Don't blame MEH for his performance Monday night

[Megatron96]

Originally Posted by jd1020

If only you could take out someones best run like it doesnt exist and act like sitting out an entire game in which both of them are likely to fail, preserving ones YPC, is evidence of anything.

That's actually the proper way to analyze any statistical data group. Or at least it was when I took statistics in HS and college.

One of the first things you do when looking at data is to remove the largest/highest and the lowest values (they are considered outliers), then average out the rest. That eliminates most of the 'skewing' effect that one or two incongruous pieces of data can create (can't remember the proper term for that right now).