Friday, May 9, 2008

Moneyball to the NBA?

Winston-Salem, NC -- I ran into this fascinating article in the New York Times this week about the front offices of NBA teams hiring stat geeks to build their rosters. My question is: what took so long? The NBA and NFL are both number two behind baseball in terms of being statistics-driven sports. Just like in Moneyball with the Oakland A's expelling the use of the sacrifice bunt and stealing bases in their playbook, Mike Zarren, the Celtics' stats guru, calculated a transposed statistic: the 3-pointer in the corner. If the A's reduce the percentage of making an out (sacrifice bunts, % of caught stealing, etc.), and thereby increasing their on-base percentage, the percentages are in their favor that they'll score a run. Consequently, in basketball, aside from the layup, the most efficient shot is a 3-pointer from the corner. Here, rather than any other point behind the arc, is where the highest percentage of baskets are made. So, be sure to have a 3-point specialist on your team, a la Ray Allen.

Conversely, the A's roster is compiled of thrown away parts who are willing to sacrifice individual statistics for the good of the team. Their job is to get on base -- period. On defense, their role is create outs -- even if the other team is attempting to move runners forward to a base closer to home. Take the sure out over the stellar double play. As Zarren puts it in basketball terms in the article: "What’s one of the most misused, misinterpreted statistics? 'Turnovers are way more expensive than people think,' Zarren says. That’s because most teams focus on the points a defense scores from the turnover but don’t correctly value the offense’s opportunity cost — that is, the points it might have scored had the turnover not occurred.'" Outs are more expensive than moving a runner from first base to second base.

As far as the other statistics that the Celtics and Zarren uses to evaluate players, they are sealed and probably going to the grave with Zarren unless he sells his soul. Within 5 years, we'll see some stat heads manning the front offices of the NFL.

No comments: