Eyegore

Seems like it shouldn't be too hard to come up with some way to buff raid damage without unbalancing pvp... How about something along the lines of a deep combat talent like: Your finishing moves have a 20% chance per combo point to grant you a 1% haste buff, lasts 15 sec and stacks up to 15 times. Not something you would ever get stacked up very high in pvp, but in any sustained dps situation that would be huge.

 

demibaka

While both the addition of a group buff and a flat increase in dps seem like viable solutions for securing spots for rogues in a raid, the flat dps increase seems like the option that is most consistent with Blizzard's past statements about the rogue class and their role in raids. A debuff related to poisons/combo points also seems like a workable idea, or perhaps some boost to threat management for other members (increase for the tank or decrease to others perhaps), but once again the flat increase to dps seems like the most "roguish" solution.

It has been noted that a boost to PVE viability like this would have ramifications on PVP balancing issues; one idea that occurred to me is linking the boost in dps to the number of living players currently grouped with the rogue. The easiest numbers to throw at this would be a 24% increase in dps, 1% for each other player in a full raid group. This would scale nicely with five and ten man content, as well as providing a rational buff in arenas. Certainly the 24% number would be subject to balancing, maybe only a half percent would be appropriate. One small issue I can envision is the buff this would give rogues in BG's like AB or AV, where the size of the groups is fairly large but rogues can still essentially play the lone wolf. (although I suppose the buff could simply be deactivated in BG's.)

 

• Maestroquark

If only something like this ever existed.

This thread would be better served, I think, if people focused less on where Rogues stand now and more where they could stand in the expansion. At this point, throwing at random %s is pretty useless, as we don't know what new skills/talents are coming and how they will interact with what we already have. Number tweaking should be the exact last step - either expanding the roles Rogues have now (uh, do damage) or adding new roles should be the early focus.

 

Adrammelech

I think it gets tricky when you make a single class the definitive highest DPS class, because people enjoy competing, and it can take a lot of fun out of that experience when there is literally no hope.

I think they need two things. A wider raid synergy buff/debuff. (hemo would be ok, but it's in the pvp tree) and a party buff. Something like expose weakness would be good, but instead of being in an off raid tree like it is for survival hunters put it in deep combat and make it party only, it'd be perfect for a melee group. Give them a poison that increases bleed damage, or can slow a bosses attack speed beyond that of imp thunder clap, some brittle bones poison that reduces armor but doesn't interact negatively with sunder... Something that benefits either the raid as a whole or larger swath of the membership.

 

Capek

Whatever happens I hope that any buff/debuff a rogue brings to the raid benefits the rogue as much as it benefits the rest of the group. I like the way rogues currently position themselves at/near the top of most WWS parses and don't want to see that change. That's why I think ideas like poisons that buff shadow damage would not be great. A: You lose DP and B: You buff another class but not yourself.

 

• Feist-Mok

Unless of course, they give us a source of shadow damage such as that found in the TBC beta, the old Shadow Strikes talent.

 

z0phi3l

There deleted my opinion/comment since it seems that people think I'm complaining when I'm not

-Kaubel edit-
Here's what you said. And yes, you were complaining.

 

• Vulajin

Let me stop this discussion for one second here to ask that you please don't just post nonspecific whines or complaints about your particular situation in this thread. This rogue viability issue is obviously a touchy topic, but there is a huge difference between constructive discussion (as in why the issues exist) and simply whining ("if they don't fix rogues, I quit" or "rogues are doomed, this sucks").

 

Mirag

I think people worry to mutch about the pvp part if blizzard would decide to boost rogues personal dps aslong as its done the right way.

Lets say a deep combat talent that increase our haste with slice and dice with 5/10/15%.
This would not stop me from respecing to shadowstep whenever i want to do arena.

Glancing blows is another way as mentioned before.

 

Arindelest

I've seen the ideas about just providing a damage increase in deep combat, but consider that...

There's nothing stopping rogues speccing Combat for PvP however; especially so if Deep Combat provides huge damage bonuses (Like 15% increased SnD would be pretty huge for a deep combat talent). Recall 1) Just because Sub is a viable PvP spec now doesn't mean it will be in WotLK (SF Daggers and Mutilate were really powerful pre-BC) and 2) Combat can be a good PvP spec, too: Combat Maces was a staple S2 Arena spec. The sum of this is that any change cannot only be *just* in Deep Combat, it also has to be something (as others have said) like weapon skill or a stacking bonus that rewards time on target that will be potent in PvE only.

 

jonnnney

I know this is a small issue but it seems to be coming up a lot in this thread. If there are abilities that increase our overall contribution to a raid they would either need to be skills learned from a trainer or talents with relatively equal buffs in both our raid viable specs, mutilate and combat, or preferably all three. I have seen many posts asking for raid buffs just in combat who seem to forget, or are unaware of, the fact that deep assassination have usable talents.

I realize that they are just words and they don't necessarily mean that your trying to say combat is or should be the only raiding spec. It was just causing a slight twinge in the back of my head whenever I saw deep in combat only.


Sorry to complain about small things.

 

Mcslice

To summerize, people seem to be throwing around the ideas of:
a) A debuff on the boss
b) A party buff
c) A raid wide poison debuff

Another solution that comes to mind, Blizzard changing the boss mechanics for WotLK.
Would making it viable to bring a second melee group instead of a second ranged group change our situation? It would effectively double the amount of positions allowable in a raid for Rogues.
If they balanced it correctly it could be completely at the raid leaders discretion what to bring for the third group, melee or ranged.
With DK's in the expansion it will bring the total number of melee DPS classes to 5, 6 if you include Feral druid's that don't need to tank for a certain fight, that would equal it with the amount of ranged DPS classes.
If you're guild can sustain it, have 2melee DPS and 2 Ranged DPS groups that are interchangeable depending on the fight (Read: Pick the 3best DPS groups your taking to a fight, based on what the boss needs most).

 

KasumiRevy

While I enjoy the role of being a very strong dps class. I am under the impression the more expansions that blizzard releases the more likely we are to slip in the dps department.

Buffs? No please, I rolled a rogue knowing we didn't have any (coming from everquest rogues never did there either), and would like to keep it that way.

Debuffs? Now this is where I think we should hit if we want to remain an important part of raids. You can't include hemo, because it requires talents. Debuffs for all rogues, it can be simple as more well thought out poisons. We gained 1 poison in BC, and I never use it. Vanish, really is all the aggro drop we need in most situations (do people still have feint on their hotbar? )

I would like to see more traps/locked chests/ and locked doors in raid instances. There really should be no excuse for not having these things required. (you only need 1 rogue to take care of these things). Can you imagine a gauntlet for a raid event, that doesn't see stealth targets, at the end with a locked device?

I am aware, that this may put to much spot light on rogues, but it seems to me, we bring so few things to raids besides DPS, there has to be a counterweight incoming or the class will suffer from this.

I think we should all keep a close eye on the progression of rogues, from here, and give as much input as we can. They still have talents in our trees no one ever specs. (Deadened Nerves, is an excellent example) If we make the same mistake we made in BC, and think the devs will figure it out, we will get new poisons that are lackluster, talents that make you scratch your head, and a slowly degrading position in raids.

I am eagerly awaiting rogue information, to give constructive input, I ask that every rogue fight to take our class back.

 

Timujingeo

Arguably, after hit and expertise, haste is one of our most powerful dps enhancements i.e. DST/haste pots etc. A deep combat talent of 'furious haste' that gives a percentage haste increase based on combo points that has say a 3yd aura range so all local melee gain a benefit. This of course would stack with SnD or it would be pointless.
The difficulty with this is that it would give us some desirability for a raid spot, but we would also be buffing our competitor's dps as well. Maybe stacking with SnD/WF etc. would give us an edge to keep ahead on dps.

 

Rensy

Random tidbit to follow, I hope this is in the right place:

I visit often, and see some things I agree with and things that I don't agree with on these boards in regards to rogues. One thing that has particularly troubled me is the notion that on a stand-still tank and spank style fight a "good rogue" will have only 1 gain of Slice and Dice, or that this is in any way an indication of quality DPS'ing. Also troubling is the notion that Eviscerate is an evil that should much be avoided.

For what I am saying, assume a 5/5 rotation. If one is to pay close attention to mongoose and combat potency procs, and especially if they use Ashtongue, there are plenty of situations in which letting SND drop for a split second (or maybe even 2!) could be beneficial to your DPS-- at least in my opinion. While most of my evidence is anecdotal... if I am not mistaken, having two mongoose procs and an ashtongue proc up (especially if you have a ret paladin in the raid or a feral in your group) would make eviscerate valid. I would like to see math to contradict that. Sure Slice and Dice is the highest DPS finisher for a rogue, but is 1 second of Slice and Dice higher damage than a 5 point Eviscerate, or even an additional rupture at the end of your cycle in all circumstances? I think not.

I don't mean to provide little math and mostly anecdote, but I can't help but bring up what I find to be a fallacy in the commandment of "never let slice n dice fall". I think it is a misleading guideline.

Here, for example, is a pretty strong parse of myself not adhering to this commandment:
Wow Web Stats

Notice some dreaded eviscerates and many many Slice and Dice gains.

 

Left

In that parse, white damage is 70% of your damage. Why would you want to cut your white damage output by 35% [2pT6], even for a second or two? Sometimes it's inevitable to do this, given luck with CP procs or whatnot, but why would you ever do it intentionally?

Consider, if your DPS is 2774 (as it was on that parse, and is amazing btw) and 70% of your damage is white, than losing SnD drops you by 2774*0.7*(0.35/1.35) = 500 DPS. To drop SnD for 2 seconds, you'd have to get back 1000 extra damage from somewhere... what ability is giving you that much extra damage? I guess I just don't see it being Eviscerate.

As for Eviscerate being situationally superior to Rupture, that might be the case. For the purposes of this calculation, I'm going to assume that you have armor penn. and crit procs up such that the boss has effectively 0 armor and you have a 50% crit rate. (IE, this is the best possible case.)

From your armory, and assuming Battle Shout, UR, Dual Mongoose procs, etc., you'd be looking at somwhere around 3300-3500 AP. Thus,

Rupture
1000 + 3500*0.24 = 1840 (w/out Mangle)
(1000 + 3500*0.24)*1.3 = 2392 (w/ Mangle)

Eviscerate
(((985+1105)/2) + 3500*0.15)*1.06*(1 + 0.5) = 2496

So, Eviscerate, under those super-buffed circumstances, will deal a little more damage. But, now we have to take into account energy cost. Rupture is effectively free, while Eviscerate costs 10 net energy. So, the true value of using Rupture should take into account where those 10 energy would go: to Sinister Strike.

Sinister Strike
From your parse, you performed 112 Sinister Strike hits for 186668 damage. Average SS damage is therefore ~1667.

Rupture + 10 energy
Therefore the true value of Rupture + 10 energy is...
1840 + 1667*(10/40) = 2256 (w/out Mangle)
2392 + 1667*(10/40) = 2809 (w/ Mangle)


So, in the best of circumstances, Eviscerate during procs will outdamage Rupture if Mangle is not on the target, but will not outdamage Rupture if Mangle is on the target. Also consider that I'm assuming enough armor penn. procs to take armor to 0. If there is even 10% DR from armor, that small advantage pretty much disappears.


IE, if you want to and have time to watch your procs that closely, you might eke out another 1500 average damage or so over the course of the fight by Eviscerating during a bunch of crit procs. Or, you could just Rupture instead and concentrate your time and energy on keeping up SnD. If you can get SnD uptime to increase by 3 seconds per fight, you will do more damage from that than all this messing around with Eviscerate.

 

• Vulajin

Before I get into any napkin math, let me just make sure we're on the same page: you are talking about running a 5/5 cycle and slipping in extra finishers (like Eviscerate) due to beneficial procs, even at the cost of Slice dropping for a short time, correct?

Let's start by looking at the best case iteration of the 5/5 cycle, where we start with a Ruthlessness proc from the previous cycle, and both finishers proc Ruthlessness. We can also assume that Slice is up from the previous iteration, which was performed "by the book" (i.e. normal 5/5). We'll assume that we're using T6 2pc. We'll assume that our weapon is [Blade of Savagery], and that we have double Mongoose at all times; thus, our swing speed is 1.4 / 1.35 / 1.02 / 1.02 = 1.00, rounded off. We'll also go ahead and assume we're hit-capped; although this is not an accurate assumption, it is an assumption in favor of your idea of squeezing in extra finishers, and therefore makes sense to consider in trying to disprove your assertions.

Since Ruthlessness procs on both finishers, our required combo point generation is 8, requiring 320 energy. Both finishers are obviously free. Our 5-point Slice and Dice gives us 30.45 seconds of Slice, or roughly 304.5 energy to use during the cycle. Now, let's say we want to squeeze an additional 5-point finisher into this cycle. Since we're assuming both finishers proc Ruthlessness, we only need to generate 4 combo points for this additional finisher, requiring 160 energy. Since our extra finisher is going to be Eviscerate, we'll also need to spend 10 energy on the finisher itself, for a total of 170. This will bring our total energy for the cycle up to 490. If we assume we want 100% Slice uptime, then we'll need to generate at least 185.5 energy via Combat Potency (about 12.33, or at least 13 full procs) during this cycle to make it work. Any less will allow Slice to drop for a length of time.

Remember that our actual cycle length is [intended to be] 30.45 seconds, regardless of any other factors, and our swing timer is roughly 1.00. Therefore, for simplicity's sake, we'll say that we have about 30 chances to proc CP during a given cycle. Our expected number of CP procs would be 6, or 90 energy -- nowhere near enough to achieve the desired extra finisher. How lucky would we have to be to get enough energy for that finisher, though? Recall that we need at least 13 procs to guarantee 100% Slice uptime, and each swing has a 20% chance to proc CP; thus, using some statistics, we can find that the probability of experiencing at least 13 procs is 0.31%. So, roughly 3 of every 1000 iterations of a 5/5 cycle allow you enough excess energy to add a 5-point finisher with no detriment -- even if Ruthlessness procs on every finisher.

What if you experienced fewer CP procs, but used another finisher anyway, taking the Slice downtime in exchange for a 5-point Eviscerate? First, let's figure out how much damage a 5-point Eviscerate would deal, given your assumptions of double Lighting Speed procs (240 AP, roughly 6% crit) and an Ashtongue Talisman proc (about 6.57%). We'll assume that your base crit before the procs was roughly 35%, and that your AP is about 3500. We'll also assume full armor penetration and [Relentless Earthstorm Diamond]. Eviscerate's base damage is 1045, and it will gain 525 from AP, for a total of 1570. Your crit rate will be roughly 47.57%, and crits will cause 206% damage; thus, your overall average damage calculates out to about 3109.

Now let's look at various combinations of CP procs during that 5/5/5 cycle iteration and how long Slice would drop:

- 12 CP procs: 0.64% chance, 180 energy generated, 0.55 seconds of Slice downtime
- 11 CP procs: 1.61% chance, 165 energy generated, 2.05 seconds of Slice downtime
- 10 CP procs: 3.55% chance, 150 energy generated, 3.55 seconds of Slice downtime
- 9 CP procs: 6.76% chance, 135 energy generated, 5.05 seconds of Slice downtime
- 8 CP procs: 11.06% chance, 120 energy generated, 6.55 seconds of Slice downtime
- 7 CP procs: 15.38% chance, 105 energy generated, 8.05 seconds of Slice downtime
- 6 CP procs: 17.95% chance, 90 energy generated, 9.55 seconds of Slice downtime

I won't go beyond this as I think we can both imagine that this much Slice downtime is probably crippling. So the question is, how much exactly does that Slice downtime hurt? Well, again, let's assume 3500 AP, 35% base crit rate, and [Blade of Savagery]. Let's assume our main hand is [Vengeful Gladiator's Slicer]. Remember that we assumed we're hit-capped. Let's calculate our white DPS:

MH weapon DPS = 103.1
MH AP DPS = 250
Windfury = (103.1 + 250) x 0.2 = 70.62
Sword spec = (103.1 + 250 + 70.62) x 0.05 = 21.18
Total DPS = 103.1 + 250 + 70.62 + 21.18 = 444.9

OH weapon DPS = 75.3
OH AP DPS = 187.5
Sword spec = (103.1 + 250) x 2.6/1.4 x 0.05 = 32.79
Total DPS = 444.9 + 75.3 + 187.5 + 32.79 = 740.4

Actual DPS = 740.4 x (1 + 0.35 x 2.06 - 0.25 x 0.25) = 1228.0
Slice DPS = 1228.0 x 0.35 = 429.8

Now, you might say "hey, that's pretty good, even if I'm losing, say, 5.05 seconds of Slice, that's ~2170.5 damage compared against Eviscerate's 3109, that's a decent gain." Well, it seems that way, but there are several factors ignored by this estimate of Slice DPS. The first is the fact that losing Slice means your off hand swings slow down, reducing your frequency of CP procs, thus reducing your DPS. The second is the fact that delaying your next cycle iteration (by lengthening the current one) also forces your next Rupture back, resulting in another effective DPS reduction. The third, a minor factor, is that you reduce your frequency of off hand poison procs. Finally, you experience greatly reduced gain from any procs which might happen to occur while Slice is down.

So the conclusion here: even if Ruthlessness procs on all of your finishers, and even if you experience more than 30% CP procs over a 30-second span, and even if you assume that you Eviscerate with several different procs active and ignore the same procs for your white DPS, it is still not likely that you can squeeze in a 5-point Eviscerate and experience a net DPS gain. And if all of these favorable conditions don't apply, then quite frankly there is simply no reason to even consider letting Slice drop.

 

tetracycloide

Let's put some of the rampant paranoia in this thread to rest about current and future rogue viability without party buffs, raid buffs, or a massive buff to personal DPS. Several examples have been posted that indicate that rogue is not the single target DPS king but it does not suffice to take the most extreme outlier of a particular class and use that as an estimate for average damage for every player of that class.

To get an accurate feel of what each classes expected single target DPS would be we need a reasonably large sample size for each class, a boss that favors DPS in time, and some measure of assurance that each sample is attempting to maximize DPS to the same extent as the others.

The obvious choice for this is brutallus being the DPS benchmark of choice in sunwell, practically requiring full use of raid buffs for successful attempts and T6 or equivalent gear in almost every slot of every player. Let’s look at the values across the entire spectrum, and not just the highest single numbers posted by each class, to get a feel for who is doing what in raids.

The WWS score board continently saves thousands of parses to maintain its rankings and calculates all the statistics for us so all that is required is to reproduce and analyze them. For the sake of expediting this endeavor I am not including and class who's highest DPS value on record is lower than the average rogue value.

Hunters
Average: 1871
S Dev: 325
# samples: 1949
95% interval: 1221 - 2521

Mage
Average: 1935
S Dev: 213
# samples: 2205
95% interval: 1509-2361

Paladin
Average: 1504
S Dev: 328
# samples: 726
95% interval: 848-2160

Warlock
Average: 1929
S Dev: 297
# samples: 2877
95% interval: 1335-2523

Warrior
Average: 1518
S Dev: 494
# samples: 1265
95% interval: 530-2506

Rogue
Average: 2118
S Dev: 237
# samples: 1975
95% interval: 1644-2592

Now, what does all this mean? Quick and dirty, for the uninitiated:

Average = Average value
S Dev = Standard deviation, a relative measure of how much the DPS of a given class varies from the average. Larger numbers are indicative of wider swings in DPS from parse to parse. Lower numbers demonstrate consistency.
# of samples = the total number of DPS values for a given class in the sample
95% interval = the range of DPS that 95% of the players of a given class will fall into

In addition to the numbers above we know some general theories about raid DPS, namely, that melee DPS is grouped with other melee DPS so they can feed off of the same party buffs. This means that, directly off the top, mages and warlocks could never be considered as replacements for rogues as they would not be buffed by their party in the melee group.

First up we have hunters, which can buff the melee party DPS by roughly 3% each and can derive a small amount of benefit from the buffs in party assuming the shaman is twisting. Assuming a party makeup of warrior, enhance, and ret with two open slots, one filled with a rogue for highest average group DPS, adding a hunter would add, on average:

1871 + 1518*.03 + 1504*.03 + 1670*.03 + 2118*.03 = 2025.23 DPS

This is plainly less than the average DPS added by a rogue by nearly 5%. Meanwhile, hunter DPS is much less consistent than rogue DPS so on a typical attempt this number could range rather wildly from 1378 to 2675

For paladins and warriors the case against is much much simpler. The raid DPS contribution of a second paladin or second MS warrior is significantly smaller than the fist, so much so that it is practically negligible. With the numbers we have here it is highly unlikely that any one 25 man raid is fielding more than one warrior or paladin that is capable of putting out rogue numbers and on average the second warrior or paladin is adding 500 - 600 less DPS, or 25%-30% less DPS than the average rogue. Furthermore they suffer from the same consistency problems as hunters, even if you factor out the number of warrior tank parses that are deriving up the s dev for warrior numbers the amount of consistency you would get from a warrior or paladin is significantly smaller than a rogue.

You will notice I haven't even listed shaman, you can look them up yourself if you wish, but their highest DPS posted is lower than the average rogue DPS which is as an open and shut case.

The actual facts we have in hand at the moment indicate very very plainly that if there are open spaces in the melee group and all of the group or raid buff slots have already been filled that they are best filled with rogues. Rogues not only provide a raid with the highest average damage in a single raid slot but they also do it consistently. Consistency is what makes farm kills farm kills. Consistency is what drive progression, wiping on a boss you have beaten a dozen times because everyone got unlucky is the kind of thing having a rogue or three in your raid is going to guard against.

Neither the high average damage, nor the consistency of that damage, is going to be something a raid in WotLK is going to want to do without. Rogues most certainly do not require a raid buff, party buff, or a buff to personal DPS to remain raid viable for 1-3 slots which is nearly the same representation that any other class can be expected to enjoy in any given 25 man raid.

My vanity is justified.

 

Dorvan

Good post, but this section seems very hand-wavey, which is always a red flag for me as assuming the status quo is correct (yes, I know there are lots of favorable assumptions towards evis, but we may as well get the right answer). As far as energy, I'm horrible with haste math, but at 1.0 speed you'd get an expected 15 energy with SnD up over 5 seconds. Without SnD, weapon speed would be what? 1.35? So you'd have an expected 11.1 energy gain with SnD, for an expected energy loss of 3.9 energy. You didn't include an average SS number, but if Left's 1600 is reasonable (correct me in I'm wrong), that means a loss of about 160 damage from combat potency. That still leaves an 800 damage advantage of eviscerate over 5 seconds of SnD. The Rupture time I'm less sure of how to calculate, perhaps you could provide some more insight there, it just wasn't intuitively obvious that "oh, when you consider these factors, the math works out to favor the 5 seconds of SnD over Eviscerate even in this case".

 

◊ gwystyl

Isn't this a specious argument? Why is a (pure) melee dps group required? The classic MS warrior/Enh Shaman/3x other melee group isn't a requirement. Stacking a Feral, 3x BM hunters and a resto druid would provide tremendous synergies and allow you to condense to 1 melee group that housed the tanks and whatever else you wanted to group with them (shaman for threat, trees for healing buff, etc).

We've seen a rogue-less bleeding edge progression kill. I don't recall any celebrity kill shots with 0 warlocks, mages or hunters?

 

• Vulajin

Well, the point wasn't really to demonstrate necessarily that Evis did not trump a couple seconds of Slice in the single very contrived situation, but rather that the parameters required for Evis to come close to or top Slice are ridiculously restrictive. Even if we say that that 1000 damage gain is worthwhile and completely ignore any considerations of the extra damage lost to Slice and Dice "feedback" into other effects -- we're still gaining 1000 damage only if two finishers proc Ruthlessness and we have double Lightning Speed procs and Ashtongue Talisman up for our Eviscerate (which, by the way, to be able to 5-point Eviscerate within 10 seconds of a 5-point Rupture is also very restrictive in terms of number of CP procs required) and we experience at least 9 CP procs during the cycle (around 10% likelihood, maximum) and all of those aforementioned buffs happen to drop off before the time Slice drops and stay off until we get it back up.

In other words, the likelihood of actually reaching the situation where it is worthwhile to 5-point Eviscerate in a 5/5 cycle is extremely rare. If you relax any of the factors mentioned, what happens? On inspection, Eviscerate's DPS contribution decreases further.

Does this type of situation happen? Surely. I'm sure that a skilled rogue would be perfectly aware of a situation in which there were enough energy and enough time left in his cycle to Eviscerate and still manage his next Slice perfectly on time. And the entire discussion here ignores bigger haste effects like Bloodlust and Blade Flurry (especially the combination of the two), which cause temporary spikes in energy intake. However, a combo point cycle is not intended to encapsulate these contingencies. It is meant to provide you a baseline for utilizing your energy and combo points, from which you may deviate in those situations in which it is appropriate to do so.

Still, I think the general rule that it is not worth Eviscerating if it will come at the cost of Slice uptime is a good guideline. Yes, a fraction of a second of Slice being down might only mean a few hundred damage lost if you are not under the effect of other hastes at the time. However, the more hastes that come into play, the more damage that fraction of a second of Slice uptime is worth. Also, keep in mind that haste doesn't necessarily apply linearly over a period of time; letting Slice drop even for an instant is essentially playing the lottery on whether your swing timers will hit zero during that instant. If they do, you lose the hasted damage of one or two swings, not necessarily the prorated DPS numbers given above in my post. In many cases, this loss will be well above the numbers I've given.

 

• Maestroquark

I see Shaman higher than the Rogue average. Also, I'm not seeing a way to force exclude tanks from the DPS list. The notes say a 500 DPS threshold, but with a quick check I found reports with tanks > 500 DPS.

 

Dorvan

Yes, I think there may be a few Prot Pally data points in there as well. You'd also want to separate Warriors into Arms and Fury as well, as the 2nd Warrior you added to the group should be Fury, which is much higher personal DPS than Arms.

 

• Vulajin

While I respect the time you obviously put into your analysis, I have to question the use of WWS Scoreboard as an analytical tool for this type of thing. There are many confounding variables, not the least of which is player skill. It may be true that, on average, hunters output less DPS in Brutallus parses over a certain subset of guilds. However, does this say that hunters are inferior DPS, or does it say that hunters are played by a larger proportion of unskilled/uninformed players, or does it say that hunters generally do not receive as beneficial group synergy as rogues receive?

As long as we're using anecdotal evidence, I have seen my guild's BM hunter (The World of Warcraft Armory) output DPS very close behind our top rogue (The World of Warcraft Armory) when given a 2 BM hunters/survival/feral/resto shaman group and two Bloodlusts (and receiving Burn once and having to move), while the rogue received a 2 rogues/arms/enhance/ret group and three Bloodlusts.

 

Left

Hmm, that certainly puts rogues more convincingly in a better light as DPS leaders. I do want to point out something that you touched on but didn't really elaborate on, however: spec differences. I would venture (and it would be interesting to investigate the data further) that the Paladin, Warrior, and Hunter numbers exhibit lower averages and larger standard deviations because those classes each have a top DPS spec and a lower DPS spec represented in a typical raid. In the case of warriors and paladins (particularly for Brutallus), the lower DPS spec is a tank. This will greatly pull down their averages and greatly increase their standard deviations. For hunters, the off-spec is Survival. Many (most?) min-maxing raids bring a Survival hunter these days. Survival is not a top DPS spec, but it represents 1 out of every 3/4 hunter DPS figures by virtue of there usually being one in the raid to buff the other physical DPS. I'm going to guess that the numbers for Ret paladins, Fury warriors, and BM hunters are higher and with tighter standard deviations than the overall numbers you posted.

Another issue regarding pure-DPS versus buffing classes is raid location. For a Brutallus kill, the raid stacks itself so as to do maximum raid damage. In such a stacking mechanism, rogues come in a stacked group or else not at all, since rogues have nothing to contribute outside of damage. If you have an extra slot in a shadow priest/lock/mage/ele shaman group and you have your choice between a rogue or a BM hunter, what do you put in there? Naturally, the hunter. He'll do less DPS than he would elsewhere, but at least he gives 3% damage to the rest of his group (which the rogue would not do). Same analogy for putting a DPS warrior in the tanking group for another shout, putting a Retadin in another raid slot where s/he can still keep up judgments, or putting a hunter in the tank group for increasing tank threat. In summary, do rogues do higher DPS because they can, as a class, put out higher DPS, or do they do higher DPS because they don't have non-DPS specs or non-stacked-group raid functions which are pulling down the averages?

EDIT: Seems Vulajin more or less raised the points I wanted to before I finished my post...