Shadowstepping air bursts on archimonde... mmmmmm tasty. (Assuming you already vanished one : P).

As far as cheatdeath is concerned, I think it would be nice to help avoid dying on wipes. Two vanishes + cheatdeath sounds like a cheap repair bill to me.

I'm not experienced in T6 instances but I could give you a pretty good idea for the T5 ones. In the end it sort of depends whether healers are on the ball, since if you have half health abilities from a lot of bosses can kill you, whereas at full there's relatively few. I'm going to assume you're not fighting at half health the majority of the time and focus on real "one-shot" abilities, a la Aran's AE.

Hydross: Nothing that really has a chance to one-shot you
Lurker: Spout, but you can still have 100% DPS uptime while avoiding it
Morogrim: Nothing that really has a chance to one-shot you
Karathress: Nothing that really has a chance to one-shot you
Leotheras the Blind: Potentially useful for staying in during a Whirlwind to eek out extra damage. Probably could buy you ~3 to 5 seconds of DPS every 2 minutes for 8 to 10 minutes, so a total of ~12 to 25 seconds of extra DPS.
Vashj: The only thing that really kills melee on this fight is the goo in Phase 3. While cheating death may help here, it can't really be used "offensively" because the tank is not going to position her in a way such that melee can't DPS without getting killed.

Al'ar: I could see Cheat Death really working here. Using it against the adds that spawn could buy you an additional 5-7 seconds of DPS time per minute, for about 6 or 7 minutes. That's about 30-49 seconds of extra DPS time.
Void Reaver: Nothing that really has a chance to one-shot you
Solarian: Can't use the ability offensively at all in the fight. It could save you from having to run away if a nearby player gets a late jump on his Mark, but most likely that situation is going to end up killing a lot of people and possibly causing a wipe, so it's not really a huge factor
Kael: You could probably actually melee the first adviser because if he starts his gaze, the first attack (which would usually be a one-shot, or close to it) would end up being harmless. However when you get to Phase 3 you wouldn't be on him anymore anyways, and that is where DPS actually matters, as opposed to P1.

All in all this an interesting take, first on offensive Shadowstep and now Cheat Death. My problem, especially with the latter, is it is far too situational. Of all the fights in the T5 instances, "offensive" CD would only play a significant role in 2 of them, and even then, only in "half" the fight for one (Human Phase for Leo). Offensive ShS seems a lot less situational but I still don't think it can bridge the gap between Hemo/Combat and Deep Sub.

This thread has taken a horrible turn, the discussion here should be around doing maximized dps with a spec, who cares about survivability on a fight or preventing a potential wipe? No fight in the game requires cheat death to prevent a wipe, moving out of AOEs is not hard. Also you claim you can eek out extra damage on fights like leo? Why would you ever put the strain on your healers by sitting through a whirlwind, that is just plain stupid. Being the rogue class leader of a guild that has had Illidan on farm for months, I would gkick any of my rogues that came to the raid shadowstep claiming "the utility makes up for the dps loss" this thread disgusts me, can we please turn it back into the right direction.

I originally shared your opinion and still do to some extent (See this post). The interesting thing was when people noted that Shadowstep and Cheat Death could potentially be great DPS boosters. Let it be clear I'm not advocating speccing Shadowstep for raids, at all, period. Nor am I advocating that people should pick up these talents in order to have an unnecessary "safety buffer". I just think their possible offensive uses are intriguing, and there's potential to design a fight where Shadowstep could out-DPS a standard Combat or Hemo/Combat spec. And no, staying in for an extra 2 seconds on a WW and taking no damage is not a strain on your healers.

You'd still take damage since his whirlwind won't 1 shot you. It's just a matter of how low you get before the next damage done will kill you. So yeah, it's an extra strain, especially if the bleed lasts longer than your damage reduction. Something like that is an awful idea. Shadowstep will save you a few seconds getting back to Leo, so you can put out sub par DPS for a few extra seconds over combat or PvE hemo.

You are also talking as if this one time save by Cheat Death is going to actually save you. Umm...whirlwind, you cheat death, as you try to get out of it...dead rogue. Under this scenario you don't gain 5-7 seconds every minute, you lose all your dps for the rest of the fight. Unless I'm missing the actual mechanics of how Cheat Death works, I only see it as really being a one-time oops I screwed up saver.

Cheat death works by preventing damage that would otherwise kill you, aka you have to be low enough for the hit to trigger cheat death, for whirlwind to trigger this on leo you would already have to be sub 3k~ intentionally hurting yourself is putting strain on your healers regardless of how you look at it.

I must be really tired because I royally screwed up that example. Having reread your post and mine, of course you're correct, CD is not very useful on Leo.

My point still stands that although Subtlety may not match Combat Hemo for DPS the mechanic of using CD offensively is interesting.

Ok lets go over the gains and losses from shadowstep/cheatdeath on a fight by fight basis for BT/HYJAL. First we need to look at what you lose by going 41 into sub.

You lose 5% crit, 2% damage on humanoids (most bosses non humanoids), combo point and energy regen.

Rage Winterchill: shadowstepping back in from a bad death and decay saves 2 seconds of dps, lets say this happens 3 times a fight (Ive had fights where every DND was on casters) thats 6 seconds of extra dps vs 5% static crit and better energy/cp gen. so for this fight its a LOSS of dps.
Anetheron: If you have to run out because of an infernal, saves a max of 10~ seconds of dps on average if you shadowstep back in. LOSS of dps.
Kazrogal: Absolutely nothing that you would use CD or SS for. LOSS of dps.
Azgalor: Running out of rain of fires, and cheat death possibly saving you from a rain of fire death, however.. I never die on this fight due to proper use of cloak timed with silences, so Im only counting the dps gain from shadowstepping in once every 30 seconds, fight lasts around 5-6 minutes so a gain of 30 seconds of dps? still seems like a LOSS of dps.
Archimonde: You should not be dieing in this fight period, you have cloak of shadows if you were to ever even take a doomfire you can cloak it. For this reason I do not count cheat death to be a dps gain, you should never even risk standing in a doomfire to get extra dps. Shadowstepping could save you around 20-30 seconds of dps time on this fight.
DPS LOSS

See the trend? I fail to realize how you guys can actually justify picking up cheat death and shadowstep, it actually boggles my mind. You act as if shadowsteping in from 30 yards once every 30 seconds is some huge dps increase, look at what you are losing 100% of the time. I could continue, however the bosses in BT are even worse to support SS and CD, 90% of the time you stand there and mash sinister strike.

My initial interest in the problem wasn't "omg cheat death owns," it was "is there a way to model additional dps uptime as a result of playing cheat death optimally on certain types of fights?" It turns out that yes, there is a way to model it, and initial results on T5 content are dissappointing against one-shot affects. Extra margins on consistent damage have not yet been discussed, and are probably harder (and riskier) to model. I don't even play a rogue, for me this is all a thought experiment.
And I do agree with you in that my intention is to turn as much of this "utility" into hard DPS numbers. Not being able to run out of WW (eg) is one thing--wanting extra dps from not having to is entirely another.

I don't mean to come across bitter, but I don't like to see the thread de-railed. Shadowstep pve specs remind me of posts I see in the WoW class forums.

What's even more boggling is the fact that people are assuming you'll have your SS up exactly when you need it. You can shadowstep twice per minute, so either you are saving your shadowstep cooldown for when you need to "OMG get back into the fray", which means you lose even more DPS, or you're using shadowstep every cooldown, risking losing the amazing utility some people think it will have. Spec'ing shadowstep for a raid is lose/lose.

CD, as said, is incredibly situational. I can count on 1 finger the amount of times this actually would've saved us from a wipe, and this occurred this reset on Vashj. She killed the last person when she had 0% health, and I died due to a second bad static/root/spit combo at around 7%. She probably had 500 health left at death, so yeah, I basically had to swing my weapons once or twice, and she would've died. That's it that comes to mind.

And that's exactly why it's worthless. Feint removes 1,050 threat, so 1,260 with Sleight of Hand, or -126 TPS assuming perpetual usage. 1800 DPS for a rogue with Salvation is only generating 49.7% of their DPS as threat, so 894.6 TPS. 894.6-126=768.6 TPS vs 779.6 without SoH.

Now, realize that if you don't Feint every time the value continues to decrease. I personally would rather have something like Dirty Tricks, more MoD or more Camo over such an utterly tiny bit of -TPS. If you really never use your rogue for anything else, then fine, but why not just get some extra benefit out of spending so many points in Sub anyway?

Sleight of Hand benefits in PvP also if you want to count it as extra resilience. It mostly comes down to Sleight of Hand vs 2 points in Camouflage (you probably have taken Dirty Tricks at this point already). I would definetly pick first one even thought I cant see myself Feinting with 3,5min Vanish cooldown (Elusiveness).

Edit: This is if you are not taking Opportunity, which I dont really see useful for nondagger hemo build.

I never have problems with threat on Teron Gorefiend and I now typically pull between 2000-2200 DPS without Blood Frenzy. The warrior, shaman and other rogue in my group use to be threat capped until we started twisting tranquil air. Only reason we didn't before is because he was resto at the time and we were still new to the fight, but now as Enhancement he can twist to his hearts content depending on mana.

Regardless cheat death and shadow step should never be considered for improving upon raid DPS. The only possible fight that I could see it helping on is Archimonde to some extent and thats only if you have a massive amounts of air bursts. Or of course if you're finding yourself waiting for that ring of fire to despawn, in which you have deeper problems anyway. Like another mentioned you really can't depend on cheat death for this fight anyways. I myself rarely get a doomfire in any given attempt and would not put myself in a position to get one just because I have cheat death. There are times where this talent could be used, but its your fault for putting yourself in the position for this talent to useful!

Did some initial testing on the 3 builds below. i didn't get to do a wws parse of it, i did use recount to gather the data. Please note this is just raw data that i had time to test today before my BT raid. If it's not useful just let me know and i'll delete it or you can delete it.


1. 20/41/0- my actual current raid spec. Talent Calculator - World of Warcraft

2. 11/28/22 trispec. I substituted 2/2 imp SS and 3/3 imp gouge instead 5/5 Lightning reflexes. I did this due to the nature of the testing. I would be attacking from the front, so i would already be dodging attacks that i would not be dodging from behind the target in a raid environment, and thought additional dodges would skew the numbers further. Talent Calculator - World of Warcraft


3. 11/27/23 trispec. the difference is 4/5 sword spec and 2/2 dirty deeds(this spec) vs 5/5 sword spec and 1/2 dirty deeds(11/28/22). Talent Calculator - World of Warcraft

Testing Conditions: unbuffed, 1788AP, 288hit rating, 29.66%crit, 2/2 weapon expertise. Human rogue with swords recieving additional 1% to crit from racial bonus. Talon of Azshara MH(2.64attack speed) with S2 arena sword OH(1.47attack speed). I have the swiftstrike shoulders which provide the static haste to my attack speeds. Lvl 57 Servant of Razelikh in the blasted lands, attacking from the front.

first spec 20/41/0 combat sword
testing time: 604.37seconds
used: AR rush x2, BFx5, 3snd/5rupture rotation
dps: 1031.07
total: 623,149 damage
melee: 399,293 damage 35.7%crit
sinister strike: 180,268 damage 38.5% crit
rupture: 36,970 damage 200ticks(25 ruptures)
eviscerate: 6,618 damage 25%crit(only used 4 times; once during each AR push, once during a string of combat potency procs and the last time as i finished testing)


second spec 11/28/22 trispec. 5/5 sword spec with 1/2 dirty deeds
testing time: 607.98 seconds
used: BFx5 3hemo/5rupture rotation
dps: 1159.1dps
total: 708,678 damage
melee: 492,983 damage 39.7%crit
hemo: 160,443 damage 38.2%crit
rupture: 53,817 damage 200ticks(25rupture)
eviscerate: 1,435 damage 0crit(only used 1 time, as i finished testing)

third spec 11/27/23 trispec 4/5 sword spec with 2/2 dirty deeds
testing time 600.33 seconds
used: BFx5 3hemo/5rupture rotation
dps: 1120.50 dps
total: 680,595 damage
melee: 452,144 damage 34.5%crit
hemo: 168,844 damage 38%crit
rupture: 58,156 damage 200ticks(25rupture)
eviscerate: 1,451 damage 0 crit (only used 1 time, as i finished testing)


observations of note: on average i used up 3-4 of the 10 hemo charges with my own white damage as i waited for energy for hemo spam.
hemo damage and rupture damage is higher than normal as it is a yellow attack below 35% mob health for the duration of the testing time. A fairly large diffence in crit rate for normal melee attacks in the 11/28/22 build, with an extremely close crit range on the primary special attack between the three builds.

I think this requires more testing to see the effects that sword spec procs, mongoose procs, WSC and TT procs affect this.

i will attempt to get a few wws reports of this so we can figure in the affects of my CoH(chance on hit) proc items. If i can get the data i would need someone to put it to formula to calculate the diffence. These were all 10minute fights. Do you want me to lengthen the sample time or change out/remove my CoH gear?

'ello. I appoligize if this has been discussed already but with a hemo build is haste still as a desirable stat as it is with combat. Haste= more ptency procs. But without having potency is it sill worth having it? Also with a hemo build im asuming no one will have imp/vile poison.

To be honest I haven't died in a fight we weren't about to wipe to since we stopped putting melees on Azgalor.

Rage - 2 iceblocks before trinket CD if you bother to wear a trinket
Anetheron - swarm + infernal but you can vanish an infernal on yourself or move away from a infernal on a melee
Azgalor - oh god yes if you use melees on him full time
Archimonde - between cloak and 2 min trinket you can save yourself everytime

Najentus - not a chance
Supremus - only if I wanted to watch TV midfight
Akama - nope
Teron - only when ghosted
Gurtogg - unless you get 2x fel rage and evasion was down
RoS - turning off auto attack while evasion tanking p1 enrages helped a lot
Mother Shaz - not since the nerf
Council - a couple of close calls but nothing cloak of shadows wouldn't save
Illidan - cheat death won't save you from bad parasites

Guys, can you please check something... those that have tested and confirmed hemo build actually goes on top of the combat one, can you verify with the dps spreadsheet 2.3.0.3(fixed for hemo)? I for once are loosing fair bit of dps going for hemo according to that spreadsheet so no idea if it can be trusted for hemo or it is too early

Does the 125% buff to weapon dmg apply to base weapon dmg or weapon dmg after AP modifiers?

I ran a test some days ago, I've posted it on the rogue forums already, so here it is:
" So I tried
Talent Calculator - World of Warcraft on the PTR.

I compared the damage to the tests I've done 1 week ago with the first hemo buff. I was attacking a Servant of Allistarj in blasted lands. These guys don't die unti you have the quest for them. The combat build I tried was 20/41(the one I use, armory me to check). They bleed and get poisoned, murder works against them as well. I've "normalized" the yellow damage of the hemo builds since they get an unfair advantage against the combat build.

The advantage in short is:
These guys don't die, they stay in 1% and have only 3k life. So the hemo builds have almost always the dirty deeds additional damage up. So I "normalized" with a simple math, as though every 1% of the boss needed the equal amount of time to be burned. I was completely unbuffed in both cases. I used Instant Poison in main hand and deadly poison in off hand. I opened with Garrote and did SnD at once in both cases(Initiative didn't proc for the hemo build)and I was using 1s/5r cycles in both cases, combat build had 2-3 eviscerates also(due to combat potency I had some spare CP). In the PTR I have worse gear than I have in live

With Combat I had:
AP->1549
Hit->208
Crit->23.84%
Exp->13

And with hemo:
AP->1540
Crit->23.63%
The rest are the same since Vitality is the only difference.

The test was simple. I was DPSing until I deal about 200k damage and then Vanish. It's not a huge sample but it's OK. Both fights lasted almost 4 minutes(2xx seconds).

With the combat build I did 202.316 damage in 225,871 seconds. That's 896 DPS.
With the hemo build(after the yellow damage normalization)I did 193.175 damage in 216,361 seocnds, and that's 893 DPS. Although I can burn only 6 charges(maximum, although I didn't count the chance for extra attack at all) every time I reapply hemo(3.5 seconds). The rest 4 charges are a 41 DPS damage loss, and there's no such loss in a raid, so the effective DPS of the hemo build is 934 DPS.

That's for now. You're welcome to correct me on things I did wrong."

Now I explain later what the "normalization" I did was:
"DD affects both Hemorrhage and Rupture(and all the other finishers, CP generating abilities)but not poisons. What I did to normalize my incorrect damage bonus is:
Find which portion of my damage(of the 200k sample)was yellow. Then I removed some of it cause DD didn't work 100% of the time(the portion I removed wasn't 100% correct but it's off less than 1-2%). Now I had about 6xk yellow damage. I nerfed 65% of it by 20%. So my final yellow damage was 20k+40k*0.8.

Why my math is not 100% correct? Well it's for two reasons. Because of the incorrect yellow damage buff
my yellow damage was 31% of my damage done. So since for 2145 HP DD didn't work(the mob has 3300 HP)I removed 665 from my overall yellow damage, instead of 27.7...% which was my final yellow damage after the normalization. Also another minor mistake is that I didn't count the 665 yellow damage when nerfing the overall yellow damage."

This test was done in build 7521(??)where Hemo got buffed to 125%. At the current moment I didn't know that Hemo got normalized, so I thought that a better Hemo weapon(Since I was using a Vindicator's Brand)would out DPS combat without the charges taken in account. Also since the mob was level 5x I didn't have any misses/dodges and Combat would be hurt more by them cause of combat potency while Hemo would only notice some dodged Ruptures(And Hemos but so is Combat with SS).

One more thing is that I'm not convinced which ability(Hemo Vs SS) scales better. We have SS ((Weapon Damage+98)+(AP/14*2.4)*1.16) Vs (Weapon damage+(AP/14)*2.4)*1.25 :
This means that if 9% of your weapon damage is 98 then Hemo gets better but that's not gonna happen so easily, only raid buffed, and I'm not sure if this will happen(I'm not sure if the 125% modifier is applied after the AP bonus).

Of course even if Hemorrhage damage is sub par to SS damage, the white damage gained through the Hemo debuff makes up for it, although while in a raid the debuff charges are spread between 8-10 people(melees and hunters). My conclusion is that while one's DPS will fall by a slight amount the raid DPS will increase for more than that loss. So 2 or maybe 3 hemo rogues in a raid, based on the amount of the melee and hunters, will increase the overall DPS.

One last thing, on my test I didn't use any cooldowns. So theoretically a combat rogue will have much better burst damage if needed(Boss on low health or DPS to help the raid catch up a timer or something). The only DPS cooldown a hemo rogue has is BF, which is better as Combat due to combat potency.

All in all, and sorry for my stupidly long post, Hemo is now seriously considered a raid viable build(As 11/2x/2x)but it's not superior to Combat. At least that's my conclusion of my hand-made test, I didn't run any spreadsheets yet. I hope I helped.

I think you should spec out of DD. Because of the nature of these mobs it means you are constantly getting the damage increase because they stay at 1% health. Your better off doing the test without DD and calculate what damage DD would add afterwards.