[WhiteWolf]

I can't comment on the analysis, but it seems to me (especially with a small sample size) that it would be impossible to seperate cause from effect here. Much of the correlation can come from just running good. If I'm catching good cards, I'm going to be more aggressive than normal with my betting, and I will also be winning more than normal. As a result, the statistical model will show the aggression/money won correlation without accounting for the fact that both of those values are a result of merely catching a good run of cards.