[dougbright]

The author raises good points which are hard to disagree with. I'm troubled by the decision to choose the odd and seemingly arbitrary "95-20 ratio" though, instead of a more typical 80-20.

Does anybody know what these graphs look like at 80-20?

[s17n]

Wouldn't gini be better than either of those?

[arcanus]

Or a straight up power-law curve fit, which is what those are both (implicitly) based upon.

[s17n]

Really? I thought that Gini is equally applicable to any type of underlying curve.