One of the problems with using something like Chauvenet's criterion (or anything else for that matter) is that in this game anyway, a lot of things that would definitely qualify as outliers probably weren't.
What if you're a THF and 99% of the time you use SA+TA+WS but one time you don't. Or a WAR and you do a Mighty Strikes rampage? For Chauvenet's criterion to work the data has to fall into a normal distribution, and since the results have too many external (i.e. player controlled) factors, I don't think the data does. I've wanted to parse relic 3x damage proc as well, but this would be hard for the same reason.
What if you're a THF and 99% of the time you use SA+TA+WS but one time you don't. Or a WAR and you do a Mighty Strikes rampage? For Chauvenet's criterion to work the data has to fall into a normal distribution, and since the results have too many external (i.e. player controlled) factors, I don't think the data does. I've wanted to parse relic 3x damage proc as well, but this would be hard for the same reason.