Kershaw's Legacy

Clayton Kershaw is the greatest pitcher I have ever seen. Since his debut in 2008, he has dominated Major League hitters, accumulating 79 bWAR and 76 fWAR. He is a sure thing to one day become a first-ballot Hall of Famer. A three-time Cy Young winner (with seven consecutive top-five finishes) and 2014 NL MVP, Kershaw has already passed the 200-win mark and is rapidly approaching his three-thousandth strikeout. We sometimes forget, bogged down by our focus on the Now, when greatness is happening right before our eyes, so go ahead and take a few moments to read through his Baseball-Reference page. Or Trout's.

But Kershaw has a blot on his otherwise otherworldly resume: his reputation of poor postseason performance. Aided by his regular-season majesty (and the pockets of a franchise worth an estimated ~5.5 billion US$), the Dodgers have reached the playoffs 13 times in Kershaw's career, including an ongoing streak of 11 years in a row. But for Kershaw, the playoff results have been underwhelming. With a win-loss record of 13-13, his career postseason ERA is 4.49, almost double his (mind-boggling) career regular season ERA of 2.50. Due to some uncharacteristically bad performances on baseball's biggest stage (and, more importantly, only one World Series title during a period when the Dodgers have been historically dominant), Kershaw has earned the reputation of a playoff choker.

This article's goal will be to examine that claim. Full disclosure, I did a study similar to this for my junior year (of high school) statistics final project, but I did it based on ERA and for a couple different reasons I'm pretty sure my methodology was flawed. (My stats professor, by the way, either did not notice or did not care. I got like a 96).

So here goes. We're going to use the principles of Fielding Independent Pitching (FIP) for this project. FIP is a stat that attempts to evaluate pitchers based on only the things that are under their control. It ignores all balls put in play and only considers walks (BB), strikeouts (K), and home runs (HR). The idea is that pitchers can't control their fielders (hence, pitching independent of fielding). The problem FIP is trying to solve is that two different pitchers could give up identical fly balls into the left-center gap and get vastly different results, ranging from a flyout to an extra-base hit, and that those results aren't based on anything the pitcher can control, but the skill of his respective fielders, their respective positioning, the wind, the weather, or ballpark itself.

FIP, again, focuses on three things that are completely in the pitcher's control, unaffected by any defense: strikeouts, walks, and home runs. Here are Kershaw's career stats in those areas, both regular and postseason:

	HR%	BB%	K%
MLB average, 2008-24	2.89%	8.35%	20.88%
Kershaw, career regular season	2.10%	6.27%	27.44%
Kershaw, career postseason	3.81%	6.47%	27.03%

Of course, the HR rate instantly jumps out. While Kershaw's walk and strikeout numbers are nearly identical no matter when he's pitching, he's given up home runs at nearly twice the frequency when he's pitching in the playoffs-- a massive number, considering how important home runs (and their prevention) are.

But is this the product of random chance? Is Kershaw the victim of too small a sample size? A statistical z-test can illustrate that for us. The question we're answering here is basically: What are the odds (p-value) that a pitcher that gives up 2.1% HRs actually gave up 30 bombs in 788 batters faced (the 3.81% rate from his postseason numbers)? What are the odds Playoff Kershaw was pitching just as well as ever, but just got unlucky?

If we plug that into a z-test, we get a p-value of 0.000416, i.e. if Kershaw in the postseason was truly a 2.1% HR rate pitcher, there would have been a 0.0004% (1 in 2500) chance of him getting these results by chance. Essentially, the answer is No. Kershaw definitely was more HR-prone in his playoff appearances.

Calculator courtesy of purdue.edu

Remember, Kershaw has pitched in the playoffs a lot. In his career, he has 39 postseason appearances, including 32 starts. That's a full season! Kershaw has pitched 194.1 career innings in the playoffs. That's more than his teammate Tyler Glasnow has ever compiled in any regular season.

If we limit our tests to only Kershaw's prime (arbitrarily, I chose 2009-2017), we get the exact same results. In the postseason, Kershaw's K and BB rates stay consistent, but his HR rate balloons.

	HR%	BB%	K%
Kershaw, 2009-17 regular seasons	1.64%	6.38%	28.32%
Kershaw, 2009-17 postseasons	3.70%	7.19%	28.34%

Interestingly, it's Kershaw's HR rate that has climbed as he has aged. During his prime years it was consistently below 2.0% (a ridiculous number), but as Kershaw has pitched deeper into his thirties (and dealt with continuous arm issues) his HR rate has fallen back to mortal levels:

As a final redeeming note, these results were not as bad as I expected. Recall that both Kershaw's BB% and K% rate were perfectly normal in the postseason. And although HRs are the most important of the three true outcomes, they are also the most subject to variance. Since the margins are so small, just a few home runs in either direction can skew the results, but they also skew game results. It's true that playoff games are harder, and the lineups he faced were better. But too often, during the most important games of the season (and of his career), Kershaw failed to live up to his otherworldly standard.

And that brings us back to the realization of how good Kershaw actually has been. Kershaw's postseason ERA is 4.49, and his postseason FIP is 3.855, still better than the average pitcher. The fact that we consider that level of performance a disappointment is a testament to how impressive Kershaw has been for so very long. He's an exemplary professional, an apparent man of conviction, and, soon, a legend in his own right. I'm glad he got a ring.

Just one, though. That's plenty. Game 6 of the 2016 NLCS (when he lost to the Cubs' Kyle Hendricks) is still one of my favorite baseball memories.