Issue with double standards with Se and defining between dps and support.

[Judge_Xero]

Pretty sure the number of MCH is way below the number of BRD, creating a statistical bias. I've checked DPS board for MCH/BRD last month and they were something like 5 BRD for 1 MCH on my server ..

I mean even if you take the top 400, you may not have the good sample compared to BRD since 400/500 or 400/2500 (just to illustrate my word, I don't know exactly the ratio) will not give you the skilled players with high DPS for MCH ^^

Yeah it's a tough call on how to accurately compare that data. Look at the case of SMN on A2S and you will notice that they dominate the first 400 data points which significantly pads their overall average.

I imagine you would want to compare 75th percentiles across the board to eliminate the bias caused by the # of overall players playing each class and use the 75th percentile to get rid of the really high (orchestrated parses) and really low values.

While it appears to be a lot of work done as well, I'm not sure about the credibility of it. Just a quick look at A1S NIN VS BRD there seems to be something off as BRD NA is averaged much lower than NIN despite having higher averages, not to mention that out of 400 data points only 100 are below 1133 DPS to a floor of 1087, so it's pretty much impossible for their average to be that low.

(maybe this data was collected a while ago?)