In Sunday’s New York Times, David Leonhardt reports on an article my Bill James that has caused a mild stir in the sabermetric community. For years, the stathead chorus has espoused the view that clutch hitting does not exist. In the article for Baseball Research Journal, Underestimating the Fog (which I have not read), James argues that just because we have been unable to identify “clutchness” using statistical methods does not mean it’s not there. Leonhardt spells out James’s argument
In baseball, luck and randomness – weather, ballpark dimensions, the pitcher – play the role of the winter coat. And the search for clutch hitters involves not just one comparison that compounds the statistical noise. It has two: the differential between a player’s normal and clutch batting averages and the difference between this differential across seasons.
I’m not sure what to make of this assertion, because, well, it’s obvious. In a theory of science you can’t prove a null hypothesis (e.g., clutch hitting doesn’t exist). But the fact that people have searched for so long and generally found little — there are a few exceptions — certainly gives me more reason to be pessimistic than optimistic about the existence of clutch hitting. I understand that there is much noise in statistics, but I still think the burden of proof is on those who believe in clutch skill, and here’s why.
I think that clutch hitting does not exist for a reason that has nothing to do with numbers or any type of statistical test. Let me pretend to be a baseball player for a moment. Consider two situations:
- It’s the third inning of a game that my team is winning by three runs with no runners on base.
- It’s the bottom of the ninth with my team down by a run with runners on second and third, two outs.
The first situation is certainly not a clutch situation, while the second one is. (Please, let’s put aside the debate of what “clutch” means for a second; clearly, this situation qualifies.) Why should I expect any player to exhibit any type of different behavior in these two situations? I don’t think any player would approach these situations differently at all. Hitting, is not an endurance sport. Players stand up and do it 5 times a game. I think they put forth the exact same amount of effort no matter what the situation, and it seems silly to me that players would exhibit some level below the maximum at any time (which is what clutch hitting theory requires) . Every at-bat appears in the box score equally and is used to calculate stats that will determine the salary a player will receive. So, unless there is some reason for players to preserve some hitting effort until crucial times, and it does not seem that there is such a reason, I think batters put forth 100% effort 100% of the time. There is no incentive for a player to ever hold back, therefore there is no room for clutch ability to exist.
But what about the nerves factor. The first situation has lower stakes than the second and maybe this can cause some players to be unclutch or chokers. In this sense, clutch ability is really un-unclutchness. I’ll grant that maybe there is some nerves factor, even aided by some physical characteristics such as adrenaline production, but I just can’t see major league ballplayers differing in this area. By the time even the worst major league hitter reaches the big show, he has endured so many nerve-racking situations that the lining of his stomach must be nearly gone. It’s a requirement to get into the show. I suspect young players may be slightly more bothered by it than veterans, but I don’t think among veterans there exist classes of clutch and unclutch players due to nerves. They’ve all been there. It’s their job.
Now pitching, on the other hand, has some room for clutch ability. Pitchers clearly do vary effort from batter to batter based on the players involved and the game situation. I do think some pitchers throw 90% most of the time to save gas for the times when some extra gas is needed. While we have yet to identify clutch pitching yet, I have tried and been unable to observe it, I certainly think it is plausible for it to exist, unlike hitting.
So while I hope the quest for clutch hitting skill continues, I don’t expect there to be anything there. And as sabermetricians continue their non-findings I hope that the burden of proof remains solely on proving clutch hitting exists.