October 11, 2016

Bayes-Running in Baseball

Is 4/10 greater than 300/1000? It may be a silly question to your Calculus teacher, but not so much to a sports fan when the numbers are associated to the players. Is a .400 hitter really better than a .300 hitter in baseball? How do we estimate the proportion of success when there is a lack of evidence? In this talk, I will show you one way to attack this problem using empirical Bayes estimation. The stage will be set in the baseball world but no coding or baseball knowledge is required.