James H Albert
Abstract: Baseball fans are fascinated with splits, where player offensive or defensive statistics are broken down by different situations, such as home/away, date, same side/opposite arm pitchers, etc. We provide a general overview of the use of random effects models to fit these data. It is well-known that hitters have different talents, but it is not clear that hitters have talents to take advantage of particular situations. For example, fans like to think that specific players have clutch ability, but there is little statistical evidence to indicate that clutch ability exists.