Help find “clutch” players, if they really exist.
Perhaps no subject in baseball is more hotly contested than the concept of clutch players. If you want to start a fight with a member of the SABR statistics committee, insist that clutch players exist. Tell him they absolutely exist, beyond any doubt. Tell him Alex Rodriguez is obviously not a clutch player, and David Ortiz is a clutch player. If you want to start a fight with a traditional baseball fan, tell him there is no such thing as a clutch player. Tell him there is absolutely no statistical evidence that any player has ever performed better in “clutch” situations.
Making the debate even more contentious, Bill James recently switched sides. He used to be a clutch atheist (“there are no clutch players”), but recently he became a clutch agnostic (“we can’t measure whether there are any clutch players”). (If you want to read the article, see http://www.sabr.org/cmsfiles/underestimating.pdf.)
In this hack, I’ll show you some quick ways to answer this question yourself using play-by-play data. Let’s break down this problem into four parts: defining what makes a player a clutch player, measuring player performance in clutch situations, comparing player performance in clutch situations, and understanding the results.
The dictionary definition of clutch is “tending to be successful in tense or critical situations” (well, the relevant meaning). Here are a few examples of clutch situations: ...