MS/SD versus Splam

[Rocki]

So, tonight I didn't have much to do, so decided to hit up the training dummies with a few rotations. Here's what I found. For the record, if armory shows another gearset, I run with some of the best gear BT and MH have to offer, minus chest, with the BoH. Glyphs: Execute and Rending. All of these tests were done with a flask up (+180 AP, mixology), Battle Shout and a +20 str foodbuff. They were done over a period of 3 minutes (Battle Shout Duration), during which my trinket was popped once at the start.

First off I started with what is at the moment considered the ideal rotation for a SD build (Talent Calculator - World of Warcraft). Keep up Rend, OP when it procs, Execute when it procs and under 50 rage, MS if neither are up or over 50 rage, slam if MS on cooldown. I was wearing 4/8 T6, so I had reduced Execute rage cost and the 5% MS damage bonus. I found that:

a) this rotation needs a lot of on the spot decisions and good observation, leaving less room for reaction to environment (fire etc).
b) it depends on a LOT of proc luck to reach 1500 DPS selfbuffed.

Result: 23% melee and Deep Wounds, 18% slam, 12% Execute and OP, 7% MS, 5% Rend.
Interested in the high spot my Slam damage took, I decided to toss up the rotation, and use nothing but Slam, OP when it procs, and keep up Rend. The result was this:

a) the rotation is terribly easy to keep up. SCT or so alerts you to OP, and keeping up rend is easy with any decent debuff timer. Other than those 2, mash your Slam button.
b) I reached 1600-1650 DPS fairly easily, while depending on much less procs (just OP procs from rend).
c) since we are not using Execute (with it's 125% threat modifier) we would actually be a little more comfortable on the threat side of the fight. This put the threat cap around 3% higher (25% [extra threat on execute] of 12% [execute damage]).

The result of this bout was Slam topping the chart at about 42% of the total damage. Melee and Deep Wounds a shared 2nd with around 20%, OP 13%, rend 5%.

I tested both "rotations" multiple times in rounds of 3 minutes as described above, and found the results to be fairly consistent. Personally, I tend to favor a different talent build now (keep in mind this is oriented towards PvE) that skips SD and Imp MS in favor of some more utility. Something like this: Talent Calculator - World of Warcraft. Note that while I would skip Imp MS, I still would take the base skill, since it has uses as both an instant attack for on the move/trash/mild PvP, and as a healing debuff in the absense of Fury warriors or Rogues.

So yeah, I'm fairly interested in how everyone else thinks about this. Feedback and opinions highly appreciated.

PS: I may be imagining this, but it seemed like with 2/2 Unrelenting Assault, OP only triggered a 1 second GCD instead of a 1.5 one. Not 100% sure though. Anyone got more info on this?

[Oda]

Just curious, did you try ignoring Rend entirely, moving into Zerker Stance, and only using Slam? I'd like to see how that stacks up.

There's also the likelihood that the SD/Execute/MS build would benefit far more from raid buffs, given that it will consume more rage than the Splam build.

I would agree that OP triggers a 1 second cooldown on other skills (though apparently a 1.5 second cooldown on itself, which is usually a non-issue).

[Gorgrim Warcry]

I was easily bumping 1600 dps self buffed on level 70 dummys. In the best bt/hyjal gear 2 piece teir 6 with alot of arp. You either have your build wrong, or your gems wrong. It takes practice to really get it down took me over a period of 2 days to consistantly accomplish it, and I never left battle stance.

[Rocki]

Just curious, did you try ignoring Rend entirely, moving into Zerker Stance, and only using Slam? I'd like to see how that stacks up.

No I have not. I specced TG after those tests, but need to respec for tonight's raid anyway (TG atm doesn't compete with arms). I'll give it a spin.

There's also the likelihood that the SD/Execute/MS build would benefit far more from raid buffs, given that it will consume more rage than the Splam build.

Splamming was already getting my rage dry occasionally (that's what BR is for), but the raid buffs I'd need to check more in-depth. I think all I'd need is a ret pally with kings, an enh shammy and a feral druid. Applying my own sunders quickly should be doable.

I was easily bumping 1600 dps self buffed on level 70 dummys. In the best bt/hyjal gear 2 piece teir 6 with alot of arp. You either have your build wrong, or your gems wrong. It takes practice to really get it down took me over a period of 2 days to consistantly accomplish it, and I never left battle stance.

My build is the first one I posted in my first post, which iirc is the standard SD/MS build. My gems are correct. If you need detailed gear: T6 helm, shoulders, gloves and legs; chest from MgT hero (don't laugh), choker of endless nightmare, dory's embrace, bracers of eradication, RBoB, dreadboots, unstoppable aggressor's ring, garona's signet ring (need hit), DST, Zul'Jin trinket, Blade of harbingers. Belt and Bracers have a purple gem, everything else is str or crit. 2300 AP unbuffed, 38.14 crit, 140 hit, about 9.7% armor ignore.

I've had about 30 hours worth of raiding as Arms since the patch. I was doing very competitive damage on Kalec and absolutely destroying everyone else in SSC/TK the other day, so I think my rotation is OK. Either way, I'll be giving it some more tests, most likely in the weekend when I'm not limited in time by raids, and let you guys see some results.

[ebs2002]

To determine if it's better to use MS on cooldown, or to splam, see my "When to Execute?" thread (which should probably be renamed now that it's been expanded).

It does calculations for Damage-per-Rage of Execute, Slam, and MS (based on your stats). If your Slam DPR is higher than your MS DPR, then you should ignore MS.

And the target number is the rage at which you should execute (ie, over 50 and you should dump, under 50 and you should execute).

Pop your numbers into there, and try following that rotation and see how it does (noting the potential human error, if you could)