clock menu more-arrow no yes mobile

Filed under:

Is Clutchness Correlated With Any Offensive Traits?

It's long been a point of contention—are patient hitters like Jack Cust or Daric Barton less productive than their free-swinging brethren when it counts more? And to zoom out and get more general, are there any offensive traits that correlate with being clutch?

I decided to calculate correlations between a whole set of assorted offensive stats and clutchness, just to find out. To measure clutchness, I used WPA's clutch score, which compares a batter's context-dependent offensive production to his context-independent production, to see if he performed better when it meant more. Click to enlarge.


  • Note that these correlation coefficients are miniscule. (1 is perfect correlation and 0 is completely uncorrelated.) Even the stats that do correlate with clutch are weak fits at best. This makes sense—clutch doesn't even correlate with itself from year to year, so why would other stats fare any better? That said, I ran the numbers for every qualified batter from 2006 to 2010, and these seven stats did seem to weakly correlate with clutch.
  • Some interesting conclusions? Batters who hit more groundballs tend to be slightly more clutch, and the reverse is true for flyball hitters. Also, batters who strikeout more tend to be less clutch. These two factors most likely explain why isolated power is negatively correlated with clutch—power hitters generally strike out a lot and hit flyballs.
  • I ran essentially every stat that Fangraphs carries, but these were the only ones that showed any vaguely consistent correlation. Some notable stats that didn't correlate with being clutch? Swing percentage, walk percentage, batting average, on-base percentage, and first-strike percentage. That first one is very interesting—patient hitters don't seem to be any more clutch than free swingers.
  • The formula for WPA clutch score is WPA/pLI - WPA/LI. More here and here.
  • But the bottom line? It's important to note that while these correlations may exist, they're tiny factors compared to context-independent production. In other words, the effect that these correlations may have are dwarfed by the effect of a slightly more talented hitter, clutch or not.

The Boston Red Sox are in town, starting a two-game set. John "Sling Blade" Lackey faces Brett Anderson at 7:05 PM.