FanGraphs has its own Clutch statistic, as you probably know. If you didn’t know that before, congratulations, you’ve already learned from this article, and it’s only just beginning. Clutch is available all over this website, and it’s fun to scroll through the all-time leaderboard. At the top you find names like Tony Gwynn, Pete Rose, and Dave Parker. At the bottom, you find names like Sammy Sosa, Jim Thome, and Barry Bonds. I should note that by “all-time leaderboard” I mean “leaderboard since 1974″, but that’s a lot of time to many of us. The baseball that existed before roughly 1974ish was a very different sort of baseball. Earlier-baseball statistics are weird.
The Clutch statistic is based in win expectancy, and a high Clutch score doesn’t necessarily mean the player was awesome in clutch situations, just as a low Clutch score doesn’t necessarily mean the player sucked in clutch situations. Clutch is relative to the player himself; a high Clutch score means a player was more clutch than you would’ve expected that player to be. You get it. You’ve probably gotten it for years.
Just eyeballing the Clutch leaderboard, I got curious about the relationship between Clutch and strikeouts. I’m sure I’m not the first person to investigate this, but I investigated anyway because it wasn’t very hard. I care most about modern baseball, so I gathered all players to have batted at least 1,000 times since 1995. This gave me a pool numbering 861. Given that Clutch is essentially a counting statistic, I plotted Clutch/1000 plate appearances against strikeout rate. I wasn’t sure what I was expecting to find, and this is what I found, in Excel chart form:
Overall, as strikeout rate increases, Clutch performance decreases. Had there been a strong correlation, that would be very interesting. Instead, there is a weak correlation, although there is some correlation. The top 20 Clutch performers have an average 17% strikeout rate, while the bottom 20 Clutch performers have an average 21% strikeout rate. Considered another way, the top 20 strikeout rates have an average Clutch/1000 of -0.71. The bottom 20 strikeout rates have an average Clutch/1000 of 0.12. You can come up with various explanations; perhaps high-strikeout guys are more exploitable in the later innings, by specialist relievers. Contact hitters might see the least variation in their performances against different pitchers, since their whole thing is putting the ball in play somewhere.
The relationship is so weak, though, that it’s hardly worth fretting. It’s just a thing that seems to exist to some degree, and it should rarely if ever be considered for strategic purposes. Bill James argued the other day that contact hitters are undervalued, and maybe that’s true, and maybe this here is a tiny reason why. Maybe contact hitters give you a little more dependability. But maybe it’s hardly significant.
My intention was simply to generate that chart and write it up with a few paragraphs. Then I looked at the source data. Among active players with at least 1,000 career plate appearances, the worst Clutch/1000 belongs to Josh Reddick, at -4.46. Next-worst is Giancarlo Stanton, all the way up at -2.39. Reddick, of course, has barely batted 1,000 times, and Clutch, of course, isn’t very predictive, but what’s happened so far is what’s happened so far, and here’s Reddick’s wRC+ breakdown:
High leverage: 41
Medium leverage: 67
Low leverage: 129
Reddick has barely 100 high-leverage plate appearances, but he’s batted .188 with two home runs. He’s got nearly 600 low-leverage plate appearances, and he’s batted .285 with 28 home runs. Josh Reddick probably is not a woefully unclutch hitter. Josh Reddick, to date, has been a woefully unclutch hitter.
At the other end, among active players, the runner-up for best Clutch/1000 is DeWayne Wise, at 1.95. I’m only guessing that Wise is still active because he doesn’t exactly stand out in my mind. And the leader? None other than one Willie Bloomquist, at 1.96. And Bloomquist is approaching 3,000 career plate appearances, so the sample’s bigger than Wise’s, and Reddick’s. Bloomquist’s wRC+ breakdown:
High leverage: 83
Medium leverage: 90
Low leverage: 66
This is sort of what I was talking about when I said a high Clutch score doesn’t necessarily mean great clutch performance. Bloomquist hasn’t been unbelievable in high-leverage situations; he’s just been relatively terrible in low-leverage situations. When the stakes have been higher, Bloomquist has been closer to passable, leading us to the present, where I’m writing this post, and you’re still reading it.
Again, Willie Bloomquist probably is not a terrifically clutch hitter. Willie Bloomquist, to date, has been a terrifically clutch hitter — baseball’s most clutch hitter, by this particular measure. One argument would be that Willie is able to elevate his game when the pressure is on. The alternate argument would be that Willie has given up when the situation hasn’t mattered. These are bad arguments to try to justify a statistical curiosity. You wonder to what extend this colors the Bloomquist perception; because of his Clutch score, he might “feel” like he’s been a better hitter than he is.
By Win Probability Added, Bloomquist’s biggest-ever career hit came on July 6, 2009. The Royals trailed the Tigers 2-1 in the top of the eighth, and Bloomquist batted against Joel Zumaya with one out and runners on first and second. Bloomquist tripled home a pair of runs to put the Royals in the lead, and the WPA impact was +45%. This is what that looked like:
Oddly, there aren’t as many references to Bloomquist being clutch as you might think there would be out there. Probably because, when we think of clutch, we don’t think of it as being relative to the player’s own performance. Here’s a headline from something:
Bloomquist comes through in clutch three more times
Here’s part of an analysis from 2011:
Bloomquist has a career .317 OBP, which is below average, and a career .336 SLG, which is way below average. At least he does have better numbers in clutch situations, batting .262/.359/.346 (705 OPS) for his career with Runners in Scoring Position, and .283/.329/.377 (705 OPS) in Late and Close Situations, compared to a 653 OPS overall.
People have written a lot about Willie Bloomquist, a lot that’s been needlessly positive, needlessly negative, and everywhere in between, but people have not written a lot about this part of the Willie Bloomquist story. Turns out there is a new Willie Bloomquist angle to take. This gives hope that there might still be a new Jack Morris angle to take. Not that I’m encouraging you to find out.
Clutch performance: it might have some sort of relationship with strikeouts. Josh Reddick has not been clutch. Willie Bloomquist has been very clutch. Even clutch Willie Bloomquist has not actually been good, but I’m not here to be critical of Willie Bloomquist. Not this time, not today. Bloomquist has earned this respite.