Is it possible to use thin slicing (as detailed in Gladwell's Blink) to make better bets about the outcome of NBA basketball games? The most important factors would appear to be FG%, turnover rate, offensive rebounding rate, and free throw attempts. (via truehoop)