The Arbitrarian

Bill Simmons wonders if having what he calls a “cooler” on a team helps that team win. What he means by cooler, I believe, is someone who is called upon to make free throws for a team at the end of games, and who doesn’t miss those late-game free throws. Unfortunately, I don’t really have the data to focus exclusively on late game situations, but if we assume that teams expect their best overall free throw shooters to be their cooler, we can look at whether good team free throw shooting in general, or having an exceptionally good free throw shooter (a Cooler), helps teams win games.

First, using data from 1979-2006, I ran a regression of offensive efficiency (pts/pos) and defensive efficiency (opts/opos) on team wins. As it turns out, every extra point scored per 100 possessions results in about three and a quarter more wins, and every extra point allowed per 100 opponent possessions results in about three and a quarter more losses. Offensive and defensive efficiency account for 93.9% of variation in team wins over this period, which is extremely high.

Now, add to the model team free throw percentage: You get nothing. There is absolutely no significant effect on team success from team free throw percentage when controlling for efficiency. What about best individual player free throw percentage (that is, the ft% of the player who would be the Cooler)? Still nothing. In fact, here is a scatterplot of team max ft% (for players with at least the league median number of attempts) versus team winning percentage:

For the uninitiated, that scatter leads us to believe that there is no relationship at all, even without controlling for efficiency. It’s a blob. Sorry, Coolers.