Matt Holliday, Colby Rasmus, Clutch, Skill and Performance
I’ve seen a few things around twitter the last couple days that I found warranted a little discussion. Most of these topics revolve around the idea of clutch hitting or hitting with RISP (I’ll use the two interchangeably although there are slight differences). The general line of conversation goes something along the lines of:
Old school guy: Holliday (or Rasmus) is a horrible clutch hitter. Look at his numbers with RISP (insert one year sample here)
Saberist nerd: Shut up old man. Clutch is a myth, it doesn’t exist. Holliday has a 0.385 wOBA and Rasmus’ is 0.360 they’re the 2nd and 3rd best hitters on the team
Old school guy: Having a good wOBA (whatever that is) doesn’t matter if they aren’t helping to produce runs
I have a couple of comments on this general line of discussion. First when the saberist speaks of clutch not existing he does not mean that clutch performance doesn’t exist. Clearly Holliday has been sub-par with RISP this year. It happened and is factual (luck factors in here, but you get my drift). The saberist is talking about clutch skill when he says clutch is a myth. By skill I’m referring to something that is repeatable and predictable. The statistical evidence points to this skill being either non-existent or too small to measure (read too small to care about). The basic experiment in The Book concluded that the player’s generic wOBA (i.e. generic hitting skill) is more predictive of future RISP performance than a player’s past RISP performance is of future RISP performance, so basically good hitters project to hit well in the clutch and bad hitters project to hit poorly. Holliday and Rasmus both qualify as good hitters, and are therefore pretty likely to be good in the clutch going forward.