Nah it doesnt defeat any purpose since its a great sourch for dagger rogues aswell.

Ty Vulajin for all the work btw!

Such things do have a very valid place in a discussion about what you would do in reality.
It allows you to see what make up produces the best results for a given objective.

You can use what you learn from that to make up groups in a Raid because you know how the buffs interact with the members of the party.

I also made no mention of being in a Raid or having to account for raid efficiency.

Too much was being read into what I typed, all the things about "raid", "best for raid" etc came from others. I try to keep what I say simple, add no hidden meaning and I try not to require that the reader has to make assumptions.

I shove the BM Hunter in the "melee group" because if the pet is not a caster then it will be doing melee and be buffing the party doing that. With all due respect to Hunters I am only interested in his pet in my Party, the owner is literally secondary to my needs.
Call it an ideosynchracy of mine if you will.
Hey, I am not disputing that Swords is better than Daggers.
I see the results in reality, where the metal meets the meat, to be different. If I were doing less dmg with my gear (Malchazeen, Heroic Dagger) than a Vindicator/Latros equipped Rogue I would be tearing my hair out.
If I were not doing significantly more dmg I would be looking for what I was doing wrong.
The spreadsheet, any spreadsheet, gives approximate results under ideal conditions.
While it is a guide it is not a reflection of what happens in practice.
As you said about my ideal group, it is not what happens in practice.

I would be horrified if the state of play in the game was actually that shown by the spreadsheet. I don't think it is either, mainly because my own experience tells me that it isn't.
Fact is that itemization -does- play a role and I do contend that Blizzard use itemisation as a tool to balance things out.
There is very strong evidence of this in every patch note where they change item stats.
They didn't just drop the ball on this, they kicked it out of the stadium too.
It would not force every build to spec it, there is no need for Combat to do so.
The effect of 5% haste would not be greater than 5/5 Lethality (which is what would be dropped).
It would remove the need for any combat build to spec Sub at all, Sub Rogues would get a meaningful boost however.

Edit: Just to make this perfectly clear. It is obvious that both the mechanics of the Sword Spec Combat Build and itemisation give Combat Sword a clear advantage over any other Rogue build both in terms of DPS and in terms of playability. I am in total agreement with you on this.
My point about itemisation as far as Arena gear is concerned is that it has dramatically widened the gap and done it much earlier. Without this then things would not be as bad as they are now.

As far as the OP is concerned:
1- both theory and practice show that in order to maximise your DPS you should start with a Combat Swords build and equip a Rogue thus specced accordingly.
2- It is advisable to use Arena PvP as a way of improving your weapons to at least SSC/TK:TE standard or better.

I just tested this on the PTR and you're right. Surprise Attacks stacks additive for both Opportunity and Aggression. I'm going to update my formulas.

By the way, the wolves in this starting area have 20 armor (level 1) resp. 21 armor (level 2) (checked with a hunter). Sometimes this *does* reduce your damage range, although only by 1 (and only if your minimum damage is only slightly above a natural number, so that this 0.189% damage reduction bring it down to a smaller natural number).

The formulas are:

Backstab with Opportunity
( (DAMAGE+(AP/14*1.7)) * 1.5 + 255 ) * 1.2

Backstab without Opportunity, but with Surprise Attacks
( (DAMAGE+(AP/14*1.7)) * 1.5 + 255 ) * (1 + 0.1)

With Opportunity and Surprise Attacks
( (DAMAGE+(AP/14*1.7)) * 1.5 + 255 ) * (1.2 + 0.1)


Sinister Strike with Aggression
( (DAMAGE+(AP/14*2.4)) + 98 ) * 1.06

Sinister Strike without Aggression, but with Surprise Attacks
( (DAMAGE+(AP/14*2.4)) + 98 ) * (1 + 0.1)

Sinister Strike with Aggression and Surprise Attacks
( (DAMAGE+(AP/14*2.4)) + 98 ) * (1.06 + 0.1)

I don't think this is a productive discussion anymore, but I'd just like to conclude by reiterating that I am not at all in agreement with your point about itemization. Comparing any pair of daggers at a given itemization level against any pair of swords, the swords will always come out on top by a significant margin. Neither has arena gear widened the gap, as you can carry out the same comparison with PvE gear and obtain the same results. The two changes that really did it were making offhand sword spec procs cause main hand swings, and introducing a large amount of hit gear (enabling rogues to approach and even reach the cap) in BC.

(edit) I just noticed how contradictory this post of mine is. What I wanted to get at was that it's not weapon itemization itself responsible for the issue, as I believe you are arguing.

I think everyone else handled all the other parts of your post, but you're forgetting mainhand special attacks that trigger sword spec as well. Your opinion that it was not a major difference is different than mine, but I'm not really willing to go into it, since I think this thread has enough of non-math discussion on completely outdated topics (I am not simply blaming you for that - my posts aren't exactly numbers-laden either).

As for the attitude of "Well that's just theorycraft", I will say that I think you're really posting in the wrong thread if every disagreement is going to come down to that - that's why we're here. Several people have made calculations in "non-ideal" conditions, but there isn't much call for spreadsheets that deal with "optimum gear selections for the Gruul Encounter" - we have to choose something to generally agree with, and while we ALL realize that the tank & spank fights are limited in BC - the principles used in analyzing those fights are useful when engaged in any fight.

I've found in the gear spreadsheet that the weighting of hit rating versus agility is extremely dependent on just one item. With my T4-5 gear but no Dragonspine Trophy hit rating is worth just slightly less than agility (2.09 AP for hit rating versus 2.11 AP for agility). Add in the DST and the relative worth flips (2.18 AP for hit rating versus 2.13 for agility). I wear two proc trinkets (WSC and RPV) and have combat potency, but without the DST the sheet never weighs hit rating over agility.

When I discuss the value of hit rating with my guild rogues I always put a caveat on pushing hit rating and the hit cap - it's much more important if you have the DST. If not, then either hit rating or agility are good choices.

To make the point a bit more fundamental: the more powerful procs you have, the more hit is worth. Dragonspine, being a powerful proc, tends to push up the value; however, it's not the only thing. Combat Potency, Mongoose/Executioner, Windfury, etc. all increase the value of hit by nontrivial amounts. The more of these you have, the more hit is worth.

"Hey, I am not disputing that Swords is better than Daggers.
I see the results in reality, where the metal meets the meat, to be different. If I were doing less dmg with my gear (Malchazeen, Heroic Dagger) than a Vindicator/Latros equipped Rogue I would be tearing my hair out."

Maybe EJ should ban these kinds of statements? As is often the guy, a guy is generalizing from his own personal experience in a theorycrafting thread. He's wrong. The differences he is seeing are based on his skill vs. someone else. There are thousands of post in the other threads backing up the actual math that proves the superiority of swords over daggers. And he says but in the real world daggers are better.

The "reality" of which he speaks is solely himself vs. someone else he knows (or a couple of people, however many rogues he raids with). It serves as some kind of counter-illumination to have these "but I know my personal setup is better" posts when the opposite has been proved.

The forum bans signing of posts, perhaps it's time for no more I'm-sure-it's-true-even-though-it-isn't posts? Anyway, not my forums, but just a thought.

Well because poisons are considered Nature (not Physical) damage, it's fair to assume that they fall victim to the same level-based resistance rules as spells. 2 points in Master Poisoner decreases the 17% resist rate of a level 73 mob to 7%.

I think most people skip it because there are 41 better talent points in the Assassination tree if you plan to use Mutilate. Poisons contribute nicely to your damage, but they're not the hinge of your performance. A resist or two is nothing to nothing to quibble about when avoiding them would cost you valuable points like Quick Recovery, Fleet Footed, or something in your 20-talent support tree.

I have to pipe in on this one.

You will be missing so many "% chance" talent procs (energy, etc.) going with this 1s/3r rotation.

As I recall hearing someone say long ago, "It's better to know what you aren't getting than what you are." If that's not an exact quote, I reserve all rights to copyright it.

What are you talking about? 1s/3r keeps 100% SnD uptime - I'm assuming (and please correct me if I'm wrong) that you're stating that 1s/3r is dropping SnD.

I don't think this is entirely correct. Let's look at those talents individually.

ruthlessness: 60% chance to get another combo point with a finisher. The value of this talent degrades linearally with the number of combo points that you build. If you always do 1 cp finishers, than each combo point is always worth an extra 0.6 combo points. if you always do 5 cp finishers, than each combo point is always worth an extra 0.12 combo points. So the fewer combo points you build before using a finisher, the greater effect you get from ruthlessness.

relentless strikes: 20% chance per combo point to regain 25 energy when landing a finisher. This scales linearly with the number of combo points that you acquire before using a finisher. 1 cp is 20%, 5 cp is 100%. So it doesn't matter if you do 1 cp finishers, 3 cp finishers, or 5 cp finishers.. you'll get the same amount of bonus energy per cp.

Let's take an example of 20 combo points:

Scenario 1: 20 combo points built, finisher used every 5 cp, so 4 finishers. Ruthlessness will give you an extra 2.4 combo points and relentless strikes will give you 4 guaranteed procs for 100 energy.

Scenario 2: 20 combo points built, finisher used every 2 cp, so 10 finishers. Ruthlessness will give you an extra 6 combo points and relentless strikes will have 10 chances at 40% each of returning 25 energy... 10 * .4 * 25 = 100 energy.

It's correct to say that the energy gains from relentless strikes are more _consistent_ with 5 cp than with less than 5 cp.. but the energy gained over the course of the fight per combo point is unchanged.

A 1s/3r rotation gains more from ruthlessness and the same from relentless strikes as compared to 1s/5r... 1s/5r is simply more consistent with regard to the energy return from relentless strikes.

Rogue Combat talent Aggression now applies its 2/4/6% bonus to Backstab. This is in addition to the previous effects of 2/4/6% increased damage on Sinister Strike and Eviscerate.



- Shadowstep reduced to 30 second cooldown. 8-25 yd range. Changed from 20 yard. (previously no min range) Also adds +20% damage to any special on next attack.

- Hemorrhage debuffed increased from 10 to 36. Now lasts 10 charges or 15 seconds.


Goodbye, nerves of steel and 2/2 endurance, hello Aggression. This is a nice step toward a needed buff for combat daggers and I'm GLAD it didn't involve nerfing swords.

Subtley is also becomming more viable for pve and pvp purposes

Let me tell you my 'realcraft' experiences, then.

On my old server, in my old guild, we ran with 2 rogues for 25 man raids. Both of us started out as daggers. Our gear was very close. So was our position on the meters; we took the top 2 slots on most fights and it was a tossup which one of us took #1 on any particular day.

When Arena Season 2 came around, we decided to test swords vs. daggers by having me stay daggers and my guildmate go swords, in part because he was in a better arena team situation and could buy a set of S2 swords right away.

Result: I started losing the damage meter battles by a wide margin. Even after I completed my S2 dagger set (bringing our gear back to near parity) it still wasn't close. The only times I ever beat him were on low-armor bosses during the short period of time when I had Warp-Spring Coil and he didn't.

I'm on a different server now, so I can no longer compare to him, but I did recently get a Talon of Azshara and a S2 OH to replace my S2 MH/OH daggers. The Rogue DPS Spreadsheet predicted about 150 raidbuffed DPS gain from doing this without any other changes to my gear. I can report that it doesn't seem to be horribly wrong in practice; I'm not actually getting quite that much but the set of buffs I have enabled in the spreadsheet is probably a bit optimistic. It's certainly worth at least 100 DPS, which is substantially more than 5% gain for me.

So yes, sword spec really is that good.

Both specializations benefit a great deal from Arena items. The S2 OH dagger is vastly superior to anything you can get before Hyjal, and there's nothing to compete with the S2 MH dagger until Vashj.

I think you may be overestimating the impact of the availability of the Arena OH on swords, too. Consider a thought experiment: All else equal, what is the DPS due to weapon specialization for sword and dagger when changing between a 1 DPS OH and a 1000 DPS OH? Sword spec's contribution stays constant, because SS procs from your OH result in MH swings; changing OH DPS doesn't change sword spec DPS at all. The DPS from dagger spec, on the other hand, changes considerably since the 5% crit scales the DPS of the OH weapon.

So the point of my thought experiment is that sword spec is quite insensitive to OH weapon DPS. If you have to use a Latro's instead of a S2 OH, you only lose out on the raw autoattack OH DPS -- about 26 * 0.75, modified by crit etc. Latro's happens to have excellent DPS stats, probably better overall than the DPS stats on the S2 OH sword. It's also 0.1 faster than the Arena OH, and that means even more for sword spec than it would for dagger spec. Thus, Latro's is surprisingly competitive with the Arena S2 OH. Not as good, but not a disaster either.

I'm afraid, then, that you are going to have to be horrified. If you're beating a Vindicator/Latro's rogue, it's because that rogue isn't up to your standard in some regard (gearing choices, talent choices, player skill). Like I said above, I've seen two of the best real world scenarios to verify what the theorycraft says: both comparison between two closely matched rogues and personal experience switching from daggers to swords.

I think nobody would disagree that the Rogue DPS spreadsheet can't predict exact outcomes. (At one time pre-TBC it came close, but it really ought to be a Monte Carlo simulator these days to accurately model the interactions between haste procs, combat potency, cycles, etc. The game's too complicated now to model in a straightforward way with a spreadsheet.) But you seem to want to have it both ways: anything that's useful as a guide obviously does reflect what happens in practice to some extent, no?

Personally, I've had great success using it as a relative ranking tool. That is, if you set up your gear, talents, race, and buffs, save the DPS you get from that setup, then start playing with the options available to you, it's quite good at predicting whether your dps will go up or down, and even at predicting the relative value of different gear upgrade options. Do I rely on it as much to tell me whether spec A is better than spec B? No, but it's at least close, and when it predicts a huge difference between one spec and another, I'd take it seriously if I were you. People have done a lot of work on it to make it reflect reality as closely as possible.

I'd dispute this to some extent; while it's certainly harder to model things in a spreadsheet than it was before, it's certainly still possible; for everything but deep-Assassination builds, I'm confident that the Rogue Gear Spreadsheet is as accurate as any Monte Carlo method. And even high-end Assassination *could* be modeled with it - I just haven't bothered, since it is fairly challenging and not really that worthwhile.

Here's the thing, though. Over a given period of time, you generate a certain amount of energy from base energy regeneration, and you also generate a certain amount from Combat Potency if applicable. Both of these effects, we will assume to be independent of cycles (as they largely are).

Now let's say you gain 5 energy back on every single combo point generator you perform (let's use Sinister Strike as an example). Let's assume only the base energy regeneration of 10/s (if there were Combat Potency, it would apply equally to both cycles):

Running 1s/5r
Sinister Strikes used: 6
Finishers used: 2
Total energy spent per cycle: 40 * 6 + 25 * 2 = 290
Total energy back from Relentless Strikes: 6 * 5 = 30
Net energy spent: 260
Sinister Strike energy spent: 240
% energy spent on Sinister Strike: 92.31%
Cycle duration: 260 / 10 = 26s
Relentless Strikes energy/s: 30 / 26 = 1.15

Running 5s/5r
Sinister Strikes used: 10
Finishers used: 2
Total energy spent per cycle: 40 * 10 + 25 * 2 = 450
Total energy back from Relentless Strikes: 10 * 5 = 50
Net energy spent: 400
Sinister Strike energy spent: 400
% energy spent on Sinister Strike: 100%
Cycle duration: 400 / 10 = 40s
Relentless Strikes energy/s: 1.25

So not only does the cycle with higher-CP finishers use more of its energy, on average, on Sinister Strikes (thereby increasing Sinister Strike DPS), it also obtains higher energy output per time from Relentless Strikes. The effect, thus, does depend quite heavily on the cycle you run.

It's based on your tooltip weapon speed.

I spent some time over the weekend taking a closer look at what the Spreadsheet is telling me and what an event driven model I have been knocking together is telling me as well.

I must admit I have completely underestimated Sword Spec. By rights a Vindicator/Latros rogue shouldn't just be close behind me I in fact should be trailing them on dmg.

This is good news for me, it shows that at least 3 rogues in the world are worse than I am

Reading this and other threads about DPS cycles:
I am specced Combat Daggers with 3/3 Ruthlessness and Relentless Strikes.

Two questions then.

I am running 1s/3r comfortably but is this the best cycle to be going for in Raids?
The other option is 1s/3s/5r.

Quite often this can stretch to a 1s/4r but that often means a short gap in SnD which many say is anathema. Is it wise to stretch a combo that far? (the gaps are somewhere in the region of 2 seconds)?
Edit: My second question arises out of the fact that Aldrianna's spreadsheet is saying my Approximate Cycle is 1s/3.7r

mutilate off hand weapon = Vengeful Gladiator's Mutilator - Items - World of Warcraft

You do not want to let SnD drop, ever. Hence, the ideal solution is to watch your SnD uptime carefully and do 1s4r when you can and 1s3r if you can tell that that will give you an SnD gap. If you are having trouble with that, it is better to just do 1s3r than to have SnD gaps.

First up thanks Vulajin for this thread, I've found it extremely useful in helping to plan future upgrades/compare pieces of gear =)

I apologise if this has come up elsewhere but I've been reading conflicting views on the value of different trinkets for Rogues after the haste nerf in 2.2

Pre-2.2 DST was the best by a fair margin but after the changes in calculating haste rating --> haste % I've noticed some rogues preferring to go for the combination of Warp-Spring Coil and Tsunami Talisman (and later Madness of the Betrayer) as opposed to using Dragonspine in one of those slots, are there any post 2.2 calculations on this to back up one side or the other? Also what would be the better option between Warp-Spring Coil and Tsunami Talisman assuming neither would put me above the hit cap?

Sorry if I've posted in the wrong format, first time posting on EJ ^_^

According to the Rogue Gear Spreadsheet, DST is still the best trinket. What's number 2 depends on your spec and gear to some extent, but generally the answer is either Ashtongue Talisman of Lethality, Madness of the Betrayer, or Warp Spring Coil. Tsunami Talisman, while certainly better than the pre-raid trinkets, is on the whole inferior to these 4.