I was researching a small essay that I was going to write regarding the efficiency of Chris Bosh, Tim Duncan, Chauncey Billups, Steve Nash, and Tony Parker and found an interesting article which concisely summarises the recent work in basketball statistics that we (coaches) are attempting to incorporate into the game.
Just as the technical analyst would argue that a stock’s price reflects all possible information, the statistician would claim that a player’s contribution to the team’s winning percentage can be entirely distilled into numbers. I disagree — stats can be tremendously helpful when used intelligently but must eventually take a backseat to the coach’s discretion. Technical analysis doesn’t always foretell events such as innovations or federal indictments (although these stocks are usually overbought when they turn south) and using statistics to identify truly effective players is much more difficult than it seems.
I’ve always found that statistics can never properly analyse how players contribute to a team’s personality. If Team X needs to play at a high level of intensity and Player Y drives intensity but jacks up questionable shots frequently, what should a coach do? Should a coach play the intense player and write off the errors as part of the game or should the burn go to a player with a higher efficiency rating?
Billy Beane possesses an excellent payroll-to-wins ratio since 1999 but has been unable to win any post-season series. The Los Angeles Angels, Florida Marlins, and Chicago White Sox have all won the World Series playing small ball and getting “hot” at the right time (although the statistician would claim the hot hand doesn’t exist and the Chicago’s performance could be attributed to other factors, such as Phil Garner). I think that the coach should play Player Y, keeping them on a short leash and focusing on the core team statistics that contribute to winning.