Okay, so seriously...we know 250 hands isn't truly enough, but I'm curious to know what info you glean from the last 100 hands. Why the dramatic upturn there? Were the cards running hot or did you make a change in strategy or plug a leak? Remember, just because sample sizes are small and subject to gigantic variance or inaccuracy, there is still some truth hidden in them (kind of like rumors) which will become more evident as the sample size grows. This is where you can start to hypothesize about the trends and see if you can generate the same results over the larger upcoming dataset or where changes may be needed.