Nice article – I think one of the next big breakthroughs in poker stats analysis will be analyzing the correlations between different traits, and thereby allowing you to infer much more from small pieces of information. For instance, correlation between cbet% early in the tournament when deep-stacked and steal% from the CO when 10 bbs deep late in the tourney. The same analysis would be useful in cash games, but perhaps a little bit less so because 1) it’s easier to get a bunch of hands, so making the most out of a little information is relatively less useful, and 2) there are fewer situations in cash games, since the structure never changes, so you have less “variables” to estimate, so fewer hands gets you more direct information about the current situation you face compared to a tourney.

-bruechips