Need help on the weird dps meter

[adilina]

Hi guys, I am looking at the log of my guild's heroic Ultraxion overall wiped attempts and I found something quite puzzling.

First off, here is the log: http://www.worldoflogs.com/reports/r...one/?enc=wipes

We run three shadow priests, with the two lower ones sub'ed in/out mid-fight.

Here are their armory links:

(L) http://us.battle.net/wow/en/characte...odead/advanced (minus the 410 chest, he had 391 FL chest on boss kill)
(S) http://us.battle.net/wow/en/characte...%B9st/advanced
(B) http://us.battle.net/wow/en/characte...essid/advanced

As you can see, L has better gear, S and B are on par but S has the legendary staff. When all spriest are in the raid, the dps is: L>>S>=B. With L and S receiving DI most of the time.

What puzzles me is their Mind Blast damage. Here is a chart for easy comparsion:

Hit/Crit Average over 35 tries
L 45987 / 94179
S 42660 / 92603
B 47396 / 98468

So, L has around ~12K spell power fully buffed, S has 11.2K. I am not sure about B, but S should have more spell power (407 more net int unbuffed) than B after 10 stacks from <Will of Unbinding>

L has no 4xT12 bonus, S and B both have that. S have legendary staff but it does not matter. Then why would S have the lowest MB damage? It looks like B received his 4T12 bonus but S did not.

Here are some more data on their other major damage spells.

MF ~~~~VT ~~~ SW:P
L ~~~10741 / 22579 ~~~ 11479 / 24060 ~~~ 5800 / 12180
S ~~~ 9522 / 20066 ~~~ 10183 / 21570 ~~~ 5545 / 11542
B ~~~ 9572/20050 ~~~ 10325 / 21766 ~~~ 5592 / 11733

It is simply strange to see S doing less of all spells given his gear was slightly superior to B.

Lastly, a comparison of their basic stats from Armory:

Int ~~~ Haste ~~~ Mastery ~~~ Crit
L(Goblin) ~~~ 7073 (+880) ~~~ 22.19%~~~ 16.84~~~ 14.12%
S(BE) ~~~ 6559 (+780) ~~~ 24.12% ~~~ 11.80 ~~~ 14.47%
B(Goblin) ~~~ 6932 ~~~ 20.99% ~~~ 11.11 ~~~ 18.07%

Your input is appreciated.

---------- Post added 2011-12-30 at 04:24 PM ----------

S joined the raid right after a server transfer. Is there a transfer sickness?

[FieryBull]

Might have something to do with volcanic potions. B uses them alot more often than L who uses them more often than S.

[sillydkarchie]

So I'm kinda lazy so I just looked at one attempt where all three were part of it, since every attempt is different you shouldn't lump them all together and "average" their stats (one priest might've had way more fading lights some attempts, other form of RNG fffff'd (shadow orbs, lol) ect)..

Anyways, attempt #11:

They all have pretty good dot uptimes, with L's being the best.
S's MB usage is, by far, the best of the three. (42 to 35, 34)
L has DI, S and B don't (this is huge, btw, don't know if you realize that).

L should be winning, with his large ilvl gap AND DI. However, looking at this attempt, if S had gotten the DI, he probably would've beaten him (DTR is pretty good, but won't make up for ~8lvl's + lack of DI)

Your thing about B and S being close on damage, though.. they aren't close, S beat him by a sizeable amount on this pull.

Also, I think you're reading too much into their mind blast dmg. The 4-piece got nerfed, and RNG is RNG. You should more be analyzing their dot uptimes, mind flay uptime, mind blast USAGE (1 every 9 seconds is pretty good), who's helping out healers with pw:s and dispersion for a fight like ultraxion HM, empowered shadow uptime, ect.. I'd probably bring L and S, once S gets equal gear to L, DI should go to him.

[adilina]

Might have something to do with volcanic potions. B uses them alot more often than L who uses them more often than S.

The logged potion usage seems way less than actual. I would not go with that.

---------- Post added 2011-12-31 at 02:58 PM ----------

Thanks for the above analysis.

DI did not seem to help S as much as it helped L. There is no discernible dps increase for S after he received the DI (it was given to B originally), which is strange as well. Comparing the proc up time, L had more than S, probably because L received the DI from an Afflication lock, S did from a Destruction lock.

I think 25 attempts is a decent sample size to compare these priests' performance. It seems S had lowest average MB damage because he casted MB the most often. However, since he has the highest haste of all, he should have produced shadow orbs the most often, which is the case. S's data still lacks a reasonable explanation as to why he seemed doing sub-par damage overall given he was executing things well enough, if not the best of all.

[Garouken]

Hit/Crit Average over 35 tries
L 45987 / 94179
S 42660 / 92603
B 47396 / 98468

Answer is very simple.... L and S try to use MB on cooldown, resulting in more MB0 MB1 MB2. While B fails and doesn't cast it on CD resulting in more MB3's so his average MB damage is higher.

If anything looking at those logs a bit, S is your best shadow priest, given a little more gear and DI he'll easily be beating L. He has better MB usage and good dot uptimes.

Easy way to tell, go to some parses and look at damage break downs. It should be Mind Flay > Mind Blast > Vamp Touch > Etc

Your L priest always has Mind Flay > Vamp Touch > MB, he ussually has 10-15 less mind blasts per fight than S, which is also why S has the lowest average MB damage. (Using MB on CD is always better, you never wait on it disregarding Shadow Orbs)

[Shaiandra]

Another thing you have to take into account is Mastery. L has 17.01 Mastery, so her (going by character gender) DOTs are increased by ~24.66%, and that means Mind Blast can get anywhere between a 24.66% and ~73.99% increase, depending on the number of orbs. S, with only 11.08 Mastery gets only a 16.07% bonus to DOTs and at most 48.2% buff to Mind Blast with 3 orbs, and once B's Necromantic Focus is stacked, he'd have 13.56 Mastery (~19.66% DOT damage and up to +58.99% Mind Blast).

While S does have the highest Haste, L isn't too far behind if you include the goblin racial 1%. And of course the 4T13 bonus generates a lot of Shadow Orbs, so L would probably have the most 3 Orb Blasts.

[adilina]

Thanks for the incisive analysis. The above posts make much more sense. One thing to mention is that while B should definitely cast more MB, L did it less probably because he didn't have 4T12, and S probably had a few MB duplicates from the legendary staff (I personally think it has very low occurrence rate). The staff proc only register in log when a dot tick is duplicated. When MB or MS is duplicated, the log uses the ability name to register.

[Strykie]

With a name like stardust im assuming its a girl because they wanted a pretty character instead of a troll or goblin which provide the best dps increase for shadowpriests.
It would also explain how they got a legendary.......................

Looking at the first 3 ~6 min attempts dot is behind and MB ahead for stardust, meaning they are having a problem maintaining empowered shadows, using dots with potions or lining up other damage boosts, because their uptimes are decent enough.

If your blessid priest did not have DI till stardust was out, his dps is pretty good, could be a little better ~1k better.
Stardust having DI and Staff doing 33-34k i want to throw my head through the monitor.

[adilina]

Looking at the first 3 ~6 min attempts dot is behind and MB ahead for stardust, meaning they are having a problem maintaining empowered shadows, using dots with potions or lining up other damage boosts, because their uptimes are decent enough.

Parse didn't show any of them having problem maintaining ES. B may be a little behind on keeping the dots up but there is no major difference. The interesting part comes from the lining up with potion/enchant proc part. It seems neither L nor B bothered to refresh dots right after these buffs or before they expire, they just maintained their normal rotation, while S made some efforts in dot refreshing with the buffs. Quite interestingly S therefore lost a few ticks on his MF ticks to L, who had less haste than S. Considering MF is the most damaging spell for spriest on single target fight, leading the runner up by about 10% on the damage chart, it would make sense to use the buff time on MF rather than reapplying dots.

I understand the spriest guides says refresh dots with temporary buffs but as the gear level increases, the % spell power increase from these buffs are diminishing, maybe to a point where dps loss due to casting and GCD outweighs just ignoring dot refreshing for single target fight. Hopefully someone with good theroycrafting skills can comment on it.