PROS ARE NOT GETTING BETTER AT FREE THROWS
Rick Larrick recently told Decision Science News that baseball players have been getting better over the years in a couple ways.
First, home runs and strikeouts have increased. The careless or clueless reader might note that this is curious, for from the batter’s perspective home runs are a good thing and strikeouts are a bad thing. What’s going on? Batters may be swinging harder, increasing the chance of both. The purported improvement is a result of the benefit of a home run being greater than the cost of a strikeout. After all, a home run results in at least one run, often more, and runs are a big deal since the typical team earns only about 5 of them per game.
DSN wondered how the players learned to swing harder from one decade to the next. Was it based on feedback from coaches? Or from fans / media attention?
According to Larrick, the number of attempted stolen bases has decreased over the years. Apparently, it is only worth it to steal if one can pull a very high percentage of the time, higher than had been believed in previous years (anyone know the stat?). So while crowds (presumably) like the action of stolen bases, players do not respond by doing it more. Winning seems more important than pleasing the crowd, which is a strike against the fan-feedback hypothesis.
After our post on winning back-to-back baseball games, some folks like our friend Russ Smith made comparisons to the hot hand effect. There is something to it. However, in the baseball example one starts with a prior of .5 (since one doesn’t even know which two teams are playing), while in basketball the chance a pro will make a free throw is about .75 (since one can condition on the player being a pro). What is surprising is that in both cases, the past success tells you next to nothing.
This conversation lead your Editor to find this NY Times article which shows that, surprisingly, pro basketball players are not getting better at free throws over the years.
So, the question to the readers is: Why do some athletic abilities improve as history marches on (e.g., running speeds, batting, base-stealing) and others do not (e.g., free throws)?
P.S. For the record, Decision Science News is not becoming a sports blog. It is just a phase the Web site is going through. That said, there has been interest in seeing this kind of result in other sports, so that analysis will be coming in future posts, in glorious, glorious R and ggplot2. (Don’t know R yet? Learn by watching: R Video Tutorial 1, R Video Tutorial 2)
Photo credit: http://www.flickr.com/photos/cakecrumb/4398699952/. A cupcake was chosen because Jeff gave us empirical evidence that people like cupcakes much more than a control food.