He refers to even more basic things, like 40 agi ~ 1% crit, and 1 agility = 1 AP, etc.

I can see reasons both to do it and to not do it. Obviously despite any intentions, this thread has become a widely-used rogue information source, so such a section likely would not be wasted, but this thread fulfills a particular function and I don't want to bog it down with more than it needs to have.

I don't really see a lot of need to duplicate that information in the first post; perhaps a link to wowwiki or some such thing, but as that information is already readily available I'm not sure it's important to include here, particularly since, by itself, that information is about 90% useless. To the casual player who wants to know "which piece of gear is better, A or B", the existing equivalence table is more useful. To the slightly more serious player that wants to know how these stats improve his damage, the in-game tooltips which provide infformation at the level of "Agility gives AP and crit" are sufficient - the exact conversions tend not to be relevant. Really, the only people who need to know what the exact conversions are are those who are actually going to be theorycrafting, most of whom already know the conversions, and the rest of whom could easily find them in existing knowledge bases. So I guess I'm not seeing a lot of motivation to include it here.

I take it from what I've read here that all other talents remaining unchanged for a combat/sub build, Deadliness > Malice in terms of overall DPS (assuming you had to make that tradeoff - I realize both can be incorporated in a build)?

Sry, I think I misread this in my earlier post. If I'm sitting at 1500 AP then according to this rule of thumb, the 5% crit from Malice would be worth much more than the 10% AP buff from Deadliness, correct?

Well, the fact of which is "better" is irrelevant in the context of a Combat/Sub build. 99% of the time, a Combat/Sub build is used for the sole purpose of Arena, where Resilience is sky-high and 5% crit means almost nothing. If you're talking damage output in a raid, we may have something to discuss. If you're talking about what's generally an intelligent decision with said spec, there's no question that Malice is a poor choice.

Ah yes, ur bringing back memories of my initial logic. I am in fact primarily interested in pvp and I did go with the AP buff for that reason (resilience), and because I assumed the AP buff would scale. I posted the question here simply because the conversation was around talents and dps so I assumed it was appropriate. Thank you for the response and you answered my question.

i was reading

Combat daggers is now the worst of the combat builds as of 2.3

What made daggers so horrible ? I love daggers so much, i like to back stab. Just my preference, and playstyle. I hate the idea of being forced to goto swords to be highly effective as a rogue. My guild, our sword rogues top the chart by a landslide, then im up there somewhere under them and over most other classes, but the gap between me as a dagger rogue and them as a sword rogue is incredibly huge.

What would blizzard need todo , to get combat dagger rogues up closer to sword rogues ? I dont desire to be higher then them, but I do want to be competitive in dps to them.

I would highly recommend searching around, since there've been dozens if not hundreds of posts in the various rogue threads on this very topic, and, moreover, at least 1 thread was almost closed for speculating on what would be required to fix them - hence people aren't likely to want to open that particular can of worms again.

Maces is worse then combat daggers when i played around with spreedsheet. I get worse dps with s3 maces then my s2 daggers!
I also played around with s3 swords and the only thing that swords is better then daggers is because they have more talents points "over" . The extra point in ruthlessness, 2 in murder and last one in lethality. And the extra in 4 in imp poisons. So without those combat daggers has higher then dps then sword! Its rather extra talents points then the specc itself.

Just checked and s3 swords need 2xmurder to outpass s3 daggers in dps. Which means that on poison immune mobs and where murder is useless combat daggers has higher dps according to spreadsheet with my gear.

So simple solution, move improved backstab and opportunity to combat tree and daggers are where they should be!

Before stating such big claims, please make sure you link the according specs and mention the assumed gear level.

Hey, Cmecu (tell Grim that Hanos says hi). Right now daggers is going to end up being behind swords when you have similar options, the main reason this changed since before the expansion is that Sinsiter Strike got new ranks, and Backstab essentially did not. However, if you have better dagger options (and with a guild full of sword rogues you will, especially once you get Vashj down) then you can be very competitive with daggers. Before the Windfury and Sword Spec Nerfs, this was not the case, but now that Sinister Strike no longer proc WF or Sword Spec, the gap has closed substantially.

Up until recently my guild had 2 sword and 2 dagger rogues (one of the dagger rogues picked up Blade of Infamy and Blade of Savagery to save them from sharding), and the 2 dagger rogues are that way because they prefer the play style. At the highest gear levels you are looking at maybe a 3% DPS difference, which is small enough that until everyone has full ideal gear sets you aren't even going to notice it (T6 2-Piece bonus is a bigger DPS difference).

Taking a quick look at the other rogues in your guild and your armory, I see some issues that are also contributing to the differences. My guess is Taigori is your highest DPS rogue, and you are probably the lowest. The main thing is you are running way too low on Hit Rating, 227 is atleast 50 below where you want to be. The places where you can pick up some hit rating are also where you will see the biggest DPS gains:
-Ranged -> Arcanite Steam Pistol is still the best ranged weapon in the game as long as you are under the hit cap (363) drops from Al'ar
-Neck -> Choker of Vile Intent from Heroic badges, only costs 25, and is a nice upgrade over what you have.
-Cloak -> Either of the cloaks from Kara (Dark Reavers is the nicer of the 2), the PVP one from Badges, or the Haste one from ZA would all be very nice upgrades.
-Chest -> Bloodsea Brigand's Vest or the one from ZA are both big upgrades
-Wrist -> Heroic Badges ones or the ones from Astromancer are both big upgrades
-Waist -> Belt of Deep Shadow (Crafted) is the second best belt in the game and has a ton of hit on it
-Boots -> Heroic Badge ones also loaded with hit and a straight upgrade from Edgewalkers
-Trinkets - Warpspring, Dragonspine, Tsunami and the ZA trinket are all huge upgrades over you current ones

Comparing your gear to Taigori's puts you about 75 DPS behind him, and he has most of the upgrades I listed above (Ranged, Neck, Cloak, Wrist, Waist, Trinket)

So while, yes you can get higher with Swords, Daggers is fine, and the DPS difference isn't a big as you would think, not to mention taking less appealing weapons might make it easier to pick up more appealing ones like the chest off Fathom Lord.

Does anyone have any data on how the hemo debuff is actually calculated? The Tooltip says "adds up to 36 dmg" (buffed to 42 with next patch). In general it would seem that most people consider this to add a flat amount of dmg. I have always assumed it meant it added between 0-36 making it an avg increase of 18 dmg. Has anyone done any tests or posted an actual calculation somewhere?

Please don't propagate the 'magic number' hit rating thing - there is no magic number for hit rating (edit: as long as you aren't hit capped (363), or missing specials, that is). Especially when rogues move from T4->T5, following SEVERAL optimal upgrade paths, their hit rating can dip into the lower 200s while their DPS is going significantly up.

Hit is strong, but the stats on the T5 set pieces (only one of which has hit rating on it) is much stronger, and many other pieces exist in T5 that are strong but have less hit on them. (Shoulderpads of the Stranger, Tsunami Talisman, the neck off Kael, etc)

And just to take that one step further: even the point where you start missing specials isn't a "magic number". In practice, of course, it's almost never a good idea to drop that far; however, even at the point when specials start missing, it's still a tradeoff of "how much damage do I gain from this item versus that item" - the value of hit is just a bit higher. If you can get 10 AP for dropping 2 hit rating, it's still worth it, no matter how low your hit is.

That said, the gear recommendations in the rest of the post are generally good, even if the logic behind them is a bit off.

Tsunami Talisman has 10 hit rating. You might want to remove it from that list :P

Sorry, I mean the general trend from "lots of hit" to not much - I'll edit the post to be more clear with that, but when you're talking about replacing a Romulo's Poison Vial with a Tsunami Talisman or a DST, you're dropping 20-30 hit. The idea of the post was that someone sitting at 270 hit shouldn't go "Oh, hm, I heard I shouldn't go below 270 hit, should I replace my Romulo's with a Tsunami?" The answer is yes.

And yeah, I didn't mean to jump on Hanos' recommendations, they are very solid ones (though after running my own numbers, I've seen that the S3 BP is damn close to the FLK one, and I'm now using the S3 hat since it's better than anything in T5) - I just have struggled to stomp out the 'magic number' syndrome in my own guild, and well, call it a personal crusade.


And for purposes of definition, when I say "magic number", I basically mean where the graph loses continuity, and there's a noticable jump in value. It's around those 'magic numbers' that re-evaluations of gear choices have to be made.

A short list of the ones I personally believe in:
- 363 hit rating - after which hit rating has a value of 0 for rogues with 5/5 precision.
- 63 hit rating - below which specials miss and the value of hit rating increases noticably
- 63 expertise rating w/ Weapon Expertise (or 26 TOTAL expertise) - after which theoretically you should have no dodges/parries (assuming 6.5% boss mob avoidance, which is correct for dodge according to best data so far)

That's just kinda odd, how all of them have 63 in them. :P Spooky!

There's also an upper limit on Armor Penetration, but that should be ~2k on caster bosses, and ~3k on normal bosses.

1) ye exactly, why should a certain weapon specc in combat have more off talents to use then the other weapon specc? thats not fair.

2) im using RogueDPS_2_3_0_3.xls spreadsheet, maybe thats why.

Just checked RogueDPS_2_3_2_2.xls, newest version, with standard sheet gear(T6 eq), s3 swords vs s3 dagger mainhand and messeger of fate offhand and get the same results.
Checked rogue dps 0.8 sheet and it show a different view. So some differents between those sheets then.

I'm getting back into playing rogue after someone made me a 1000g bet that I can't do 1200 dps with a rogue in T4/ZA gear (I just so happen to have a toon at close to this gear level). I'm confident I can do this with a combat or hemo build but I am curious about mutilate.

I like the idea of "solving" the mutilate/sealfate problem. I think the way that I'll approach it is to write a combat simulator instead of the way the spreadsheet does it (which essentially calculates averages). It may take many trials of the simulator to get results that converge, but there's also value in knowing how much a spec can vary based on luck. A simulator makes it much easier to factor conditionals associated with the luck-based nature of mutilate.. you can make 4cp with 60 energy perform a different action than 4cp with 0 energy.. or based on how much SnD is left.

I think what's really missing from the mutilate modeling is understanding how to best utilize find weakness. Are there ways to leverage a lot of abilities in a short window of find weakness and then spend some time letting energy recharge to be able to do cycles in bursts (without ever letting SnD down, of course)? I don't think this is the kind of thing that you can model easily in the spreadsheet.

This isn't a promise that I'll write such a simulator.. but I will think seriously about it. =)

Doing 1200 DPS is going to depend 100% on debuffs at that gear level. In all BT/Hyjal level gear, my unbuffed DPS is only 903, however, my fully buffed, debuffed, potting DPS is almost 2k. Make sure you have an Enhancement Shaman (dropping SoE and WF), a Warrior with Imp BS, a Survival Hunter, a Warlock dropping CoR, a warrior sundering, a druid dropping Faerie Fire and keeping Mangle up, as well as Flask, Food Buffs, and Haste Pots. Without a full set of DPS debuffs and buffs you aren't going to come close to 1200 DPS. Also, only certain bosses are conducive to that level of DPS... choose wisely.

No, it's not, and this is why I've been toying with the idea of writing a DPS simulator for quite a while now. The issue is that there are quite a wide variety of procs and abilities that need to be properly modeled to obtain an accurate simulation. It's a non-trivial task, and most rogue mechanics are straightforward enough that closed form evaluation is satisfactory, hence why you don't see a plethora of rogue DPS simulators around.

Oh yeah, this is definitely raid buffed with consumables, which won't be a problem.

The bet was definitely setup in the context of raiding.

I agree that we're starting to run up against the limits of spreadsheets in terms of rogue modeling; however, I remain unconvinced that simulation is the correct answer. While it might be viable in terms of banging out custom calculations, I think it would be impractical for doing a gear-comparison type computation where you need to run your damage estimates many times. I think a better approach would be to write a program to do the statistical analysis, taking into account the exact probability distributions and their variance in addition to the mean (which is all most of these spreadsheets use). With such an approach you could be every bit as accurate as a simulation, with the added benefit of not requiring hundreds of hours of simulated combat per damage computation. The down side is that such a program would be a great deal of work to write... which is the primary reason I haven't done it yet. But I do think that if one could find a couple of motivated programmers to work it out, it would be the right approach.