On another note. Good finding! ;]
We always automatically discount Expose Armour with a scoff. Are we doing it a disservice? Read on to find out!
a) Assuming the boss is 7700 armour (typical for SSC/BT).
b) Assuming the warrior is using a Mallet of the Tides and 1000 attack power. This gives us an average weapon hit of 165 at a speed of 1.7. 1000AP -> 71.43DPS, or +171 damage per hit. We'll also assume 700 damage per Shield Slam.
Case 1: Expose Armour
Improved Expose Armour reduces by 3075 leaving the boss with 4625 armour, or 30.46% mitigation (69.54% taken)
Zero Sunder Armour means no bonus damage, so each Devastate will hit for 1/2 x (165 + 171) = 168. That will get mitigated by armour to 143 x 0.6954 = 117 per hit
Zero Sunder Armour means 109 innate threat, so total threat from one Devastate will then be 117 + 109 = 226 threat.
700 damage Shield Slams get mitigated to 700 x 0.6954 = 487 damage. Threat from one Shield Slam will then be 487 + 307 = 794 threat.
Revenge's average 460 damage will get mitigated to 460 x 0.6954 = 320 damage. Threat from one Revenge will then be 320 + 200 = 520 threat.
Now we can work out the threat for one cycle (Shield Slam, Revenge, Devastate, Devastate) = 794 + 520 + 226 + 226 = 1766 threat for a standard cycle with Expose Armour up, or 1766/6 = 294 threat per second
Case 2: 5x Sunder Armour
5 Sunder Armour reduces by 2600 leaving the boss with 5100 armour, or 32.57% mitigation (67.43% taken)
5x Sunder Armour means 175 bonus damage, so each Devastate will hit for 1/2 x (165 + 171) + 175 = 343. That will get mitigated by armour to 343 x 0.6743 = 231 per hit
5x Sunder Armour means 176 innate threat, so total threat from one Devastate will then be 231 + 176 = 407 threat.
700 damage Shield Slams get mitigated to 700 x 0.6743 = 472 damage. Threat from one Shield Slam will then be 472 + 307 = 779 threat.
Revenge's average 460 damage will get mitigated to 460 x 0.6743= 310 damage. Threat from one Revenge will then be 310 + 200 = 510 threat.
Now we can work out the threat for one cycle (Shield Slam, Revenge, Devastate, Devastate) = 779 + 510 + 407 + 407 = 2103 threat for a standard cycle with 5x Sunder Armour, or 2103/6 = 350 threat per second.
We see that under the given assumptions, we see that using Expose Armour over Sunder Armour results in a net decrease of 57 threat per second.
Plugging in different typical tanking weapons while keeping the same 1000AP, 700 Shield Slam, and 7700 boss armour (Sun Eater, King's Defender, Unbreakable Will, The Brutalizer), we find that the net decrease remains at 57 threat per second in all cases (the variance appears in the decimals)
Varying the boss armour to other common values (as per EJ: [RAID] Boss armor values - Elitist Jerks) we get: 6200 armour -> 59TPS, and 8800 armour -> 55TPS. As before, varying the weapon does not alter this value overly much and the variance appears in the decimals.
Note: All numbers above do not include any multipliers! With Defensive Stance and 3/3 Defiance, the 7700 armour value increases to 57 x 1.495 = 85TPS, and similarly the 6200 and 8800 armour values increase to 88TPS and 82TPS respectively.
One thing that isn't modelled here is how autoattack and Heroic Strikes fit into the picture. Certainly with EA up, each of these hits will land a little bit harder and as such generate a little bit more threat. You could call that 2% more as a generalisation (in line with the overall DPS increase of EA), reduce the net TPS loss of your cycles accordingly, and call it 80TPS net loss overall.
So, is Expose Armour a good idea? Given that you need 5 combo points to apply it in a worthwhile fashion, it certainly is viable for a rogue to build those 5 points while the tank applies 5 Sunders. This neatly sidesteps the loss of the 1500 threat (2242 with stance/defiance) from applying 5 Sunders at the beginning of the battle - just have the rogue apply EA when the 5th Sunder appears on the mob. Depending on how close your DPS hounds your tank, it may not be such a bad idea. The 85TPS loss (51000 threat on a 10 minute fight) translates to roughly 2% more DPS overall. This may or may not be an issue for you.
More notably, the 85TPS decrease won't be an issue when your tank peaks their threat output in the 1000+ range. The case to keep an eye on is when your tank's threat output is bottoming out (rage starvation, boss abilities that freeze the tank, etc.). In those cases is where 85TPS is more likely to make a difference.
As always, the best people to ask are your DPS. Only they know how much they really hold back in general to remain under the tank. (If they say they don't need to hold back, well, you probably need new DPS actually =) Seriously, though, if your DPS aren't holding back then perhaps EA is a viable option.
Cut and paste error!
After an 8hr shift, I admit I just skimmed the math-heaviness, but the thought I had: I was of the impression that, after there are 5 sunders on the mob, Devastate isn't giving the bonus threat from re-applying the sunder effect.
If we assume this is true, if you build up 5 sunders, then have the rogue overwrite your 5x Sunder with Expose Armor, wouldn't the threat from Devastate be a hair higher than it would be with Sunder, because the mob has lower armor and you wouldn't get any bonus sunder threat anyway?
Edit: I just forgot about the bonus threat of 35 for each sunder effect already on the mob. So you'd lose a significant amount of threat
Last edited by ebs2002; 02-22-2008 at 03:50 PM. Reason: Realized I'm dumb
Other things to consider are the DPS loss incurred by applying Expose Armor over a different finisher (though this is minimal), as well as other tanks -- though as long as a warrior is required to build threat, that's not much of a difference.
Dovie'andi se tovya sagain - it's time to roll the dice
This is news to me! I never looked into it but was always told by warriors that the reason you took them was because sunder armor was better or so it was implied and thus youd want a warrior tanking a boss.
haha this is indeed an interesting find =)
Very interesting article. Seems to make enough sense. I'll try this out the next time I run with a rogue.
Good work, Satrina!
DW-Prot never looked so sexy...
So you would also have to consider the numbers the rogue chosen to use EA would lose, and compare that to the overall gain in raid dps, after you have figured if your tank can hold threat without sunders up.
It really comes down to the same as blood frenzy, do you have the melee dps available to make up for an individual loss inb dps.
It would be interesting to see the comparable analysis in raid DPS, yep.
One thing to conisder is that often enough DPS does have to hold back to some extent. If you have a rogue that is already holding back, then assigning EA duty to that rogue might not change anything at all for that rogue, while giving lesser geared/skilled DPS a free boost.
The biggest problem is I dont think rogues spec imp EA ever...
And rupture represented 1.6% of a rogues damage on one fight i measured the other night meaning the 2% gained would probably be better overall...
The thing to consider with this really is it makes having a druid or paladin tanking a mob very viable. Infact it increases raid DPS without losing TPS because paladins don't do alot of threat from melee damage (5-10%). For DPS races that could be an option...=)
That was one of my first thoughts - in any case, the use of EA for mobs tanked by druids and paladins would be ideal.
Rupture should be more like 5-10% total dps on boss fights, it is ot used much on trash, which is why it shows so low over a total raid. If it is under 5% of total dps, (combat roguyes) then theya re not using the right rotation.
If you are going to do the math on the benefits of armor reduction debuffs you should also assume that Faerie Fire -610 and Curse of Recklessness -800 are up there, both of which stack with Sunder. This gives additional benefit to stacking more armor penetration. In a post on my guild forums I calculated out the benefits of armor penetration assuming a total raid debuff of -4010, the benefits of the -475 bonus of Improved EA would be about 4% DPS increase against a 6200 boss.
So if you are doing 1000 DPS on the meter for a 6200 boss, your actual damage output is 1207.43 if the boss was completely unarmored. Assuming the same DPS, your meter value on a 7700 boss would be 894.7
Code:Penetration 6200 Boss DPS % increase 7700 Boss DPS % increase 0 baseline 1000.0 0.00% 894.71 0.00% 100 1007.9 0.78% 901.04 0.71% 200 1015.9 1.59% 907.45 1.42% 300 1024.1 2.41% 913.96 2.15% 400 1032.4 3.24% 920.56 2.89% 500 1040.8 4.08% 927.26 3.64% 600 1049.4 4.94% 934.05 4.40% 700 1058.1 5.81% 940.94 5.17% 800 1067.0 6.70% 947.94 5.95% 850(Executioner) 1071.4 7.14% 951.48 6.34% 900 1076.0 7.60% 955.04 6.74% 1000 1085.1 8.51% 962.25 7.55% 1100 1094.4 9.44% 969.57 8.37% 1200 1103.9 10.39% 977.00 9.20% 1300 1113.5 11.36% 984.55 10.04% 1400 1123.4 12.34% 992.21 10.90%
Berg's Tank Gear Rankings
have rogues completely overlooked expose armor as well?
The question is does the improved version override sunder armor?
There are fights where threat is never a problem and id think that even at a 2% increase to raid melee dps its still more beneficial than the ruputre damage and the envenom talent points they miss out on...Ofc it depends how useful it is but there are guilds who do have fun and use a paladin MT and for those this information is pretty golden.
Furthermore I wonder how an offtank prot warrior would keep up 2nd on threat if there are no sunders...
Hate to say it, but no rogue will ever be spec'd into Imp EA and raiding at the same time if they actually want to be pumping dps.
The sole build, Mutilate, which should have Imp EA, will never be used for raiding because quite frankly, it sucks for PvE. A mutilate rogue has all the combo points in the world to spend, you can get 5 in 6 seconds if you have decent gear, the problem is doing actual damage with them.
A combat rogue on the other hand, with 2 pieces of T4, can barely manage 1pt Slice'N'Dice / 5pt rupture or evis. As soon as you deviate from the pattern, your dps goes through the floor and you're not worth your raid spot.
It's a nice idea, but for the dps 'increase', the rogue would personally be losing FAR more than 2%, thus really setting the raid back.
You can ask on EJ forums for them to plug it into a spreadsheet for you and say that while it's a neat idea, it really isn't beneficial until blizz lets EA + Sunder stack.
And yes I have a rogue:
The World of Warcraft Armory