the best trinkets for any DPS that can use it, regardless of the crit you lose imo. [quote]
This is the conclusion I reached as well, but depending on how Ulduar shapes up, if there are lots of Malygos and/or Grobbulus type fights where it falls off all the time, then its effectiveness for Arms would dimish greatly for the same reasons Mirror isn't AS amazing for arms as it is for say rogues or Enhance shamans.
Started experimenting with how my rage feels with and without 4 piece and I have to say that its quite a help, especially when you consider the high rage most of MS.
[quote=Romoose;1039310][quote=Healranktwo;1039147] FotFF is amazing though, it's definitely one of the best trinkets for any DPS that can use it, regardless of the crit you lose imo.
This is the conclusion I reached as well, but depending on how Ulduar shapes up, if there are lots of Malygos and/or Grobbulus type fights where it falls off all the time, then its effectiveness for Arms would dimish greatly for the same reasons Mirror isn't AS amazing for arms as it is for say rogues or Enhance shamans.
My stack from FotFF never falls off on Malygos except in the phase 1 > 2 transition. Use your heroic throw towards the end of the blender when its about to fall off to keep it rolling. Sometimes I can even attack during blender (slams and all).
I'm Murderizer, and you can compare me to my buddy Fallentrousr our fury warrior. I was using the gimp axe from H HoL till the Sartharion attempts there on the first page. Unfortunately switching from axe to sword lost me my expertise cap.
Its pretty clear that arms is really strong on some fights, and less impressive on others. For example I generally do terrible on Patchwerk...mostly because I get punched in the fase but still, not as competitive. But on fights with AOE dmg like Sapphiron and KT I rock. I think it was last week or maybe the week before that, I did so much dmg on KT, I pulled aggro at 12%, got killed and still held #1 dmg done at the end of the fight.
Yah I've been using H. Throw, something else you can due assuming you don't get put in melee range (It's really random some attempts I get it every vortex) is use thunderclap while you're in the air, you should be able to get 2 off and you basically have unlimited rage while you're in the air.
How were you looking on Loathed compared to say the hunters? It's funny because our better BM hunter will only get something like a 400 DPS boost on it while I get an extra 1500-2000. Just another example of raid buffs and crit stacking.
Just wanted to clear something up that pdpi mentioned earlier.
Execute has a flat bonus of 20% AP.
Slam and MS, for example take 100% AP bonus. The calculation for the AP bonus damage is as follows:
Slam: (AP/14)*(Weapon Speed) = (Bonus DPS from AP)*(Weapon Speed) = Bonus damage from AP
MS: (AP/14)*(3.3) = (Bonus DPS from AP)*(Normalized Speed) = Bonus damage from AP
This is one reason why Raids are more effective for testing than dummies.
The other thing that I am most curious about is the following scenario:
Weapon, Armageddon. Primary focus on Haste, DST-like trinkets. A 53/10/8 Sword-spec. Glyph of Heroic Strike. Most of us remember the BC DST+Cataclysm's Edge Spec. You focused primarily on HS because it is easier to play when you have to move. This is the same idea and there's plenty of gear which would make this possible again. The fact that you can recover massive rage through the HS glyph and you have an innate 15% more crit on HS with this 53/10/8 spec could make this possible.
Aside from that there's not much to say until the next ID. I hope I'll be able to pick up Cryptfiend's Bite and Armageddon. I'll try out The Jawbone this time and I'll focus more on Slam again. Betrayer of Humanity just doesn't convince me anymore.
I'd also like to add my 2 cents on FotFF since it came up a few times now. It certainly is a great trinket. I am currently still using Grim Toll because of the Hit on it. I am trying to get the Darkmoon card that boost strenght, because I personally think its superior to FotFF. I have to agree with most of you that loosing crit over the AP from FotFF isn't a big problem but what about stats like Hit or Expertise?
The Darkmoon card provides a base strenght of 103 or 206 AP and has a proc for 342 strenght or 684 AP. The AP bonus from the proc should average out to 170 AP per minute. A buddy from another guild on my server even mentioned that the proc averages to 225 AP but I cannot confirm that myself since I don't have the trinket myself yet. Also, I'm not willing to pay DKP for FotFF if I can get a great or even superior Trinket through crazy ammounts of farming herbs.
i've seen a lot of discussion about major glyphs (with the majority coming to the consensus that MS/Execute/Rend are the best three), but i haven't seen anything on minor glyphs. which should be used?
note: currently i'm using charge/battle/enduring victory
loatheb is silly to look at since the up-times for the crit buff are all over the place, for instance last time I had a 53% uptime while the mage who topped the meters had 85%.
I was just looking at the last 3 weeks of Thaddius and became quite confused by my numbers. I was confused because I did less DPS each successive week despite gaining a couple upgrades each week. It went, 7700 -> 7300 -> 6800 for the past three weeks. I guess that's a prime example of RNG helping/screwing arms.
3900 3 weeks ago, 4100 2 weeks ago, 4220 last week, 4655 this week on Patchwerk.
8000 3 weeks ago, 8200 2 weeks ago, 8400 1 week ago, 9470 this week on Thaddius.
Reason I type this is to state that in my opinion, ppl blaiming low numbers on rng are false. Each succesive week I ve been getting some upgrades and each succesive week I do much better than the previous one (Gear plus getting better and tighter on how to play the spec). RNG definitely has a role to play on whether you do your best dps or a bit less, but we are talking about a range of probably 100-200 dps difference up or down from your average potential is what you can blame or thank RNG for.
As I posted a few posts above here it is again the link from this weeks Naxx 25 for me so that you can see the numbers yourself WWS Loading...
The above comments make me also want to comment on something a lot of people also say about arms. All that about how it is not too competitive vs Fury and you cannot come near other classes DPS Wise and and and.....In My opinion (and I have been Arms for a loooong time) Arms is just fine and if you maintain well balanced stats and manipulate the rotation properly and stay sharp at all times to get the right decisions about all the SD and ToB procs that pop and use em in conjunction with your other attacks properly Arms Warriors can do VERY good damage, amongst the best currently actually (not considering bugs such as HAT Rogues etc).
Every time is the same thing with classes....people complain about how the class cant do damage (when in fact it can do just fine)...Blizz gets convinced, they buff the class and then it does too much dmg and they nerf it to worse than before. Lets discuss about how you can reach to the level of doing the same damage as me or Negiva or a few other Arms Warriors rather than thinking its some divine exception to the rule of Arms being sucky. If I or Negiva or anyone else can do high dps as arms under COMPLETELY normal conditions and buffs etc, so can everyone else.
Sorry if I go a bit overboard here, I aint trying to be mean to anyone just defending my beloved spec
Howdy,
Is this understanding of how weapon speeds affect abilities accurate?
Execute - Doesn't matter, 0.2*AP + rage and ability bonus Mortal strike - Uses normalized weapon speed of 3.3s for AP bonus, but non-normalized speed for base damage: (Wep. DPS)*3.x + (AP/14)*3.3 + ability bonus, where 3.x is non-hasted weapon speed. Slam - Non-normalized speed, so you do (Wep. DPS + AP/14)*3.x + ability bonus Rend - Non-normalized speed, (Wep. DPS + AP/14)*3.x + ability bonus Deep wounds - '48% of average weapon damage'... guessing non-normalized: 0.48*(Wep. DPS + AP/14)*3.x
Seems like the big components of arms dps like slow weapons. I guess benefits of faster weapon are smoother rage generation and more crits -> more deep wounds (but weaker deep wounds). Without working anything out, I'd guess the benefits to slam and bleeds make slow more desirable. The more tricky question then becomes comparing, say BoH to Cryptfiend's bite.
Id say minor glyphs should include the battleshout-glyph for a longer shout. Other than that its mostly nitpicking and getting what you want and or have some kinky loveaffaire for, really.
About the comments about "not coming near other dps classes" as arms, its just plain wrong. I wreck havoc on thadius, for instance. Just note that on some fights, rng and the general weirdness of arms will make you not win wow, but that should be ok. We still do pretty damn awesome in my opinion and we should concentrate on improving your priorities / cycles instead of qqing about arms not working properly in raiding. This includes me
Did a quick test again today with axe-spec. Im REALLY not feeling the love on the dummy (...heheheh)
Going to regem hit / exp cap tonight and try out swordspec on malygos 3 dragons. Probably not the ideal fight to experiment, but hey ... screw pveprogress compared to testing out arms-stuff with my fellow compadres here.
Also pondering a massacre chant on armageddon, mostly because I keep rolling 12 every time an item I want shows up. The gear on my armory is from early december, I think. Major sigh :\
About the HS-build, Gorrog. Go for it and check it out if you`re in a position to dump 5/5 commanding presence. Im pretty sure I would be yelled at bigtime if I tried to skip that. The build seems fun, tho. I can see a scenario where we would have rage to hs at least 50% of the time + ms slam. I think I have to check it out on a dummy later just for my damn curiosity. Its costing me a fortune, yey
Howdy,
Is this understanding of how weapon speeds affect abilities accurate?
Execute - Doesn't matter, 0.2*AP + rage and ability bonus Mortal strike - Uses normalized weapon speed of 3.3s, so you do (Wep. DPS + AP/14)*3.3 + ability bonus Slam - Non-normalized speed, so you do (Wep. DPS + AP/14)*3.x + ability bonus, where 3.x is non-hasted weapon speed. Rend - Non-normalized speed, (Wep. DPS + AP/14)*3.x + ability bonus Deep wounds - '48% of average weapon damage'... guessing non-normalized: 0.48*(Wep. DPS + AP/14)*3.x
Seems like the big components of arms dps like slow weapons. I guess benefits of faster weapon are smoother rage generation and more crits -> more deep wounds (but weaker deep wounds). Without working anything out, I'd guess the benefits to slam and bleeds make slow more desirable. The more tricky question then becomes comparing, say BoH to Cryptfiend's bite.
You hit the spot just right. I think this is the point where you can push damage the most. Deep Wounds is a prime example because it has two components.
0.48*Weapon DPS + The AP component
You can obviously see that with BoH you will have a greater first component but since the second takes weapon speed in account it will become far more potent at high AP and with a slow weapon. I always though weapon DPS is everything. Now I'm pretty certain I was very wrong.
Howdy,
Is this understanding of how weapon speeds affect abilities accurate?
Execute - Doesn't matter, 0.2*AP + rage and ability bonus Mortal strike - Uses normalized weapon speed of 3.3s, so you do (Wep. DPS + AP/14)*3.3 + ability bonus Slam - Non-normalized speed, so you do (Wep. DPS + AP/14)*3.x + ability bonus, where 3.x is non-hasted weapon speed. Rend - Non-normalized speed, (Wep. DPS + AP/14)*3.x + ability bonus Deep wounds - '48% of average weapon damage'... guessing non-normalized: 0.48*(Wep. DPS + AP/14)*3.x
Seems like the big components of arms dps like slow weapons. I guess benefits of faster weapon are smoother rage generation and more crits -> more deep wounds (but weaker deep wounds). Without working anything out, I'd guess the benefits to slam and bleeds make slow more desirable. The more tricky question then becomes comparing, say BoH to Cryptfiend's bite.
Not quite accurate, for normalized attacks it is wep dps*wep speed + 3.3*ap/14 . The weapon damage range is not normalized, so a slower weapon with identical dps to a faster one will hit harder even on normalized attacks, but the difference is much less than with unnormalized attacks.
What makes weapon speed extremely important is mainly deep wounds, due to it being a damage modifier for all your attacks that can crit. Cryptfiend's Bite is around equal to Betrayer of Humanity(although BoH comes out ahead due to superior stat distribution) and Wraith Spear around equal to any ilvl 213 3.4 speed.
I always though weapon DPS is everything. Now I'm pretty certain I was very wrong.
This is spot on. I was using Saliva Corroded Pike a few weeks ago and was doing 4400 on Patchwerk, then I got Skull Clad Cleaver and my dps dropped down to 3900 the next week with the same group/buffs.
This is spot on. I was using Saliva Corroded Pike a few weeks ago and was doing 4400 on Patchwerk, then I got Skull Clad Cleaver and my dps dropped down to 3900 the next week with the same group/buffs.
In that case wouldn't the [Sword of Justice] and [Wraith Spear] be our best options with the 3.8 and 3.7 speed? I seriously doubt its that simple.
Update: Maybe it has more to do with the max damage of the weapon than the DPS or even the speed?
Last edited by Ariakis_Uther : 01/06/09 at 4:00 PM.
Here's my WWS to give you guys something to compare to
I died on pathwerk and got battle rez'd still got 4800 dps, was at about 5300 before i died
Fights like thaddius(only 8 stacks) and kel'thuzad are hard for me to get high dps because of people dieing and not raiding with 25 people(lawl) and me having to keep sunders up on most of the bosses because we dont like having our warrior be a MT.
Well max. damage of a weapon is not as relevant as the speed and it's true that even on normalized attacks slower weapons hit harder but I don't know how far down you can go to compensate weapon dps for weapon speed. I'm not sure if I want to use a 169 dps weapon because of a 3.8 speed. This is actually something that is open for some testing. I know for certain that I'll try "The Jawbone" tomorrow and see if the 3.6 will make some difference.
Another thing that I see on afenis WWS and armory is that he is also using the darkmonn card and has an amazing 3.6k AP unbuffed. That is the second WWS that I've seen so high and both were using high ammounts of AP. Especially through trinkets.
EDIT: I've just figured out a way to test the speed factor. The Bloodied Arcanite Reaper is the Account bound 2hd Axe which scales with lvl. It has a 3.8 speed, 156 dps, 68 str and it's easy to attain. I bought one and I'll try it out tomorrow on some dummies for testing.
EDIT2: I was just too curious and tried it out. I used Skull Clad Cleaver because BoH has Berserking on it and SCC has only 85 AP in my case and the Bloodied Arcanite Reaper I mentioned earlier in this post. Here's the data on a 5 minute dummy run, I know 5 minutes aren't very accurate but I didn't want to spend the entire night testing stuff.
Skull Clad Cleaver:
2750-2850 DPS range on a lvl 60 dummy, total damage = 762796
Attack / Count / Damage / %
Slam / 92 / 281453 / 36.9%
Melee / 73 / 170047 / 22.3%
MS / 37 / 134525 / 17.6%
Deep Wounds / 218 / 103368 / 13.6%
Rend / 78 / 73403 / 9.6%
---------------------------------------------------
Attack Min. Avg. Max. first row is non-crit
Slam 1752/ 2069/ 2540
4475/ 4916/ 5740 34.8% crit
Melee 1451/ 1633/ 2030
2945/ 3448/ 4130 38.4%
MS 2165/ 2446/ 2884
5164/ 5590/ 6404 37.8%
Deep Wounds 212/ 474/ 817
Rend 705/ 941/ 1033
Bloodied Arcanite Reaper:
2750 DPS range on a lvl 60 dummy, total damage = 747728
Attack / Count / Damage / %
Slam / 91 / 311007 / 41.6%
Melee / 64 / 155559 / 20.8%
MS / 35 / 125977 / 16.8%
Deep Wounds / 198 / 112758 / 15.1%
Rend / 66 / 42427 / 5.7%
---------------------------------------------------
Attack Min. Avg. Max. first is non-crit
Slam 1867/ 2111/ 2451
4559/ 5084/ 5954 44.0% crit
Melee 1455/ 1666/ 1972
3143/ 3548/ 4201 40.6%
MS 2072/ 2274/ 2618
4976/ 5587/ 6674 40.0%
Deep Wounds 216/ 569/ 1146
Rend 705/ 941/ 1033
Well, I know it's a pretty short test being only 5 minutes and not that accurate but it shows what we've been talking about. I've actually ran the Arcanite Reaper test twice because I though the crit was a little high but on the second run it remained high. Anyways as you can see while using a slow weapon your damage shifts to more Slam and Deep Wounds. It is rather pathetic that you can compare a 156 DPS to a 186 DPS weapon. That's 30 dps difference on the weapon and 0.4 difference in speed and yet you are almost able to achieve the same dps. I even wonder how this changes when you are raid buffed with more AP. For the future I am certain I will pick slow weapons since AP scales best with them. The hard part now is to figure out the delicate balance between weapon dps and speed. One thing that I've noticed while doing this test is that your rage generation if far inferior with the Arcanite Reaper but that's an obvious thing when one considers the dps.
It would be great if someone could figure out what's better:
Wraith Spear 186 dps and 3.7 speed or
Cryptfiends Bite 203 dps and 3.6 speed
If you're using a non-normalized attack (which arms relies heavily on: rend, deep wounds, slam), then as tichphys posted above, you're going to see something like:
(Weap. DPS + AP/14)*3.x
as your non-normalized base. If you're comparing a 3.4 to a 3.8 speed weapon, then we can see the difference by doing (just assuming AP is the same for both weapons for no particular reason other than brevity):
where X is the DPS of your 3.8 speed weapon and Y is the DPS of your 3.4 speed weapon. AP obviously refers to total character sheet AP, not AP granted by the weapons. So basically on a non-normalized attack, a slower weapon with a 0.4 speed difference is scaling at about 1 modified DPS* per 35 AP.
(* - Modified DPS refers to the DPS of the weapon after being multiplied by swing speed.)
So consider the bloodied arcanite reaper: its modified DPS is 156*3.8=592.8 and the SCC is 186*3.4=632.4. So when you land a non-normalized hit with your SCC, you should be expecting 39.6 more damage if you didn't have any AP at all. To make the reaper equal the SCC, you need 39.6*35=1386 AP. Every 35AP past that 1386 amount increases the actual damage of your slam by 1 damage. So if you're at 3300AP, that's around 2000AP going towards making your reaper slams better than your SCC slams, and 2000/35 is roughly 57. So you should be expecting around 57 more damage per slam.
Of course, the 35AP:1 modified DPS ratio changes depending on the speed difference. In the case of Negiva vs. Gorrog, you've got a 0.2 speed difference and the AP to modified damage coefficient becomes AP/70 as you might expect. So for Negiva's Jawbone to match your BoH, he needs to make up a 223*3.4 - 203*3.6 = 758 - 730 = 28 modified DPS difference. To match your BoH on non-normalized skills, that's 28*70 = 1960AP, and every 70 AP after that puts him ahead of your BoH. At around 5k AP raid buffed, that puts him 42 damage per swing/slam/etc ahead of your BoH, even with his inferior base DPS. Factoring in his 15% ArP (which affects his whole swing, not just the bonus 42 damage from speed) may just be the icing on the cake, pushing him way above your non-normalized attacks.
So it's starting to look like weapon speed is extremely important for arms, since the bulk of our damage comes from non-normalized attacks (25% slam, 20% melee, 20% deep wounds, 10% rend gives 75% of our damage from non-normalized sources). And to boot, one of our next biggest contributors, execute, is completely weapon damage agnostic so it simply benefits whoever's got more AP. With execute coming in at around 10-15% of our maximum DPS, we've got about 10% of our DPS coming from normalized sources.
Edit:
As tichphys posts below, a melee swing is definitely not a non-normalized source of damage in the same way that rend, deep wounds and slam are. It is, however, agnostic in the sense that the swing timer scales the damage. So it'd probably be more accurate to say that more like 50% of arms damage comes from non-normalized attacks, 35% comes from agnostic attacks (execute and melee) and the rest comes from normalized attacks (overpower (?) and MS).
So it's starting to look like weapon speed is extremely important for arms, since the bulk of our damage comes from non-normalized attacks (25% slam, 20% melee, 20% deep wounds, 10% rend gives 75% of our damage from non-normalized sources). And to boot, one of our next biggest contributors, execute, is completely weapon damage agnostic so it simply benefits whoever's got more AP. With execute coming in at around 10-15% of our maximum DPS, we've got about 10% of our DPS coming from normalizd sources.
I agree that much of arms DPS comes from non-normalized sources, I don't think it is clearly as one-sided as you make it seem. I think white swings should be considered normalized in the sense that independent of weapon speed, white DPS scales at AP/14. Swing DPS looks to be ~20% of damage dealt for some of the WWS parses above, overpower comes in around 5%, and MS at 10%. This is 35% of overall DPS that is ("*normalized"): it scales independently of weapon speed.
*As sulliwan corrected me above, what a "normalized" attack really means is that the damage of the hit scales with AP independently of the weapon speed. However, given two weapons with same DPS and different speeds, a normalized ability will still hit harder with the slower weapon.
MS (Rank 8) for example:
With a 186DPS, 3.4s weapon ([Colossal Skull-Clad Cleaver]): damage1 = 380 + (3.4*186 + 3.3*AP/14)
With a 186DPS, 3.7s weapon ([Wraith Spear]): damage2 = 380 + (3.7*186 + 3.3*AP/14)
Damage2 - Damage1 = 55.8ish more damage with the slower weapon, on this normalized attack.
Since they scale with AP at the same rate, the difference becomes smaller relative to the overall size of the hit as AP increases. On the other hand, since white DPS is tied to swing rate (and not 1.5s fixed GCD), it cares only about weapon DPS, fast/slow makes no difference at all.
The tricky thing is comparing: BoH (223DPS, 3.4) to Cryptfeind's Bite/Jawbone (203DPS, 3.6), or at a tier lower: Black Ice (203DPS, 3.5) to Wraith Spear (186DPS, 3.7). Based on ljl/Barag's post above, there is some threshold AP at which (holding all else equal) the lower DPS weapons WILL do more damage for every use of slam, rend, and deep wounds. But the question is: When do those gains make up for DPS losses on normalized abilities? How about the generally lower stats on the weaker weapons (or in the case of Jawbone, an entirely different weapon specialization mechanic)? The answers here aren't clear to me, and I'm afraid would require some good raid testing.
We think a single application deep wounds is better with a slower weapon. In fact, at high enough AP values, deep wounds can even be better with a slower weapon that has lower innate DPS than a faster one with more. However, a faster weapon will generate more deep wounds from additional swings critting. So in the end, which weapon will give more deep wounds damage done?
Had a lot of coffee today, so figured I'd try to see how this balances out.
Notation:
C = crit rate. Ex) 0.45
WS1 = weapon speed 1 (slow). Ex) 3.6
WS2 = weapon speed 2 (fast). Ex) 3.4
S = average slams per white swing. Ex) 1.0
H = effective haste. Ex) 1.3, applied as in WS1' = WS1/H (raid buffs plus a bit of haste should get this)
Assumptions:
a. Each GCD is used with a crittable ability, and each GCD ability takes 1.5 seconds (in reality, overpower takes 1.0s)
b. Each GCD ability crits at the same rate (C) (in reality some GCDs won't crit: rend, battle shout; others crit more (overpower, recklessness))
c. Each swing crits at the same rate (C) (I think white swing crit is reduced by 4.8% for bosses)
d. Changing weapons doesn't change other stats (same crit rate, AP, and haste), just perhaps different speed and built-in DPS.
Step 1: Figure out how many less deep wounds would be applied using a slower weapon over a faster one.
Calculate crits per second (CPS).
1) CPS from GCD crits: C/1.5. Ex) C=0.45 means you crit on average 0.3 times per second for GCD abilities happening every 1.5 seconds.
2) CPS from swings:
Swing speed: WS/H + S*0.5 (hasted swing speed, plus some number of slam delays)
Crits per swing is C.
Swing crits per second is C/(final swing speed): C/(WS/H + S*0.5)
Overall crits per second:
Combine GCD-driven crits per second and swing-driven crits per second.
CPS = C/1.5 + C/(WS/H + S*0.5) = C*[1/1.5 + 1/(WS/H + S*0.5)] Ex) Using example parameters above,
CPS=0.4421 for 3.6 weapon speed
CPS=0.4491 for 3.4 weapon speed Results are not much different between the two speeds, due to crits per second being driven heavily by GCD abilities and swing critting being suppressed by fixed-time slam delays.
So, when you change to WS1 (slower) from WS2 (faster), how much less crits per second are you doing?
After some simpliication:
CPS1/CPS2 = (1/1.5 + 1/(WS1/H + S*0.5)) / (1/1.5 + 1/(WS2/H + S*0.5)) In our example, CPS1(3.6)/CPS2(3.4) = 0.9845, so the crit (and therefore, deep wounds) rate is that much lower with the slow weapon than with the fast.
Step 2: Calculate damage from one deep wounds application. Your critical strikes cause the opponent to bleed, dealing 48% of your melee weapon's average damage over 6 sec.
From prior posts, we've said deep wounds damage is calculated in a non-normalized manner:
Deep wounds damage (DWD) = 0.48 * [WDPS*WS + AP/14*WS]
The first term in the brackets is base weapon damage, and the second term is non-normalized AP scaling.
Ex1)
Let's have our slow weapon be [Wraith Spear] and our fast be [Colossal Skull-Clad Cleaver], and let's use a ballpark AP of 5000.
Slow-weapon DW damage (DWD1) = 0.48 * (186.5 * 3.7 + 5000/14 * 3.7) = 965.509714
Fast-weapon DW damage (DWD2) = 0.48 * (186.5 * 3.4 + 5000/14 * 3.4) = 887.225143 At same AP and DPS, each deep wounds application with a slow weapon will do more damage than with a faster one.
Ex2)
Let's have our slow weapon be [Cryptfiend's Bite] and our fast be [Betrayer of Humanity], and let's use a ballpark AP of 5000.
Slow-weapon DW damage (DWD1) = 0.48 * (203.6 * 3.6 + 5000/14 * 3.6) = 968.963657
Fast-weapon DW damage (DWD2) = 0.48 * (222.9 * 3.4 + 5000/14 * 3.4) = 946.629943 At the same (high) AP, even with lower build-in DPS, each deep wounds application with a slow weapon can do more damage than with a faster one.
So, when you change to WS1 (slower) from WS2 (faster), how much more deep wounds damage are you doing?
After some simplification,
DWD1/DWD2 = WS1/WS2 * (WDPS1 + AP/14)/(WDPS2 + AP/14)
Ex) In our above example with [Cryptfiend's Bite] and [Betrayer of Humanity], DWD1/DWD2 = 1.0236, so each deep wounds application does slightly more damage with the slower weapon than the faster weapon.
Step 3: Bring it together.
From Step 1, we saw that the deep wounds application rate (deep wounds applied per second) was the same as the crit rate, and we worked out what that was based on some simple parameters and assumptions. I'm going to sort of be hand-wavy and argue that for each weapon case, the net deep wounds damage-done per second is going to be equal to CPS times the DWD per application:
Deep wounds DPS (DWDPS) = CPS * DWD.
So does weapon case1 do more DWDPS than weapon case2?
DWDPS1/DWDPS2 = (CPS1*DWD1)/(CPS2*DWD2) = CPS1/CPS2 * DWD1/DWD2 Ex1)
For the BoH/Crypt's example, CPS1/CPS2 = 0.9845 and DWD1/DWD2 = 1.0236, so
DWDPS1/DWDPS2 = 1.0077, or 0.77% more deep wounds damage, over time, with the slower, weaker weapon.
Ex2)
For the equal-DPS SCC/Wraith example, CPS1/CPS2 = 0.9772 and DWD1/DWD2 = 1.0882, so
DWDPS1/DWDPS2 = 1.0634, or over 6% more deep wounds damage, over time, with the slower weapon of same DPS.
Conclusions:
The increased crit-frequency effect of faster swing sweapons does not overcome slow-weapon gains in deep wounds damage, in the parameter regimes included in the examples above. However, the crit-frequency effect does work against slow-weapons and reduces the magnitude of a slow-weapon's deep wounds advantage.
[I could be making all of this ^ up, any feedback is welcome ]
tichphys, awesome job on the math there but lets see if I can put it into English so that I know I understand (and others I am sure).
Fast weapon Vs Slow weapon deep wound
Due to GCD the uptime of deep wound itself is normalized (6 seconds, so you get 2 attacks and possibly a white hit per DW, due to the GCD) We know that a slower weapons DW will do more damage then a fast weapons, however a fast weapon does not mean you get more DWs then a slow weapon because of the "normalization."
That sound about right?
=============
@ gorroks post on the HS spec
I'm starting to think as well that this will be one of the best scaling specs. I don't believe however that the gear itemization is out yet to put it into practice.
High haste/crit gear with Str sockets would most likely end up yielding some nice numbers. It has already been tested that at 2.8 swing time (with haste) that heroic strike becomes better DPS then Slam.
When I tested it on dummies a few posts ago, I felt very rage starved at time and very steady rage at other times. and since sword proc will proc off of any attack that makes up for the 2 rage you don't gain with the glyph. So with high crit and haste and a healthy AP pool I think it could very well become a viable choice. But I don't think we will see gear that will fully show this until a bit later into wrath
==================
And lastly here is a WWS of sword spec in 25 man naxx (Only the guilds 2nd time in there due to the holidays so I'm not pushing the numbers a lot of you are yet)
If you're using a non-normalized attack (which arms relies heavily on: rend, deep wounds, slam), then as tichphys posted above, you're going to see something like:
(Weap. DPS + AP/14)*3.x
as your non-normalized base. If you're comparing a 3.4 to a 3.8 speed weapon, then we can see the difference by doing (just assuming AP is the same for both weapons for no particular reason other than brevity):
where X is the DPS of your 3.8 speed weapon and Y is the DPS of your 3.4 speed weapon. AP obviously refers to total character sheet AP, not AP granted by the weapons. So basically on a non-normalized attack, a slower weapon with a 0.4 speed difference is scaling at about 1 modified DPS* per 35 AP.
(* - Modified DPS refers to the DPS of the weapon after being multiplied by swing speed.)
So consider the bloodied arcanite reaper: its modified DPS is 156*3.8=592.8 and the SCC is 186*3.4=632.4. So when you land a non-normalized hit with your SSC, you should be expecting 39.6 more damage if you didn't have any AP at all. To make the reaper equal the SCC, you need 39.6*35=1386 AP. Every 35AP past that 1386 amount increases the actual damage of your slam by 1 damage. So if you're at 3300AP, that's around 2000AP going towards making your reaper slams better than your SCC slams, and 2000/35 is roughly 57. So you should be expecting around 57 more damage per slam.
Of course, the 35AP:1 modified DPS ratio changes depending on the speed difference. In the case of Negiva vs. Gorrog, you've got a 0.2 speed difference and the AP to modified damage coefficient becomes AP/70 as you might expect. So for Negiva's Jawbone to match your BoH, he needs to make up a 223*3.4 - 203*3.6 = 758 - 730 = 28 modified DPS difference. To match your BoH on non-normalized skills, that's 28*70 = 1960AP, and every 70 AP after that puts him ahead of your BoH. At around 5k AP raid buffed, that puts him 42 damage per swing/slam/etc ahead of your BoH, even with his inferior base DPS. Factoring in his 15% ArP (which affects his whole swing, not just the bonus 42 damage from speed) may just be the icing on the cake, pushing him way above your non-normalized attacks.
So it's starting to look like weapon speed is extremely important for arms, since the bulk of our damage comes from non-normalized attacks (25% slam, 20% melee, 20% deep wounds, 10% rend gives 75% of our damage from non-normalized sources). And to boot, one of our next biggest contributors, execute, is completely weapon damage agnostic so it simply benefits whoever's got more AP. With execute coming in at around 10-15% of our maximum DPS, we've got about 10% of our DPS coming from normalizd sources.
I have been following this thread from start and with some minor changes made with Slam, SD just before whithe hit and such i hit 1st spot in DPS 7 times on bosses yesterday(naxx 25 farm), i´m now getting closer to the numbers i have seen from Gorrog and Negiva.
1. Currently i notice my rend watching fails somethimes :What appropriate addon do u use to monitor this? i use Dotimers atm but i would only like to see my Rend and nothing else.
2. When using Bladestorm i use it together with [Loatheb's Shadow] and my orc racial even better would it be if i combine it with MoT proc or Berserking enchant. What are good addons to see it proc and time untill it fades??
01/05/09, 6:33 PM
Lets keep em coming
Similar Threads
