Combat/Sub Spreadsheet for 4.0.1 - Page 3

Rerox · 10/07/10, 7:14 AM

Originally Posted by sp00n

I'm showing about a 100 DPS drop if I switch from dagger to non-dagger (not sure if Slayer101 was meaning the same), so assuming the sheet is somewhat accurate, it's a no-brainer. Especially with virtually no mastery rating available at level 80.

That's why I am wondering: with 0 Mastery rating I see a tiny (6) dps increase when using the 1.5 OH axe instead of the 1.4 OH dagger. When reforging some of the tons of critrating I get on my gear with 4.01 to mastery (reaching 299 mastery rating) I get a 30 dps drop when using the dagger instead of the axe.

Since a 0.1 faster OH for combat nearly always meant "more dps" I am just wondering why I get these results, since even without the poison talents I expected the faster OH to be clearly superior.

/edit: by the way, great "cleanup" of the sheet.

Lightshadow · 10/07/10, 11:38 AM

Thanks for the graphics update spoon, I really like the changes. I'm going to make one or two more additions and merge it into the main spreadsheet. The haste issue you mentioned isn't a bug, per se, though it's something I would like to find a way to mitigate. The calculation for AR/KS is fairly coarse at the moment (1s intervals) and as a result there are haste points where you have just enough more energy to knock a second off the cooldown for both. As you can imagine, that's a fairly large DPS bump, which is what's driving the change you're seeing. I've got a couple of ideas in mind to make this a little bit smoother.

Side note: there's a bug in the crit cap where yellow attacks are being capped far too low (which is to say the crit cap appears to be much harsher than it is in reality). You're still never going to want to go over the cap if at all possible, but I'll get that fixed in the next release. Would it be worthwhile reporting the current distance from the crit cap in both agi and crit rating terms on the front of the sheet?

For those of you playing with different offhands, please remember that there are 4 things you need to change in order to test a different offhand: the speed, the dagger flag, the average base damage of the weapon, and any stat changes. I haven't (yet) found a situation without mastery where an equal ilevel axe comes out ahead of a dagger, but it's not impossible. I'd be interested in a copy of someone's spreadsheet if it's happening for them.

Edit: New version of the sheet with sp00n's changes, some bug fixes, and a much smoother response to haste. See OP.

Lightshadow · 10/07/10, 9:35 PM

A very minor update to the sheet posted. I added the black bruise proc (both normal and heroic) using exactly the formula in Aldriana's old sheet. The proc is now substantially stronger than it was before, mostly because with shorter cooldowns on AR/KS, MG procs, and extra haste its uptime is significantly improved. (~64% instead of 57). All in all the heroic proc is worth about 255 agility, enough to push it ahead of Heroic Havoc's Call by about 180 dps given the default gear in the spreadsheet.

Sleightt · 10/08/10, 5:43 AM

I may be reading this wrong, but when you say "wound poison and deadly poison are much closer than before", did you mean to say Wound and Instant? Or are you saying IP/WP or WP/WP could be better than the usual WP/DP or IP/DP?

Edit: Saw your post about it in another thread. Sorry I misunderstood.

Slayer101 · 10/08/10, 8:13 PM

Originally Posted by Lightshadow

A very minor update to the sheet posted. I added the black bruise proc (both normal and heroic) using exactly the formula in Aldriana's old sheet. The proc is now substantially stronger than it was before, mostly because with shorter cooldowns on AR/KS, MG procs, and extra haste its uptime is significantly improved. (~64% instead of 57). All in all the heroic proc is worth about 255 agility, enough to push it ahead of Heroic Havoc's Call by about 180 dps given the default gear in the spreadsheet.

Normal BB seems to be 100 dps better than Heroic Havoc's on the current spreadsheet... Anyone else finding the same results?

Edit: Redid this and finding Heroic Havoc's about 50 dps better than normal BB.

Slayer101 · 10/08/10, 8:16 PM

Originally Posted by Lightshadow

For those of you playing with different offhands, please remember that there are 4 things you need to change in order to test a different offhand: the speed, the dagger flag, the average base damage of the weapon, and any stat changes. I haven't (yet) found a situation without mastery where an equal ilevel axe comes out ahead of a dagger, but it's not impossible. I'd be interested in a copy of someone's spreadsheet if it's happening for them.

Edit: New version of the sheet with sp00n's changes, some bug fixes, and a much smoother response to haste. See OP.

Not higher but I am finding Scourge within 4 dps of Lung which is well within the margin of error

Lightshadow · 10/08/10, 11:49 PM

EDIT: This information is significantly outdated, see this post later down the thread for an update.

I've added a new tab to the combat sheet to do subtlety calculations, and I have a few results to share along those lines. I have to stress that the cycle modeled in the sheet is in some senses very optimistic about being able to use everything optimally. That said, it was intended to answer some very basic questions about the sub spec at level 80 after 4.0.1 and I'm reasonably happy doing at least that. If you'd like to use it, simply select "Subtlety" from the dropdown box near the projected DPS area and make any changes you'd like to gear and things on the left.

Conclusions from the spreadsheet

Even more than for combat, agility is a dominant stat. This isn't really a surprise given the difference in specialization bonuses. After that the sheet puts spell hit, expertise, and haste all about equal, though I know the sheet undervalues both expertise and spell hit a little bit, so it's likely worth capping those. All of the above are solidly above crit, and mastery is dead last by a mile.
The optimal non-subtlety points appear to be 3/3 precision and 2/3 coup de grace in precision and lethality. (3 Precision if your're not spell hit capped with only 2 points and 3 in lethality otherwise).
Initiative and looks to be approximately 0.5% dps increases per point and premediatation looks to be about 0.37%. (My guess is that overstates the value of initiative ever so slightly). They're definitely DPS increases, no doubt, but assuming blizzard is coming to similar conclusions, I'm not surprised they disagree with the prevailing view of how many "utility" talents the spec has.
Assuming you have at least 60% crit raid buffed, the value of the 3rd point of improved ambush is approximately a 0.4% dps gain. (The crit part of the talent is no longer helpful). This requires some pretty high end gear but isn't at all unreasonable. The DPS loss from dropping the 3rd point of Imp Ambush for Elusiveness is ~0.1%
While we're talking about it, the DPS loss for going from dropping a point of initiative for elusiveness appears to be roughly 0.18%, and you pick up some pretty substantial benefits for the loss.
Backstab and Shadow Dance are both very valuable glyphs. I've assumed the use of the hemo glyph (without it hemo isn't worth using unless you need the raid buff). It's not entirely clear to me that it wouldn't be a minor dps gain to forgo hemo entirely and just take the glyph of eviscerate instead. Neither slice and dice nor rupture are really worth using.
The priority system for finishers is more or less SnD > Recuperate > Eviscerate. Recuperate full time is a solid dps gain (plus the health). It should be the case that fully raid buffed there isn't a huge danger of Rupture dropping off even without the glyph, but it requires a little bit of forward thinking to be sure.
As Aldriana has argued elsewhere, a 1.8 speed dagger in the mainhand is a must. The offhand is much more speed/weapon type agnostic, since there are no normalized attacks with it in this spec the only difference is deadly poison. (Thus faster is slightly better, and a fast dagger will be optimal all other things equal).

Assumptions

I've assumed that EA flatly reduces the mob's armor by 50% and stacks multiplicatively with sunder. (this turns out to be true)
I've assumed that a HAT CP comes approximately every 2.05 seconds. This is pretty quick, but with crit rates as they are at level 80 I don't think it's unreasonable now that it's raid-wide. Changing it to 2.1 doesn't have a huge effect on the outcome.
I've assumed that ambush applies EA before damage is dealt. (this is not true, I need to adjust for it.)
I've assumed that the main effects of the glyph of shadow dance are to allow an additional ambush and add an extra 2 seconds/minute to EA uptime.

• Aldriana · 10/09/10, 12:23 AM

I think you mean Find Weakness, not Expose Armor. And word on the street is that Vanish doesn't trigger Master of Subtlety until the vanish buff wears off - i.e. until you've been stealthed and not attacking for 3 seconds. This may or may not be a bug, but assuming its intentional it basically kills Elusiveness as a DPS talent - it will only be in very rare situations that it will be useful in that respect.

As an aside, a while back (I think around the start of Wrath) Blizzard asserted that their goal was for a talent to be worth about 1% DPS. Now, a lot of stuff is worth more than that, and particularly with talent consolidation, that may have changed their target; but assuming that that's anywhere close to what they're aiming for, a talent that "only" does half a percent DPS is pretty clearly still a DPS talent - it's a weak DPS talent, about half as good as it should be, but it's unquestionably doing enough that it can't be completely neglected.

Lightshadow · 10/09/10, 12:32 AM

I definitely did mean Find Weakness, you're right. (Post corrected). That said, the DPS gain from elusiveness isn't coming from master of subtlety (I accounted for the way it's working on the PTR), it's from the extra shadowstep/ambush & 10s of Find Weakness uptime you get. If master of subtlety kicked in as well, elusiveness and initiative would probably be equivalent DPS talents.

Slayer101 · 10/09/10, 2:12 PM

Originally Posted by Lightshadow

A very minor update to the sheet posted. I added the black bruise proc (both normal and heroic) using exactly the formula in Aldriana's old sheet. The proc is now substantially stronger than it was before, mostly because with shorter cooldowns on AR/KS, MG procs, and extra haste its uptime is significantly improved. (~64% instead of 57). All in all the heroic proc is worth about 255 agility, enough to push it ahead of Heroic Havoc's Call by about 180 dps given the default gear in the spreadsheet.

Are you including the 93 AP of BB when your flag is set? If not, should we be adding the AP as agility (using the formula 93/2.1)? Adding the AP as agility I am getting over a 400 dps increase using normal BB instead of Heroic Bloodvenom (currently 2nd BIS) which is huge. 290 dps increase if I don't add the AP as agility which is still huge.

Lightshadow · 10/09/10, 2:30 PM

You can (and should) add the attack power manually (the flag is just for the proc), but the easiest place to do it is under enchants instead of as agi (since the agi also provides crit). The other thing that's changed since I posted what you quoted is that I updated the sheet to refect the fact that there no longer partial resists which works out to be a bit more than a 6% boost to the proc.

ABUSEDGOAT · 10/09/10, 3:34 PM

Originally Posted by Lightshadow

[*] I assumed MH berserking is up 57% of the time and OH 37%. This is pretty much what all the spreadsheets give for current gear levels, I assume it’s unchanged on the PTR.

I added the black bruise proc (both normal and heroic) using exactly the formula in Aldriana's old sheet. The proc is now substantially stronger than it was before, mostly because with shorter cooldowns on AR/KS, MG procs, and extra haste its uptime is significantly improved. (~64% instead of 57

I have a question in relation to these posts. Is it safe to assume that berserking (or mongoose) uptime in the OH is 37%? Shouldn't MG and the haste also increase its uptime? If MG follows similar mechanics to motes from Tiny Abom (motes can proc berserking), I'd assume that there would be a higher uptime for berserking than what the sheet currently lists.

Lightshadow · 10/09/10, 3:59 PM

Originally Posted by ABUSEDGOAT

I have a question in relation to these posts. Is it safe to assume that berserking (or mongoose) uptime in the OH is 37%? Shouldn't MG and the haste also increase its uptime? If MG follows similar mechanics to motes from Tiny Abom (motes can proc berserking), I'd assume that there would be a higher uptime for berserking than what the sheet currently lists.

The uptime for either on the offhand should probably be a little bit higher, it's true, though the haste is probably more meaningful than the <5 MG procs/minute you can expect with 0 mastery rating. Using the default stats in the sheet berserking catches up to black magic at ~41.5% uptime, and in order for it to be 20dps better you need to be up around 46-47%.

That said, Black Magic may have a nasty interaction with bandit's guile that isn't being modeled. Berserking is up more or less randomly whereas the cooldown on black magic (35 seconds) is very close to the total time it takes to totally cycle bandit's guile. (32ish seconds). Thus (at least for the first few minutes of a sustained fight) you're very likely to have black magic up during the "low damage" phase where it's less useful.

• Aldriana · 10/09/10, 4:46 PM

Your Black Bruise modeling is wrong. The multiplier applies to the armor-mitigated damage, not the raw (unmitigated) physical damage. Basically, you want to add a *1.04*(1-B31) term to the Black Bruise formula, which accounts for a roughly 35% reduction in the potency of the ability relative to the current estimate.

I think you also have the poison damage estimates wrong - it looks like you're applying Ebon Plaguebringer only to the AP scaling, and it should apply to the base damage as well. Either that or I don't understand the formula you have entered, which is also possible.

Lightshadow · 10/09/10, 8:28 PM

You're right on both counts, a new version of the spreadsheet is now available. The change puts H:BB ~75 dps behind Heroic Havoc's Call and a bit above heroic bloodvenom.