Elitist Jerks Combat Guide for Cata [12/01/2011]

06/13/11, 8:26 PM   #466
Sarvius
Don Flamenco

Goblin Rogue

Maiev
 Originally Posted by Alash Roughly how do you evaluate weapon speed? E.g. how much more EP would an 1.80 OH dagger have to have to beat a 1.40?
An extraordinary amount. Offhand speed is a huge factor in instant poison dps and combat potency procs, which are in turn a large portion of our dps.

Here's why I don't have a better answer: Shadowcraft is showing offhand speed's value at -2260 for my gearset, and mainhand speed at 699. So going from a 2.6 mainhand to a 2.7 would be a bump of 699 EP, and going from a 1.4 to a 1.5 would be a loss of 2260 EP. Offhand dps is valued at 1.584, so you could theoretically do better dps with a white 1.4 than with an [Organic Lifeform Inverter].

I really don't think that's accurate (the number dump from shadowcraft may not be correct or maybe I'm applying it improperly), but the moral of the story is that you want a 1.4. If the numbers were correct....it would mean we should be using a librarian's papercutter. And probably cruel barb.

Can somebody show me why this is all nonsense?

 06/13/11, 11:21 PM #467 • Aldriana Mike Tyson   Aldriana Night Elf Rogue   Doomhammer I guess I'm not sure where you're finding those EP values in Shadowcraft; I admit I haven't used the UI for it very much, but I don't see the weapon speed EP values exposed anywhere, so I'm having a hard time diagnosing the problem in any detail. That said: they do sound roughly an order of magnitude high relative to what I recall seeing last I investigated this, so is it possible that those are EP values for a 1.0 difference in speed instead of a 0.1 difference in speed?
 06/14/11, 4:52 PM #468 Sarvius Don Flamenco   Kensington Goblin Rogue   Maiev Of course. That was a silly mistake on my part. So .1 speed is worth 226 EP which is in turn worth ~142 weapon dps. Which means that a 1.8 needs to be 904 EP more valuable in order to be superior to a 1.4. I just ran the numbers on a 308 [Toxidunk Dagger] (the tooltip will display a 346 that doesn't exist, but I used the level 78 version) and it was worth ~874 EP while heroic [Organic Lifeform Inverter] was worth 1550 EP in my gearset. Which means switching from a 1.4 308 weapon to a 1.8 372 weapon will lose you about 220 EP. So: You want a 1.4. @Aldriana the numbers were contained within the Pawn String below stats on the gear tab. Weapon speed was the only useful stat that isn't displayed in stat weights (and rightly so based on these conclusions). You've obviously already diagnosed the problem fully, but that's where they are.
 06/14/11, 5:02 PM #469 • Aldriana Mike Tyson   Aldriana Night Elf Rogue   Doomhammer Well, the other reason is that weapon speed EP values tend to be highly nonlinear. If you look at the damage difference between a 1.4 speed weapon and a 1.5 speed weapon, and then try to use that to extrapolate the value of a 2.6 speed weapon, you'll be *very* badly wrong. There are ways that EP can be applied to weapon speeds, and for purposes of Pawn its probably better than nothing, but in general I think its a good idea to avoid putting forward information that is as easily misinterpreted as weapon speed EPs.
 06/18/11, 10:00 PM #470 Bananabacana Glass Joe   Troic Goblin Rogue   Kalecgos Is anyone else having smashing problems during AR due to energy pooling? With all the reforging to haste I could get, I was nearly 15 ene regen, and when I use AR as a cooldown right after KS on the boss, I have to smash the SS button like there's no tomorrow so I can try my best at not pooling energy, which is way hard. Am I doing anything wrong, or is it how it's supposed to go? During AR: SS (4x or 5x) keeping SnD up and Rupture and keep SSing (4x or 5x) so I can dump with Eviscerate. Man, that just feels plain wrong.
06/19/11, 1:19 AM   #471
Sulphuric
Von Kaiser

Human Rogue

Frostmane (EU)
 Originally Posted by Bananabacana Is anyone else having smashing problems during AR due to energy pooling? With all the reforging to haste I could get, I was nearly 15 ene regen, and when I use AR as a cooldown right after KS on the boss, I have to smash the SS button like there's no tomorrow so I can try my best at not pooling energy, which is way hard. Am I doing anything wrong, or is it how it's supposed to go? During AR: SS (4x or 5x) keeping SnD up and Rupture and keep SSing (4x or 5x) so I can dump with Eviscerate. Man, that just feels plain wrong.

It's quite normal to energy cap during AR, and well, it's inevitable.

 06/19/11, 10:29 PM #473 nextormento Von Kaiser   nextormento Night Elf Rogue   Uldum (EU) Color me confused: if that were the case, you'd need to extend that to a fourth and a fifth proc and so on, which is a geometric series that evaluates to: $x + x^2 + x^3 + ... + x^n= \frac{1}{1-x} -1=\frac{x}{1-x}$ That is not a J-curve, nor an exponential one: it's a rational curve with a vertical asymptote for P(main_gauche_proc) = 1 = 100%. That is, the closer we get to the mastery cap the closer our damage gets to infinite (which is fairly different than what we had with arp). Do we have logs showing such behaviour?. In fact, I though the first iteration on the ability -proccing as off hand-, was to avoid exactly this. I must confess I never gave it a thought after the change to it proccing as main hand strikes, but I very much doubt such a recursive retro-feeding mechanic could fit in the game as is. Besides, we do have records of procs proccing themselves, but they fixed that somewhere in the past. Last edited by nextormento : 06/20/11 at 5:36 AM. Reason: latex
 06/20/11, 4:41 AM #474 Aminagorum Glass Joe   Aminagorum Orc Rogue   Frostwolf (EU) Hi, yes, you're correct, that we hadn't an infintie increase in damage wit arp but i just wanted to compare the characteristic of them both in increasing our damage superproportionally by stacking it up futher on. If it was fixed in the passed, why is there an assumption by the statvalues and some people on this forum, that it is getting more attractive in damageincrease the more you got? Is it, that with more mainhandactions/min you could get your IP procc close to the very optimum (which is already capped i think and only be triggerd by DP proccs furthermore out of internal cd)? That would mean, that you would stagnate at a certain level if it's the poisonal explanation. Is it a combination of having more swings and getting your enchants at a higher uptime? Could it be having a higher uptime on different stuff like trinkets? As i said, i can only think about the fact of coming close to infinite chainreactions in swinging as an explanation to this assumption of our Statvalue.
 06/20/11, 6:46 AM #475 nextormento Von Kaiser   nextormento Night Elf Rogue   Uldum (EU) Such claims have not been backed up. The term J-curve itself is completely foreign to me (I've heard it in CNN once or twice and that's about it) so, if the claims fail to caraterice the behaviour of mastery in the logn run, I'm not willig to believe any of it. See that exponential growth of stats is something that still happens for certain specs but that's something we cannot say about our stats without deeper study; under the real testing we can find all through these forums mastery seems to behave just like any other stat. Also, SimulationCraft has its place, but computing stat weights is not one of its strengths. With current gear -and possibly next tier of content- we are able to use the energy we get on average; furthermore, we usually can manage the surplus during AR/KS. On the other hand, energy capping is a real issue with the spec; also, historicaly no other class has ever been closer to push the abilities per minute ratio further than 40 (which ties in with the lack of decision making). Saying that mastery *can* be stronger when we start to inevitably cap during strings of combat potency is all there is to it. If/when we run into those issues we'd need to assess not only the haste 'tip point' but also how much stronger mastery becomes (if it does). I believe claiming that haste has diminishing returns or the J-curve -that I still cringe over when I read it- is not such analysis but merely guesses this far. Last edited by nextormento : 06/20/11 at 6:53 AM.
 06/20/11, 6:55 PM #476 Scoparli Glass Joe     AlgebraÃ*c Goblin Rogue   Kil'Jaeden I had a hard time keeping up with all of that but is it confirmed that MG itself can proc another MG?? Something like that seams unlikely because of how strong it would be in later content with high levels of mastery.
 06/20/11, 8:25 PM #477 Sakuratei Piston Honda     Sakuratei Night Elf Rogue   Frostmane (EU) As far as I know, Main Gauche cannot proc off itself (I know I've read it somewhere here on EJ but can't find it to reference). The J-curve is referring to Main Gauche being able to proc Combat Potency, which gives more energy for more Sinister Strikes, which brings more Main Gauches. However, the overall power increase of Mastery in this J-curve is still quite low (15% chance to proc Combat Potency on top of a 16%~ chance to proc Main Gauche to being with, as well as 2.66 procs needed to get the energy for one Sinister Strike), which is why the EP value of Mastery will not change to great effect in the near future (keep in mind that only part of the EP increases in value due to the J-curve).
 06/20/11, 10:26 PM #478 • Aldriana Mike Tyson   Aldriana Night Elf Rogue   Doomhammer I'm actually about 90% sure I've never seen any formal testing on whether Main Gauche can proc itself. When it was switched to a MH attack, the question was posed, and we sort of assumed that Blizzard was going to have it work the sensible way while we were waiting for someone to test it... but then no one ever got around to testing it. We should probably do that at some point. Fortunately, it should be fairly easy to figure out. Mutilate rogues run ~20 Mastery, so if you just respect Combat you should be able to get a ~40% Main Gauche Proc rate, at which point 100 MH autoattacks should be plenty to deduce whether its chain proccing or not. So its literally a 5 minute test - we just haven't gotten around to doing it, afaik.
 06/21/11, 1:00 AM #479 liftir Von Kaiser   Snowcrashed Worgen Rogue   Ysera I seem to recall some testing done on this in the past but can't find it. In any case, it would appear a MG attack can not proc itself. I did a 10 minute test at a training dummy with 1725 Mastery (35% proc chance), MH weapon only with Bouquet of Red Roses in OH just in case there's some strange interaction of not having an OH equipped. 300 melee attacks, 97 Main Gauche procs. Here is the log, it appears to of cut off some of the testing as only 213 melee attacks (69 MG procs) were recorded. You'll have to do a "Show events where source or target is Snowcrashed" to clean up the log browser a bit. While I was watching the combat log in game I saw no Main Gauche procs back to back, there was always atleast one melee attack between Main Gauche attacks.
 06/21/11, 3:14 AM #480 Aminagorum Glass Joe   Aminagorum Orc Rogue   Frostwolf (EU) If it cannot procc by itself the it's getting even worse by not only not having mentioned the internal cd and limitation on how many seconds it cannot procc again (like IP), its further on the question: Has there an Mainhandautoattack to be in order to have the possibility to procc again or is there even a timed internal cd like 1 or 1,5seconds? If it's the second than a lot of haste would also decrease the value drasticly during Heroism/Bloodlust and AR, which would be like "Make your choice, haste OR mastery".

 Elitist Jerks Combat Guide for Cata [12/01/2011]