Thank you Bowen for the reply.
I wanted to know if those numbers are specific to a combat build only or if they apply to a mutilate build aswell.

Shaker, would you mind explaining to me why this WF change affects hit so much? I am quite puzzled about this as I canlt find any explanations about this anywhere.

Its a combination of things. AP now affects poisions, well. More hits do not buy you more wf procs. Add it all in, and hit cap isnt something to aim for anymore. Im running 32.5 crit, 20% armor pen, 259 hit, 2325 ap, 19 expertise, and a passive 10.15% haste. Raidbuffed, 5/51/5 build with agi food/agi pot, I ran 4200 dps on a 3 minute brut kill. Dropping excess hit and turning it into agi had a good affect.

Here's your answer about TBC Windfury and why it made hit so valuable for rogues who had an Enh Shaman in their group.

I.e. more hits created more WF procs and more damage. There's other things going on there but it's not important anymore. Just note that the removal of the TBC Windfury mechanic is part of the reason rogue poisons were made to scale.

Mutilate or Combat/ss for both, personal preference basically which you want to use, so at 70, spec to match your gear or what you have access to, or just want to have fun with since everything is so easy now.

Mutilate will pull ahead a little bit on murderable targets (humans,giants, beasts, dragonkin), but it is more difficult to play optimally, hence the personal preference thing.

Actually the fundamental difference isn't that it created more WF procs - WF's interaction with Rogue DPS was rather nicely laid out as such:

AP - caused WF procs to hit harder
Crit - caused WF procs to hit harder
Haste - caused increased chance for WF proc to happen

Hit - caused increase chance for WF proc to happen AND increased WF proc damage by reducing number of WF attacks that missed.

That "double dipping" is one of the major contributors to rogues being so hit-focused in TBC. Now that the WF mechanic has changed, we are no longer as interested in hit. It's a good stat, but we have lots of good stats, and currently Agi/AP are showing out better than everything else.

As a combat rogue, I have been using ClassTimer since tuesday and it has performed flawlessly.

I ran AToL and Naaru sliver all night. I also ran dp mh, wp oh and people keep telling me that is the wrong combo. I have a stasis parse up, but Im asking the host if I can link it first, Im not sure what sort of traffic he can sustain.

Eternal Reign Raid History

Couple notes. We didnt wipe on felmyst, not sure why it shows 2 attempts.
We wiped on muru due to confusion the first attempt.
We wiped on kj due to a bomb starting just after the orb buff had worn off of our dragon controller. He had kept a dragon from phase 2, into phase 4 and its duration expired. We had no bombs in p2-3.

So I guess I should have been a bit more clear in my post (it was late, I'm sorry). However, you kind of completely missed my point .

So first of all, to be clear the gemming advice for WotLK has been something like this. Fill all your red slots with Bright Scarlet Rubies. Only put one Balanced Twilight Opal in one blue slot for meta, everything else stack Bright Scarlet Rubies. For yellow you may as well put in either Glinting or Wicked Monarch Topaz, since the socket bonus will be better.

However, it seems like for my gear level, even with the calculated "on spreadsheet" DPS values I should gem for Expertise. This makes me think that there either a.) is an inflection point where gemming for Expertise becomes > gemming for AP or b.) perhaps Expertise was under appreciated.

So about the spreadsheet modeling. First of all, Vulajin's sheet currently does not actually model the variable number of Envenoms in your cycle. It's just a simplified 4+n/4+r cycle. So actually modeling that will also slightly increase Expertise value I'd venture. Furthermore, even when it is modeled unfortunately with something as complex as the WotLK Mutilate cycle, it's just not going to be 100% effectively modeled by a spreadsheet. It's not something that spreadsheets specifically are good at. I'm sure Vulajin and Aldriana will come up with excellent models, but they will probably only be 98 or 99% accurate. It's one of the things a DPS simulator would slightly more effectively model. That's fine, it doesn't have to be completely 100%, but my point was that if the "calculated" DPS loss was 20 DPS, when you're doing over 5.2k DPS, 20 is insignificant and well within the margin for error. I'd err on the side of it not appreciating Expertise quite as much as opposed to over appreciating.

As for why expertise, it's not to make the cycle "simpler" or anything along those lines. "Go combat LOL" is a pretty stupid thing to say without actually evaluating it. Rather, it's to make the cycle more efficient. Take a look at this scenario:

Your Rupture just faded. You have 5 seconds left on your SnD, 40 energy, and 4 CP. What do you do?

So if your expertise is capped, it's safe to immediately hit a 4 CP Rupture, then you have a high probability of immediately building 3-4 CP and be able to Envenom in time to refresh your SnD.

However, if you aren't expertise capped, that's not safe at all. If your Envenom gets dodged your SnD will drop. So to be safe, you pool energy 2-3 seconds and Envenom. However, you've now lost at a minimg 5-6s of Rupture uptime.

That's sort of what I'm talking about...maximized expertise lets you optimize your cycles, and I'm not sure that effect is well-represented by spreadsheets.

I find myself in the same predicament i had gotten so used to using rogue-power-bars for tracking my procking buff that i am desperately searching for a replacement. Can anyone reccomend a suitable bar timer for tracking proc or item buffs?

Numerous suggestions were given by various people just a few posts after the one you quoted.

Some of these are:
Needtoknow
classtimers
slicecommander
discodice
dottimer

For those wondering the exact AP values per rank of improved poisons where wound and instant diverge (for spreadsheet calculations), here they are listed below.

770 ap imp 5 = wound 71.4 average
1365 ap imp 4 = wound 83.3 average
2356 ap imp 3 = wound 103.12 average
4341 ap imp 2 = wound 142.82 average
10290 ap imp 1 = wound 261.8 average


For the obviousness, since some people miss it from time to time. At the above listed AP values with the above rank of improved poison talent, wound poison and instant poison are the exact same average damage. Any higher AP will net you better damage with IP than with Wound.

I was trying to find the divergence point to better sort out where to stop with improved poisons and put into relentless, figuring that this was the point break where imp poisons would be better than relentless.

I have nearly identical gear to Wodahs, barring his SWP Chest (damn you, RNG). Last night we cleared SWP carrying 3 toons that were buying gear and a green machine dps re-roll rogue. My action bars suddenly disappeared mid fight on Brut so I was flipping / clicking out of my spell book (no joke). I posted 3900 dps over 3:28 compared to Wodahs ~4200 over 2:36. So, similar findings here. It does bear noting that there was absolutely no semblance of a cycle for me. There were times without AR that I had so much energy incoming I couldn't dump it fast enough because of the GCD.

My guild's ret is very talented, but was unfortunately unable to raid last night so I can't tell you where he stood in relation.

More important than this would be if non-glaive set rogues could share their observations. That's a truer State of the Union.

Classtimers works fine, you just have to make sure you delete the saved variables from pre3.02. For energy watching, combo points and SnD Icehud works very well, as well as having health, target info, CC timers, and a threat bar all nicely laid out around your character.

I'll see if I still have my logs from brut when I get home tonight to try to get a parse (when I learned WWS wasn't working I didn't bother paying attention to cleaning them up, but Loggerhead means they were saved, at least in theory), but with my gear (glaiveless, and no crafted chest, but otherwise pretty much everything a Combat rogue could want up to and including Brut drops), I posted 3300 on brut last night according to Recount. Our kill was around 3 mins and change if I remember correctly, with our Ret Pally putting out about 2600-ish. I ran a 5s/5r/xe cycle, w/ E being variable on Compot procs. Popped Killing Spree twice, and energy capped when enough haste procs lit up at once - I'm going to experiment a bit w/ Shard of Contempt in place of DST tonite to see if it makes energy management a little easier.

Check my armory, the gems/enchants/gear are slightly off (the armory seems to be buggy and showing partial pre-3.0). My guild started Sunwell in July, we were working on M'uru before the patch. (That should give you an idea of raid gear level.)

Wednesday Night Brutallus:
Spec'd 5/51/5; Glyphs were SS and SnD; 6/8 T6, Cursed Vision, Tabi Boots (got my T6 last night though), Kara cape, SSO neck, DST, AToL; gem'd heavily for agility. I ran as smooth of a rotation as I could last night but I didn't have working timers and popping Killing Spree was a bad idea as it threw my little Gnome body right to the center and spun me at least once losing DPS on the way.

Most of my rotation was 5s5rXn, I didn't get as many envenoms as I wanted as without timers my SnD time was buried in buffs and I'm sure I overwrote it by 10-12 seconds on several occasions. I really was just face-rolling the whole encounter.

Despite all that, Recount put me at just under 2800 DPS (Valujin's sheet put me at ~2950). I was number 2 on damage just behind the ret pally who is an excellent player although he admits he was pretty much face-rolling it as well. With timers and some practice (I just had my time in SW that night), I'm sure I could get much, much higher. Pre-3.0 I was between 2200 and 2300 on Brut with a single Heroism.

I plan on spending some real time with the dummies now that I've got working timers again.

Edit: Running Deadly on mainhand, Wound on offhand.

What I wonder is :
WP/WP and Xs/5r/Xe rotations or
DP/DP|WP/DP|DP/WP and Xs/5r/Xn rotations?

Since my MS Office is a victim of Windows XP's Servicepack and Updates mess (pretty much unfixable) and OpenOffice won't work for the spreadsheet, I am currently not able to consider the sheet.

Then, "working timers": ClassTimer seems to be broken for 3.0, some alternatives? (in the ClassTimer style)

And has anybody information on the Rawr.rogue? Seems to be actually weird that with tier 6 gear, combat spec and raidbuffs a rogue does not top 2k dps...

I updated my post with the poisons I ran for that fight. Evis probably is preferable to Envenom but I was eager to try different things. Wound did keep the buff up nicely though.

I ran into the same problem on the Kalecgos fight.
http://elitistjerks.com/f31/t26242-r...40/#post939347

The problem with Rawr when it first came out is that it was poorly suited for rogue theorycraft. It simply didn't do a lot of things we want our modeling applications to do. Hence, I chose not to involve myself with it at that stage, and apparently most of the other theorycrafters made similar decisions. Instead, I chose to start working on my own thing, to model the stuff that rogues actually *do* care about.

Now, Rawr has certainly improved since then. It may now be sufficient for our needs - I don't know. But in the meantime I've set up a framework that I'm relatively comfortable with, and is better targeted for the needs of rogues (rather than being rogue information grafted on to a tool originally designed for other classes). So I'm not feeling particularly inclined to try to switch over to Rawr at this point. If someone else would like to, great; but I'm not that interested in it.

I think there's also a factor that the modeling tools rogues have available tend to be a little better than some of those available for other classes. I've always felt that the rogue spreadsheets - particularly the Gear and DPS sheets - were among the better, more accurate, easier to use sheets available across all classes (I omit Vulajin's sheet purely because - by his own admission - it's not really designed for user-friendliness. It's a perfectly good sheet, and it's reasonably accurate, it's just not quite up to the level of polish of the some of the others). Hence, because we have good spreadsheets, I think we've felt less need to explore alternate modeling approaches.

So, lets be clear: I have nothing but respect for Rawr - I think they've done a good job with it, and it does a lot of nice things, and I'm sure it's an asset for most classes. It just wasn't what rogues were looking for at the time it was introduced, so it hasn't really caught on.

With regards to my particular project: yes, I do plan on updating it for WotLK. Yes, I do plan to expand it to include Mutilate. Yes, I eventually hope to get a better UI on it. All of these things simply take time, which is a fairly limited resource for me right now. So it should happen, at some point. But it probably won't be ready day one of the expansion.

As long as we're on the topic: shameless plug. If there is anyone interested in helping with RogueCalc, I could definitely use the assistance. Even if you're not that up on the details of modeling, there's lots of work to be done in terms of UI and other functionality to improve it's usability - and, frankly, such things would be help for me in developing the models as well. So if any of you are interested in helping and have some programming skills, or modeling skills, or any such thing, please let me know - it'll help us get this thing up and running for WotLK that much faster.

In terms of modeling before the expansion: I frankly don't care. First, it's only for a month or so, so spending time to model it doesn't seem very efficient - I'd rather work on the level 80 stuff and get it out that much sooner. Second, even were that not the case, the content has been nerfed enough that optimizing DPS for raids really isn't that important right now. With the massive HP reductions and DPS buffs to many classes, raids are a game of getting the overall strategy and execution right, and the DPS really isn't that important. So I, personally, am simply not worried about optimizing for the moment. It's not needed, and it won't be at all relevant for very long, so I'm not going to worry about it.

I cant seem to find any definitive answer/post but could someone link me what everyone is feeling is the most optimal raid DPS spec for combat ( swords )? Are the last 6 points worth taking in combat to lose out on imp poisons from Assassination?

Regarding addons, I tried out the fan-altered Disco Dice from Curse last night in our BT zerg, and noticed a lot of oddness coming from it. It insisted I had a slice and dice going, when I didn't (according to my bufflist). It also seemed to pick up ruptures that other rogues put up on the mobs, rather than only timing my own. On the upside, it has a working smooth-energy bar. I did not look at the "regular" Disco Dice from Curse, nor did I try my old copy of DD (I haven't updated my old copy for a year or more).

I'm kind of thinking about keeping the fan-altered Disco Dice, and turning everything off except the combo point bar and the energy bar, and using Need To Know bars below that, sized approximately the same as the DD bars would have been.

So while we're talking about this, lately I've been thinking quite a bit about writing a DPS simulator. Spreadsheets/calculators are great for getting EP values, making gear comparisons, and all that. Without a doubt, we've had some great spreadsheets over the years in Rogue theorycrafting, and they are great resources. But I also feel like the focus on solely the spreadsheet/calculator front has resulted in some questions perhaps not being answered as effectively as they could be. For instance, I feel like a simulator would work better for things like:

1. Comparing/calculating optimal cycles (which is especially important with the new age of 3 finisher cycles)
2. Calculating the affect uptime variation (procs, cooldowns, etc) and the synergies they have on overall DPS
3. More accurately modeling fights of very short <3m durations
4. Being able to more accurately model the DPS affect of time off target

Now, most spreadsheets can model these fairly well, but there is some margin of error, whether it's maybe 1-2% it's hard to tell, but there's definitely some inaccuracies which are introduced by a reliance on infinite series and the like. The other thing is whenever some new item/enchant comes out with a crazy new proc, we have to add special-case modeling for it in the spreadsheet.

So the idea would be to write an extremely configurable framework, no calculations etc hard coded in the simulator, everything based on some simple scripting model. It would be very event-based as well. So if there's a new proc off of crit for instance, rather than having to jump in there and screw with numbers, you can simply be able to register it a "crit" event set what the behavior on proc is and voila. I'd also like to be able to add a mode where you can spit out a combat log for the simulated combat, and then that gives you more tools like Stasis to use to evaluate everything.

Now, I'll be honest I haven't really looked into any of the current simulator frameworks, including Rawr, so I don't know if something similar to this has been implemented for other classes. From glancing around though, it's seemed like most simulation models have been a bit more class-specific. I've got a lot of crazy ideas for how to make the "ultimate" simulator.

Unfortunately it comes down to a function of time. The way I'd want to design the system, with that level of flexibility, would be quite the time investment. I know it can be done, and I know I'm capable of doing so, but it would take a long time, and may be time I don't necessarily have .

This whole topic did remind me about something though:

So IIRC I believe Vulajin was saying a few days ago that the sheet isn't able to properly model fights shorter than a certain amount. The number may have been like 3 or 4 minutes, I don't recall exactly.

I actually just tested this out, so I can't be sure in a raiding situation, but the SnD bar was perfectly accurate for me on DD. Tested it against my Buffalo Icon and NeedToKnow. I'll check it out tonight during the raid to make sure.