1. Need to add test dataset
x. Need to make another abstraction to run algs multiple times
3. Find Max for how far to prune back
4. Run weka on data set looking for a soft-margin
5. cluster on similar scores
-- Then look at move clusters
6. cluster based on moves
-- advantage for color based on moves less than x?
7. map moves:scores, then cluster on score, what do we see w.r.t moves
8. look at classification for each iteration of moves
-- on first turn, second, nth, etc
-- (see 2 and use prune funct)
x. look at mean score of the game as a predictor!
10. weighted move scores.