Spreadsheet for WMO Data - PvE Class DPS

#0 - June 10, 2009, 3:03 p.m.

http://spreadsheets.google.com/ccc?key=rNQYS221OWlH1bGTNIdY-aw

I've used the stats from the infamous top 50 DPS rankings at wowmeteronline (http://www.wowmeteronline.com/rank/classrank7) to get some statistics and trends showing in DPS amongst classes.

You'll find min/max/average/stdev/skew calculations for every class, per boss. There's a full breakdown of every class's average DPS versus every other class's average DPS per boss. Everything is color coordinated so you can easily differentiate between numbers.

There are two outlier statistics, but, as many people have noted in the thread, please don't get all up in arms over them. They are purely there for show and none of the stats in the spreadsheet are influenced by them. Basically what they do is remove fights from the average. You'll find that only the average drops -- ranking doesn't change amongst classes (however... the distance between averages increases dramatically when you remove the outliers).

The first outlier removes each class's highest DPS and lowest DPS fight. The next outlier removes those fights plus an additional fight from each end, again. In short, instead of averaging 21 fights, you average 19 fights. And then you average 17 fights. But these fights are sorted by DPS, so that the highest and lowest get cut off.

This isn't accurate as far as "total damage" and "real DPS" is concerned. It merely removes "gimmick" fights. Not every class has the highest DPS coming from the same boss, nor the lowest, so it may remove Boss X for one class, and Boss Y for another class when removing the highest.

A better way to look at this without removing outliers is to scroll down to the so-called gimmick fights and compare straight DPS. You'll see the percentage gains of the highest class versus every other class (for instance, Mage on Hodir, or DK/Rogue on Freya Hard).

Enjoy!

#51 - June 10, 2009, 9:21 p.m.

Q u o t e:
It always amuses me a great deal when Ghostcrawler responds with something like, "Thanks for the parse." Now, I understand their side of things: if they released any data, people would scrutinize it and reference it and say, "I'm not where you said I was supposed to be!"

Yep. The other thing they say is "Your data are wrong because they show me doing better than I actually do, implying that I don't desperately need buffs." Sometimes that's a misunderstanding of statistics, sometimes it's borne of frustration somewhere else, and sometimes it is just simply fishing for buffs.

The Ulduar fights are complex and it's hard to deconstruct them all. If you have a fight where the mages and hunters are asked just to manage adds, does that mean you shouldn't include their numbers? If there is a fight with a lot of running around, does that mean you don't include it because it's a gimmick fight? If someone inflates their dps by attacking targets that don't matter (say Ignis constructs before they are molten), do you include that?

Thanks for the parse. :)