A few questions i have regarding personal gear selections, right now i have 4 valorous 1 heroes (gloves never drop >.<)

-4 set bonus, worth it not worth it? gloves of patch and helm off OS 2 drakes better than having the 4 set?

thats my main concern right now

Enderlin, server Velen if you have any other suggestions, i know how to play ive been playing fury for a long long time but allways looking for correct math data etc. to help me maxamize my damage.

After playing as both arms and fury it really seems the four set bonus is far more valuable for arms where rage is much harder to come by and all GCDs must be utilized. It's nice for fury but I'll drop it in a heart beat for best in slot items once our hunters and rogues already have them. Personally I wouldn't drop the four set bonus for the items you've mentioned but I would for the blue aspect helm, frost adroit handguards, and chestguard of the recluse. At a quick glance at a recent wws it looks like the four set bonus procs about an average of eight times in a three minute fight. Not a tremendous amount of rage saved but slightly helpful none the less, sort of like unbridled wrath, not much of a dps incease.

I'd say no, based on my experiences, since the proc rate seems to be too unreliable.

* Patchwerk: 5 procs over 2 minutes and 50 seconds.
* Gluth: 5 procs over 2 minutes and 15 seconds.
* Thaddius: 8 procs over 5 minutes.
* Loatheb: 10 procs over 3 minutes and 45 seconds.
* Noth: 12 procs over 3 minutes and 15 seconds.
* Razuvious: 21 procs over 2 minutes and 15 seconds.

I'm not sure what happened on Razuvious, other than I may "accidently" have hit the offtanked adds with whirlwind a few times. :) Same would be true for Noth as well. It doesn't seem to be worth it for single target encounters though.

I'll second that, I've experienced a pretty low proc rate too:

Anub'rekhan: 9 procs over 2:41
Faerlina: 12 procs over 2:01
Maexxna: 7 procs over 2:25
Noth: 16 procs over 3:34
Heigen: 10 procs over 3:37
Loatheb: 12 procs over 5:13
Patchwerk: 10 procs over 5:35
Grobbulus: 10 procs over 3:41
Gluth: 10 procs over 4:55
Thaddius: 19 procs over 5:35
Razuvious: 20 procs over 3:24
Gothik: 19 procs over 3:25
Four Horsemen: 4 procs over 2:10 (I died close to the end if I remember correctly)
Sapphiron: 17 procs over 3:53
Kel'Thuzad: no procs, I died thanks to Mr. Ret Pally chaining a Frost Tomb between the two points that melee stacks up and not getting any heals.

First off, the proc numbers were kind of high because this was from a raid from a long time ago, and thus my guild's raid DPS wasn't nearly as good as it is now (we down Patchy in 2:30 easily) so my proc rates would be even lower if I tried it today.

Secondly, you'll notice that fights with adds have way more procs.

The results are just awful in my opinion. The proc rate is decent but unreliable, although if it was something like some strength or crit rating, it would be far more worth it. -5 rage to your next attack is just pitiful compared to the stats you could get instead by using different pieces.

gyus , i have 1 simple question : so can anyone tell me about 2h enchants ? i have 2*berserker. is it right ? or massacre will be better or offhand ?

There is a fabulous spreadsheet which can tell you your best choice. for me it is 2 times berserker
http://elitistjerks.com/f81/t37462-f...n_spreadsheet/

Was looking for more of a challenge last night, so in anticipation of the 3.1 stance rage loss change, I tried working Rend uptime into my DPS rotation. Used a macro that looks like this:

/castsequence [stance:1] reset=2/target/combat Rend, Berserker Stance
/castsequence [stance:2] Rend
/castsequence [stance:3] Battle Stance

On fights where I could maintain decent uptime did between 3-5% of my overall damage. Can go here for the reports. If the numbers seem a bit wonky it's because guild was doing 20-man.

The two major caveats here are available GCD's and rage loss. Right now I found by stance-dancing to get a Rend up I could be losing as much as 100 rage, which obviously impacted my Heroic Strike/Slam overall damage, but as I said in 3.1 this will be mitigated a bit. As for available GCD's, I didn't sweat them. When I had one where BT, WW, or Slam weren't up, and shouts didn't need refreshing, I danced and got Rend up, providing it wasn't already up (temporary buffs that boost Rend's damage make it impossible to reapply until the DoT fades). On some fights this was easy and on some it wasn't so easy. If you happen to be the Sunder-guy, it's another variable to juggle and can get very disorienting, especially if there is a fight mechanic that demands attention. On Thaddius (genius Prot War missed the jump) I let Sunders fall off twice, resulting in an obvious loss in raid DPS, and a loss in personal from having to waste time reapplying.

[edit] whoops, fixed casting berserker stance from berserker stance.

If we take landsouls rotation (from post #269), we can modify it, to include Rend thusly:

0.0 Whirlwind

1.5 Bloodthirst

3.0 Rend

4.5 Bloodsurge Slam / FREE GCD

6.0 Wait

6.5 Bloodthirst

8.0 Whirlwind

9.5 Bloodsurge Slam / FREE GCD

11.0 Wait

11.5 Bloodthirst

13.0 Bloodsurge Slam / FREE GCD

14.5 Bloodsurge Slam / FREE GCD

Rend runs for 15s unglyphed, which means we'll lose 1s of uptime every rotation, unless we want to delay Whirlwind (My first look at a glyphed Rend seems to indicate less uptime-loss but lack of convergence between the Rend CD and the 16s rotation, thanks to having to delay Rend at various times. It merits further investigation though). By putting in a Rend at 3s, we lose possible Bloodsurge procs from any Heroic Strikes we put in between 14.5s and 16s. They will run out before the GCD after Rend is up. Procs from WW at 0s and BT at 1.5s wont be affected.

Mileage may vary of course, but if we assume one HS that procs one Bloodsurge in that timeframe, then we lose the amount of Bloodsurges that proc, but do NOT proc on either WW or BT. Which means that we have a 0.2x0.8x0.8=12.8% chance to lose a Slam, given a HS between 14.5s and 16.s.

The rage loss is a whole different kettle of fish, but in terms of available GCD’s there doesn't appear to be a huge problem fitting in a Rend into a fury rotation. Latency and human reaction times will of course factor into it all.

[edited for spelling errors]

Why not go with a glyphed Rend then? Currently, most Fury Warriors seem to use Glyph of Cleaving because it helps a bit on add fights and trash. Rend may very well be better, assuming that there are no other mechanic or glyph changes that would make a better candidate. As you can see here, you'll have to Rend less frequently, although you will have 1.5 sec of Rend downtime assuming perfect play.

0.0 Whirlwind

1.5 Bloodthirst

3.0 Rend

4.5 Bloodsurge Slam / FREE GCD

6.0 Wait

6.5 Bloodthirst

8.0 Whirlwind

9.5 Bloodsurge Slam / FREE GCD

11.0 Wait

11.5 Bloodthirst

13.0 Bloodsurge Slam / FREE GCD

14.5 Bloodsurge Slam / FREE GCD

16.0 Whirlwind

17.5 Bloodthirst

19.0 Bloodsurge Slam / FREE GCD

20.5 Bloodsurge Slam / FREE GCD

22.0 Wait

22.5 Bloodthirst

24.0 Whirlwind

25.5 Rend

Thats not a proper "rotation", since it doesnt "loop". If you wanted straight loop of what you posted, next WW would be at 32sec into fight, and then BT after. An attempt to make a Glyphed rend table would be as follows:

Find best duration for skills with 18 5 and 8 sec cds. Since we dont want to overlap rend ticks (if we do we might as well not glyph it, making the whole discussion pointless) The duration of rotation would have to be within 18-19 sec or 36-38 etc (if we have more then 1 sec of downtime on rend, again - the 16 sec rotation with 15 sec rend is better).


Next step is to determine whether we prioritize BT or WW - normally its agreed to prioritize WW for a 16 sec rotation, but maybe with the rend glyph and using rend we can switch it around a bit. 16 sec rotation is not compatibile with 18-19 sec intervals to a huge extend (3 min rotation and 8 cycles - meaning rend will end up being used on about every position of GCD - colliding for SURE with WW or BT). We can try the 72 sec rotation with 9 WW and 4 glyphed rends.

Step 1

0.0 Rend
18.0 Rend
36.0 Rend
54.0 Rend
72.0 Rend - LOOP

We can see though that rend being moved 2 sec behind every 2 WW will eventually collide making it useless.

So maybe BT oriented roation - 36sec 2xRend 7xBT 4xWW

0.0 Rend
18.0 Rend
36.0 Rend, Loop

Addint BT

0.0 Rend
1.5 BT
6.5 BT
11.5 BT
16.5 BT
18.0 Rend
21.5 BT
26.5 BT
31.5 BT
36.0 Rend - LOOP (1 sec lost time on BT)

Now to fit in 4xWW

0.0 Rend
1.5 BT
5.0 WW (To allow for WW after next BT)
6.5 BT
11.5BT
13.0 WW
16.5 BT
18.0 Rend
21.5 BT
23.0 WW (2 sec delay)
26.5 BT
31.0 WW
32.5 BT (moved the 1 sec delay on BT from next one here)
36.0 Rend - Loop

overall loss of BT/WW cooldowns and Rend uptime:

1/5 BT every 36 seconds
1/4 WW every 36 seconds
100% rend uptime

vs the 16 sec rotation which yields us
1/5BT every 16 seconds
1/3 Rend tick.

Putting it at same fight lenght - 144 sec

4/5 BT 1WW lost vs 9/5 BT and 3 Rend ticks
1 WW loss vs BT+3 Rend ticks.

The 18 sec rend rotation IS producing more dps since WW wont hit for more then BT + 3 Rend ticks. But frankly using data from Landsoul's spreadsheet the difference will be in range of 2kdmg /144 sec above 75% (when rend deals more dmg), and 1k DM /144 sec below. Is the average of 7.5 DPS gain worth spending a major slot for rend? I beg to differ. Glyph of Execution will prove better in that case for single target with ease (7 executes on a 5 min fight already guarantee you better result).

Whirl Wind and Blood Thirst should never be pushed back to Rend. Both WW and BT are cooldown oriented and it is foolish to ever push back their cooldowns.

Edit: I'm also not sure why you are basing your (flawed) rotation on Rend having an 18 sec duration. Rend is 15 sec base, 21 sec glyphed. Furthermore, you seem to insist that we should wait five full seconds to WW at the beginning of a fight. Whirl Wind is our best damage dealing ability, and should absolutely always be used ahead of any other ability or any other pressing matter.

[Item not found!] wouldnt that quantify to use rend every 21 seconds? How does glyphed rend actually work? I'm not in game to check but doesn't it wait 3 seconds after application to go, ticking 7 times over 21 seconds? If you refresh it early, the 7th tick does not happen?

Also, because of slam, Rend refresh will ultimately end up getting pushed back in the Free GCDs in the 16s rotation, just because the loss of chance of overlapping a slam is > loss of rend tick and sometimes you get 2 Slams in a row. Each time you get 2 slams in a row the rend will get pushed down to the next spot where you don't have a slam. I don't know if this makes a difference or not, but rend will be really low on the priority list after 75% health and you can push it back as long as you want as long as you keep getting Slam procs. This is just my opinion at the moment, don't have anything to back it up yet but it's a hypothesis I'm willing to follow. Maybe Rend Glyph will help in this way minimizing the downtime of rend due to Slam! procs?

0.0	Whirlwind
1.5	Bloodthirst
3.0	Bloodsurge Slam / FREE GCD
4.5	Bloodsurge Slam / FREE GCD
6.0	Wait
6.5	Bloodthirst
8.0	Whirlwind
9.5	Bloodsurge Slam / FREE GCD
11.0	Wait
11.5	Bloodthirst
13.0	Bloodsurge Slam / FREE GCD
14.5	Bloodsurge Slam / FREE GCD

Edit: So yeah the DPGCD of rend is way less than 20% of a slam. So it's safe to say that if you have a Slam! unused you better use it before thinking of using a rend dance in a GCD.

Edit: I also did some additional thinking about rend dancing. When using a macro, there is a defined time that you would sit in battle stance, losing a definite amount of attack power and crit rate on a possible heroic strike or melee attack during that time. It's kind of similar to dancing to use improved overpower on a free GCD if you don't have any expertise. I experimented with that at the end of Classic and beginning of TBC. So really there is a lot of hidden opportunity costs involved with rend (including now slam loss, melee loss, and HSDPS loss) which makes the 16s rotation and the glyphed 21s "use whenever" rend more attractive.

Is the cleave glyph that much better than execute for fights that have adds?

Even on single target boss fights I rarely use more than two Executes. It mostly depends on whether or not BL is still up below 10%. I would think that the Cleave glyph would definitely be a lot more useful for trash (for now) and any boss fight that has more than one add up at a time that you're able to hit.

Moved.

While I am tempted to say "it depends on your gear and play style," it really is a clean sweep on add fights. If you have one, take a look at your WWS parse for, say, a Grobbulus fight. Look at the number of times you used Heroic Strike, and then consider that a talented Cleave will deal just a little less damage (WD+495 for HS versus WD+488.4 for cleave) and it will hit up to two slimes in addition to Grobbulus as long as the slime tank keeps the slimes near the boss the whole time.

Just by glancing at my WWS reports from my last few Grobb kills, Cleave has done ~30% of my overall damage.

Like I prefaced though, it's all relative to your play style. Some Fury Warriors like to just spam Execute when the boss hits 20% but this is generally considered to not be as optimal as simply maintaining your rotation and throwing in an Execute when you have an open GCD and enough rage. This has been covered, and you can find a great discussion on it here:

http://elitistjerks.com/f81/t36784-w...p7/#post991326

Once you get to a higher level of gear, your AP becomes so high that it's always better to WW, BT, and Slam over Execute, in which case I agree with Graul:

This is what you would see if you looked at my WWS reports. During an entire Naxxramas raid, I may execute ten times. Quite frankly, unless there are some changes to Execute or it's talents or glyphs, I see little reason to use it as we move into later content as Execute scales so poorly compared to WW and BT.

Heh, and I just realized that Graul was also the guy I linked. Seems like he knows what he's talking about.

There's been plenty of focus on rage related issues in both the Damage Dealing and Tanking special forums the past week or two, attracting Ghostcrawler's attention in both areas.

Reading between the lines from his most recent response, I'm not so certain the previewed change will make 3.1. While the new system would benefit high rage environments (relevant for Fury DPS), it potentially breaks things in low rage situations (eg. PvP, where you could no longer switch from Battle to 'Zerker at 19 rage or less directly into a Pummel, assuming fully talented TM).

I've also experimented with inserting Rend into spare GCDs and while I also didn't have much issue with 'finding the time' to do it, I found the rage loss ultimately not worth the trouble in most cases. Specifically, if I had the misfortune of no white crits in the first 3-5 or so seconds after the stance dance, I found myself missing my relatively smooth HS rhythm (also couldn't help but wonder if I was missing out on a Bloodsurge proc or two as a result).

To use properly Rend and reduce the rage loss you could use a swing timer.
Unless you are using 2 weapons with the same speed it's very difficult since the swingtimers don't recognize mh and oh hits.
If you could use a swingtimer you would be able to Rend with no loss of rage only when you have 25 rage(3/3 tm) and your swing is gonna hit in more than 1s.
If the changes in 3.1 will go live you will spend 30 rage to do a Rend and this means you will have the same situation as above at 55 rage while you'll be unable to do that swap at 25.

I don't know how many times you are in a situation with 25 rage and a spare gcd and 1s from the swing, but the only other option is forget about the swing and use BT in battle stance if you happen to have a swing while rending and before stance is out of cd (1s indeed).

Rend with 2/2 Imp rend and Trauma/Mangle up does (weapon damage + 380)*1.56.
On a target over 75% it does (weapon damage + 380)*2.106
Anyway it cant crit.

Now the thing you should consider are
1) you have to keep your bt/ww rotation without clipping nor delay (possibly)
2) you have to choose if on a free gcd is more worth using slam or rend

Slam is weapon damage+250 mitigated by armor and can crit.
Rend is higher damage but is a dot and opposite to Slam! procs can be delayed without loosing a proc chance.

At first glance I would put Rend in a free gcd even if it could cost some seconds.
The only reason to use a Rend over a Slam is when the delay you will accumulate is higher than slam damage, and this means you should be sure to use every gcd for like 15 seconds.
Since this could happen really rarely and cant be prevented we should just try to evaluate how many gcds you can loose before a Rend is worth more than a Slam.

Rend = (380+wd)*1.2*1.3(*1.35 if the target is >75% hp)
Slam = (250+wd)*1.1(*1.1 with 2 t7)*armor reduction

When target is over 75% and u have 2 t7
Rend = 592.8 + 2.106wd
Slam = 302.5 + 1.21wd

Rend deals (592.8+2.106wd)/10 damage on gcd. The Glyph adds 2 more ticks so increases the damage but not the dpgcd nor the dps.
Slam deals the whole damage on a gcd.
Rend gcd = 59.28 + 0.2106wd

Slam with 0% arp and sunder/ff with 50% crit the multiplier is 1.115
Slam gcd = (302.5+1.21wd)*(1-0.3075)*1.115 = (209.48 + 0.84wd)*1.115 = 233.57 + 0.93wd

So you need 5 gcd for Rend to go over Slam if the target is >75% (in this case Rend IS better than 20% slam).
If you are sure you'll have to delay Rend for more than 7 seconds then you will not use slam in that gcd regardless if you can use it later or loose it completely.

5 gcd is quite much, I think reasonably you would have something like this situation only if u have to dwish+reck+ww+bt+slam. Other situations are consecutive Slam procs between bt.
It's very similar to arms rng

It is possible to use Rend, but it is not worth replacing other abilities and procs so you should use it on a free gcd possibly in-between swings.

In therms of damage it is surely an increase unless the swap makes you ragestarved when bt/ww are out of cd, and it wouldnt be good to keep it in a fixed rotation.

In terms of GCD usage alone sure it will be an increase in DPS (5 GCDs is quite long) but the idea is still a little shaky when looking at less powerful melee attacks during the 1 second every 15 to 26 seconds one would spend in battle stance using a macro to hit rend.

You'd also have to subtract some factor of damage in all heroic and white attacks over some interval 1sec/avg_rend_period based on losing 3% crit and the 10% ap bonus. Also, you would have to subtract in some form the rage. I'm already taking rage loss into account in my 3.1 spreadsheet, but I'm having to do a lot of work putting together an estimate of the white / heroic damage loss. Things that are taking into account are ratio between offhand and main hand attacks, total time spent in battle stance, etc.

The main issue is not the rage loss (at least not in the calculus itself) but the swings you could do in that 1s period.
If you can calculate the average free gcds in-between a swing you will anyway run around the problem.

What I was thinking is a rough math to understand how many optimal chances to use a rend dance you can have in a given timeframe, that includes:
- having 25 rage or few more (and eventually understand how much rage loss you can afford)
- you have a free gcd with no slam
- your swing will hit after you are able to reswap. I would consider 1 - 1.2 secs since lag is really dumb on stance swaps.

The average count of this free good gcds for rend could be then used to understand the average uptime of rend.
And then you can consider the overall rend damage as its dps*uptime.

this is where I see Rend have more chances to happen

9.5 Bloodsurge Slam / FREE GCD
11.0 Wait
11.5 Bloodthirst
13.0 Bloodsurge Slam / FREE GCD
14.5 Bloodsurge Slam / FREE GCD

since you have 3 free gcds in the space of 5 seconds, but even here odds are that you can fill those with shouts or your hs procs like crazy.

Modelling Rend seems far from being easy if you plan to consider possible loss of dps from specials used in battle stance. It's also very tricky to understand when these good gcds happen for real ingame unless you have an addon that tells you when to swap ~1.5 seconds in advance.

if (nextswing>1.2 && gcd<1 && ww>1.7 && bt>1.7 && noSlam && rage>=10 && rage<=36) alert (Swap for rend)
36 rage is what I supposed could be the limit at which you can dance (i used +12 from hs as a good reason to not swap)
Maybe it just pops you a message or a timer bar.
The real problem will be with different weapon speeds.

That would be totally crazy if some kind of addon-code could do that. Blizzard has mentioned however that they too were sketchy on the TM / stance switch change so I dunno if all this thought about rage would be worth it yet.

Last post from GC seems to say they probably won't change the stance dance mechanics. Anyway an addon can easily take those things in consideration and alert you when it's a perfect time to rend. Odds are that can never happen in a fight.

The speculation on the possible rage changes are that Rend will cost you 30 rage with 3/3 TM but there is no way to make it cost 10 like it is now. So if those changes go live you can probably use Rend whenever you are over 55 rage so to ensure you'll be able to use a WW or 60 to use a BT after the swap to berz stance.

Actually Rend full time with 2/2 Imp Rend is worth around 5% of total damage, that's really good, but doesnt scale at all with crit, haste or ArP and generally not much with expertise/hit too.

We should see what's the gear expectation and what's the ArP changes if there will be any before thinking how much Rend will provide in future content.

I'll try to make that addon anyway. I was thinking about a timer that pops up after an hs/mh hit when you have no procs, spells on cd and low rage. The only issue is being able to clearly understand when next swing will hit while dualwielding.