Monday, January 12, 2009

Who Was Mr. Clutch in 2008? Not who you Think...

There's much debate about whether clutch ability exists as a repeatable skill or if it's just a function of sample size randomness and fan favoritism. A lot of statheads discount it greatly, if not dismiss it entirely. I'm not going to worry about that, because regardless it definitely makes for good barstool chatter.

Baseball-Reference has a function on it's batting splits page where you can see how players performed in high, medium and low-leverage situations. (Leverage is simply a measure of the relative importance of a game situation--two outs in the bottom of the ninth of a tie game is a much higher leverage situation than leading off the game, for instance.) Just click on the Nationals team batting splits page, scroll down to leverage, and then click on the orange highlighted "High leverage" link and each player's performance in those situations pop up. Or just click here if you just want to skip to the full results.

Here are the Nationals ranked by OPS in high leverage situations in 2008:

PA High Lev. OPS
AHernandez 22 1.12
AGonzalez 13 1.084
RBelliard 69 1.015
NJohnson 30 0.95
JFlores 89 0.835
EDukes 73 0.822
DYoung 44 0.79
AKearns 86 0.784
KCasto 40 0.767
JEstrada 8 0.75
RLangerhans 34 0.744
CGuzman 108 0.655
ABoone 57 0.646
PLoDuca 32 0.629
LMilledge 131 0.593
WHarris 85 0.588
RMackowiak 8 0.583
RZimmerman 86 0.555
POrr 13 0.462
FLopez 81 0.454
RBernadina 17 0.45
WNieves 49 0.436
WPena 45 0.367
EBonifacio 24 0.308
LMontz 8 0.25

Wow, not what I expected. Go AHern! The AG rocking it. And Ron Belliard (who I would have expected to see) tops the list among players with anything you could possibly consider a meaningful sample size (although really none of these samples are meaningful, but I digress--back to the barstool). Certainly the names on the bottom of this list can't surprise anyone. Pena, Lopez, Bonifacio all near the bottom, and poor ol' Luke Montz is the caboose.

OK, but really if you were going to say, "who was the most clutch," it's not really fair to look at it this way, just based on raw performance in clutch situations. That's because you're mixing in basic hitting ability with the ability to elevate performance in the clutch. To find out who's the most clutch, we should look at the difference between a player's overall performance and their performance in clutch situations. That way we can see who's best able to elevate his play with the game on the line.

By subtracting season-long OPS from OPS in clutch situations, we get this list:

PA High Lev. OPS 08 OPS Clutch Diff.
JEstrada 8 0.75 0.37 0.38
AHernandez 22 1.12 0.789 0.331
RBelliard 69 1.015 0.845 0.17
AKearns 86 0.784 0.627 0.157
KCasto 40 0.767 0.61 0.157
AGonzalez 13 1.084 0.938 0.146
JFlores 89 0.835 0.698 0.137
RMackowiak 8 0.583 0.462 0.121
NJohnson 30 0.95 0.846 0.104
PLoDuca 32 0.629 0.581 0.048
DYoung 44 0.79 0.794 -0.004
RLangerhans 34 0.744 0.776 -0.032
ABoone 57 0.646 0.683 -0.037
EDukes 73 0.822 0.864 -0.042
RBernadina 17 0.45 0.544 -0.094
POrr 13 0.462 0.589 -0.127
CGuzman 108 0.655 0.786 -0.131
LMilledge 131 0.593 0.731 -0.138
WPena 45 0.367 0.509 -0.142
FLopez 81 0.454 0.619 -0.165
WHarris 85 0.588 0.761 -0.173
WNieves 49 0.436 0.65 -0.214
RZimmerman 86 0.555 0.774 -0.219
EBonifacio 24 0.308 0.649 -0.341
LMontz 8 0.25 0.593 -0.343

So there you have it. The Washington Nationals' Mr. Clutch 2008: Johnny Estrada. And if it wasn't for the brief appearances by Montz and Bonifacio, Mr. Choker would have been none other than Ryan Zimmerman.

Wow. Who would have guessed that? Not me. Does it matter, and should we expect these performances to continue? No and no. But danged if it doesn't make you go hmmmm... Another round, bartender!


ckstevenson said...

Wouldn't it be worthwhile to weight these by number of plate appearances?

Estrada evidently was really clutch in 8 plate appearances, but Belliard was half as clutch in 8+ times the number of appearances.

Also, by subtracting the regular 08 OPS Estrade gets elevated merely because he sucked so hard/bad/much the non-8 times in the season that weren't high leverage.

An Briosca Mor said...
This comment has been removed by the author.
e poc said...

this kind of makes me wonder whether players are even aware of high v. low leverage situations, even in an intuitive sense. which in turn makes me wonder what "clutch" even means. presumably, the hitter would have to be aware that his at-bat is unusually important in order to access his clutch ability. but does ron belliard know (or intuit) that his at-bat down by 1 with a runner on in the seventh is more important than his at-bat down by 2 with one out and nobody on in the ninth? maybe clutch ability does exist, but most players aren't aware of which situations are actually clutch. who knows. but i'd be curious to learn whether players' senses of the most important moments in a game are actually the most important moments.