Like it was said, Overpowers are way more frequent too. In my parses Overpower sometimes goes up to 18% of my dps - usually a bit lower. Its definitely your highest damage skill - and if you couple it with the 1 sec gcd, its really working nicely. You can say that 10% buff is basically "net gain", although obviously you do lose some slam in the process. Add 10% armor penetration from battle stance you get "for free", and arms do look more promising. Another thing really adding to arms is ArP buff. While of course it adds to fury as well (and other physical classes), arms was always in dire need of "more stats to stack". After you got your hit/expertise, all you could stack was STR and Crit basically. Now ArP gets up there, and helps arms scaling.

I haven't really paid enough attention to recount I guess to determine whether or not Overpower actually lights up more (very easy to check) compared to pre 3.1 Arms. On live it's definitely more RNG based, but it seems like it's actually up essentially the same. You'll get multiple back to back procs, where as in 3.1 you cannot get more than one every six seconds or in other words 10 PPM. The main thing this does is allows you to set the priority of Overpower back quite a bit and do one or two additional damage options before having to burn the TfB buff.

Given the high amount of arp from gear and buff, I'm starting to see a point where you can't stack ArP.
Let's assume FF and sunder stack multiplicatively and then stack multiplicatively with the sum of other buffs.
After FF and Sunder the mob is down to 76%, then you can have .1+.15 from battle+mace and there is a lot of good arp gear.

A mace specced arms should be able to ignore 43% armor just with talents and base debuffs.
A non-mace specced arms would be able to ignore 31.6% armor with talents and debuffs.

An arms hitting a debuffed boss would benefit of 76% of his equipped arp. Less buffs, more %, since the nature of the formula is multiplicative, while mace spec only lowers the gear cap being additive.

From Blizz pov this should be a great achievement since the higher your arp, the less important the debuffs, and sunder is much less mandatory.

From a player pov I'm not sure yet if it will still be better "stacking" arp to cap since the relative damage gain for every increase is progressively lower.

If you are hitting a mob and you are doing 60% damage, getting 1% more damage done increasesyour total dps by 1.7%.
If you are up at 80% damage done the same 1% damage increase would be a 1.25%.

Armor doesn't scale linearly so for an arbitrary value of armor you get more or less reduciton depending on the initial armor value.

10% ArP of 10k value = 1k armor ignored
1k armor will be x damage at 10k and x*(1+y%) at lower values
1% damage from ignored armor will never be less than 1% relative increase untill 100% armor penetrated.

Before stating anything we need to know armor classes for every boss, no bugs on arp (actually on ptr is 81%) and also to know what are all the gear options so we can play around possible numbers.
This is also needed to clearly understand how this changes will affect mace and how much of a buff is the battlestance effect.

30% chance on tick vs 100% chance every 2 ticks? Its pretty obvious that on average the number of procs will be 1.66 times higher. Back-to-back procs if anything LOWER the actual proc rate on live, becuase they often cause wasted procs (or losing SD etc). In general ptr gives you guaranteed 10 PPM like you said (save for your mistakes with keeping rend up etc). On live assuming you dont have to waste procs its 30% of 20 ticks = 6PPM, and then you add the occasional wasted procs. Therefore we can assume around double overpower value.

I agree however - it would be good if they buffed other arms abilities - which are a bit unexciting and lackluster in pve.

Bladestorm could use a decisive buff - maybe they could switch the WW glyph with imp WW in fury tree, to not really change much for fury, but give 20% buff on Bladestorm (or buff the bladestorm glyph considerably). Current Bladestorm math isnt too exciting.

7xWW= 7x Normalized weapon damage = 7x Weapon damage+7xAP/14*3.3 =7*WD+1.65*AP

or

2xMS (if you already time bladestorm to not lose overpower procs, you usually lose 2xMS cooldowns basically),1 Execute (on average), 1 Slam

(2*121%*WD+2*121%*380+2*121%*AP/14*3.3) + (2236+AP*0.2)+(110%*WD+110%*250 +110%*AP/14*3.4)=
3.52WD + 3155 + 1.03*AP -

Which for BoH main hand and 5700 AP Gives us a comparison of

14655 (Bladestorm) vs 11666 (Rotation).

That gives Bladestorm a whooping 33dps increase around on boss fights (27 if you use glyph of execution).

Mortal Strike is just unexciting in my opinion. And Execute could honestly be uncapped - I understand the change was made mostly for pvp reasons, but it really doesnt seem like warriors are so dominant atm . Mostly though, bladestorm feels often like a nuisance - you can easily lose dps by badly timing it, and well timed is just a minute effect.

Ive already got a set of gear that will give 59% armor penetration with raid debuffs at all times using a axe/polearm. I just need grim toll to finish the set and make it work. I didn't calculate how much i would get out of grim toll but the proc must give me close to 100% armor penetration if it doesn't. This would skyrocket dps versus any other trinket.

Quick trinket-question.
Anyone done any math on Bandits insignia vs mirror of truth? Got Bandits today, and it seems to proc a whole lot, plus the ap-boost is pretty sick. Don`t quite know if it`ll outweigh mirrors yet, tho. Anyone with experience here?

On a happy-happy-joyjoy sidenote I got cryptfiends bite tonight as well. Finally something with more than 830 topdmg. Looking forward to what I hope will be a bit more dps.

If you're talking about on a non debuffed Heroic dummy with just Battle Shout up, I sit between approximately 3450 - 3500 over six minute spans consistently. And I don't stress about trying to make sure a Slam comes directly after an auto, it's not worth worrying about. My breakdown ends up being like this for the most part:

Deep Wounds 19%
Melee 18%
Overpower 16%
Mortal Strike 14%
Execute 12%
Slam 10%
Rend 7%
Bladestorm 4%

There is no set rotation for me, it's all priority, plus I'm most likely still going to screw up but not enough that it should greatly affect DPS. This is what I do, although maybe others are having a better success with something else:

1. Keep Rend up obviously, trying to reapply it right before it falls off, but often you can't due to the previous being affected by Wrecking Crew. If it's about to fall off and you have a choice between putting it back up or hitting Sudden Death or even Execute (if it's the difference between waiting for a GCD or not) because you can't overwrite your current, do Rend after one of the other abilities.
2. Bladestorm when it's up.
3. Sudden Death when it lights up, prioritize this over both Mortal Strike and Overpower except possibly if you only have a second left before the TfB cooldown would be up.
4. Mortal Strike
5. Overpower; you have to get used to being able to mentally guage how much time you have left on this without looking at it after it lights up so that you can do Sudden Death or even wait a second for Mortal Strike to come off cooldown before using this or even Slam. There is no reason to hit it if it just lit up and Mortal Strike is about to come off cooldown *unless* you won't have enough rage to Mortal Strike anyway. Plus, this can also lead to Mortal Strike > Sudden Death > Overpower > Sudden Death.
6. Slam in between cooldowns when you have ~50 rage and Heroic Strike when you spike ~70, especially after Bladestorm.

I must admit i really hate Arms playstyle , anyway this addon might help : Curse.com

Good old Class timers , at leats you can see when to reaply rend etc.

"Slam after an auto" is worthless after patch 3, you just waste a lot of gcd time and Totem of wrath really does nothing (you probably meant WF). During BL you can replace some slam with HS, less damage but at very fast speeds you would cripple a bit too much your white damage.

Refreshing rend before it runs out is not necessary at all. To optimize t4b procs you need to refresh on a 3s basis, ie on last tick or 3s after. Refreshing 1 or 2 s before or after is just a waste of t4b proc time.

I generally prioritize SD over MS only if i don't have rage for next gcd. At 40 rage if you have both up it's much better burning MS first so you can use both. In theory at 60+ rage you should prioritize MS from a dps perspective. Anyway it often falls under the "reaction" time of a player using one or the other.

If you have a customizable buff bar (I use ELK Buffbar but there are many, quartz aswell) you can track rend, sd and t4b time left, that helps you much in taking decisions.

Slam every time you have a free gcd and nothing else up

Isn't Sudden Death both more DPS and DPR even at 30 (20) rage? Priorotizing Mortal Strike, even if both were equal still shoves potential DPS lower due to lowering your chance for more Sudden Death procs or just overwriting your current before you use it. That also lowers any potential extra Deep Wounds and Wrecking Crew from happening.

Yes it is more damage per gcd, but Shha had a point on this. I still find myself prioritizing exe often but it's due to reactions.
Even if SD is more dpr and more damage per gcd what is more important is the flow of rage. Theorically if you are sure you will have rage to MS after exe (supposing both up and we are at 40+ rage) you can prioritize either one or the other without breaking next gcd and with marginal damage difference.

At exactly 40 rage with both SD and MS up if you use EXE first you are left with 10 rage, so more delay on MS (at least till swing) and more waste of gcd time.
Using MS first will grant you an EXE in the next GCD and 10 rage left after it, and in a raid environment you swill surely swing before 2nd gcd ends.
At 30 rage MS is a complete dump of rage so it could be better using EXE to keep 10 rage.

I don't see the waste of potential deep wounds and wc due to prioritizing MS over exe even considering potential loss of sd procs. They both have same chance to crit and same gcd, the only differences are damage and rage usage (that is the same at 40+ rage).

It's just the same gear I have on right now sans the off hand. I don't really have an optimized Arms setup yet, although I do have a few unenchanted pieces of gear in the bank and I also have a Grim Toll (which means I am going to try to get equivalent replacement items for slots that have extra hit). I'll try prioritizing Mortal Strike and see what happens, but I've seen so many Sudden Death procs happening after a Mortal Strike that was delayed 1.5 seconds due to hitting Sudden Death that I'm not convinced, but the DPS will be proof enough.

After some discussion in PM with Graul, i tried to reevaluate my math and found an important mistake.

You should not compare the lost execute vs lost MS damage, but rather Lost Execute-Slam vs lost MS -Slam.

As such the difference between Execute vs MS is much higher (since instead of lets say comparing 8000 dmg execute to 7000 MS crits, we compare 8000-6000(Slam)=2000 vs 1000 of MS).

Math that ill save for now here until i can make it cleaner suggests that:

- On dummy indeed it might be better to prioritize Execute (save for rage problems)
- With raid buffs MS is better - both due to lower comparative Slam damage (which mostly results from higher haste and more white damage lost by slamming), and slightly faster MS scaling.
- The resulting difference is pretty minute either way
- The gear improvement will tend to shift it VERY slightly towards MS
- The set bonuses will alter the decision much more. Its pretty obvious that extra 10% crit on MS will favor MS more, but its less straightforward that losing the 10% damage on Slam bonus will ALSO push the balance strongly towards MS. Yes altering Slam damage will change the result of Execute vs MS dilemma.

Overall right now its almost balanced and unimportant what you will prioritize, but as gear progresses we can expect MS to pull ahead.

It seems like that's what I have been missing. I had a few lucky runs with timing white hits and slams so I kept it in the back of my head. My gear is not perfectly attuned for arms, but sitting in the lower 3ks is good for me at the moment.

RNG is not my friend. 40% glancing and 26% crit (42% geared crit on PTR). This is one of the many times this has happened and causing my opinion on rotations to sway a bit. Rage gain in a raid shouldn't be a problem.

Thanks for the extra tips.

- - - - -

One thing I've tried a few times is the stance/ww trick when nothing is available but slam. It still takes you a bit more time that it would to slam, but in the right moment you won't sacrifice a white hit. It's a big rage killer and very hard to find the moment, but I've had it lead to higher dps when it did work out for me.

The last time I played arms was back in 2.4 for Sunwell, so I'm not sure if they've changed slam yet or not;

Does slam still reset the swing timer? Or should I be disregarding my swing timer and guiding off GCD's?

It doesn't reset the swing timer anymore. IIRC it now pauses your swing timer.

Edit:Confirmed. "Slam now suspends the weapon swing timer rather than resetting it. "
Taken from: WoW -> Patch Notes -> Current Patch Notes

Right now ArP doesn't add like this. The huge question mark on ArP is what formula is going to go live - multiplicative, additive, or some sort of strange combination. If Arp became strictly additive, your numbers would be correct, but there's just no indication that that's going to happen AFAIK.

[Edit: 3.1 ArP info at http://elitistjerks.com/f31/t29453-c...evel_80_a/p16/ -- Read Ralik's post. So far I haven't seen anyone refute his formula and it matches with other observed data]

According to me playing with numbers in spreadsheets only, even Rallik's bugged PTRdata-fitting formula yields ArP the best for Arms and Fury. If people are curious what gear to get or what spec to go, the answer would probably be if it has ArP on it and its a decent item, might want to look into picking it up cause there's a chance fitting more items in with ArP will help your output in either spec.

Current results seems to say that you can cap armor penetration only getting 100% arp from gear and only YOUR buffs will lower this cap. So with 10% from battlestance and 15% from mace you can collect up to 75% arpen from gear to reach 100%.

A full debuffed boss should have(1-0.2)*(1-0.05) = 0.76 so 76% armor left.
if you are mace specced you have 25% and if everything was additive you would be left with 51% arp from gear.
However equipping 51% arp is not giving you 100% total but only 0.76*scalingFactor*0.51 where the scaling factor is a proportion of the boss armor with half armor (formula is a bit more complicated, you can check it in the link posted above).

Executioner becomes better the more arp you gain from gear. It will surely stack additively since it gives rating (while once it was a debuff so should have stacked multiplicatively with sunder and ff).
Anyway ap works on 100% of our damage while arp only on mititgated one (totalDamage-(DW+Rend)) so there will be a point where executioner becomes better, but we need good SEP values to compare them.

1 STR is = 1.04*1.1*2*1.1 = 2.5168
1 AP = 0.3973
Berserk is 400AP = 400*0.3973 = 158.92 SEP
Supposing both have the same uptime you need a SEP value of 1.3243 for arp to make it viable.
It is very high, but since it increases both with higher str and arp it is possible that in ulduar Executioner could be worth.

Nevermind

Out of curiosity, do you think Arms Warriors should be gemming for armor penetration? It sounds like it might be a mix of Fractured and Bold Rubies depending on how much ArP your gear comes with.

Nah, full ArP until you get full penetration which is explained in the combat ratings at 80 thread. Gemming a balance of on and the other is kind of pointless when one is better and gets better the more you have.