Measuring improvement

[JeffSpicoli]

Does anyone have a good method for judging before/after results? I've got the equivalent of 2000 hours of play with one strategy...and 2000 with another. My hourly results have improved with the new strategy but I want to use some sort of statistical tool to figure out if it's for real or just randomness coming into play.

I've got session data in a detailed spreadsheet so I can run just about any stat.

Jeff

[Bacchus]

Tell us what variables you've entered into the spreadsheet--although correlations are the best you are going to get. Although you still won't know if you have neglected to include variables that are potentially important.