Clayton Kershaw Chases Zack Greinke, Nerd History

Chasing records is fun. Sometimes it’s meaningless and fun, like the famed Ryan Webb, Matt Albers games-finished-without-a-save streak. Some chases have limited meaning, but they’re fun, like Zack Greinke‘s pursuit of Orel Hershiser’s consecutive scoreless innings streak. Other are meaningful, like the 2015 Cardinals pursuit of the best park-adjusted ERA since before World War I.

Record chases exist on a spectrum of notoriety — from “Mark Simon would tweet about this,” to “my mother-in-law has heard of it and wants to discuss it.” That’s a big range. Some records are fun quirks that could happen to anyone. Some are the product of a truly great player performing at a high level. The record up for discussion these days falls into the latter category, but it’s also closer to Mark Simon tweet status than mother-in-law conversation fodder. Clayton Kershaw is trying to post the best single-season, park-adjusted xFIP ever.

Immediately, this record sounds impressive. We have xFIP dating back to 2002, and a lot of good pitchers have pitched since then. Posting a low xFIP is hard and is reflective of an excellent pitcher season. Leaving aside any potential objections to the statistic and its place in player evaluation, we can all agree posting the best single-season xFIP ever would be an impressive feat, even if you’re someone who prefers a pitching statistic that has a little more actual runs allowed flavor mixed in.

This isn’t Kershaw’s first dance with the single-season xFIP record. In fact, last season, Kershaw put together a similar effort, falling short of Curt Schilling’s record.

Kershaw needs to lower his xFIP to 1.944 to fall below Schilling’s park-adjusted mark from 2002 (i.e, to show up as 51). If we assume he will throw roughly 60 more innings, that’s an xFIP of 1.47 in those starts. Doable, but very tricky. The best way to get there would be some sort of serious fly-ball prevention — since you’re probably not going to do much better than Kershaw’s strikeout and walk rate. After all, in his last nine starts, he has a 0.91 ERA, 1.34 FIP and 2.00 xFIP.

Of course, this is sort of a head fake. The xFIP record is only as interesting as you find xFIP. The stat exists in a weird middle ground between people who only care about ERA and people who think xFIP is overly simplistic. The point, really, is Kershaw is arguably having one of his best seasons, and he’s setting himself up for another Cy Young Award.

From a who’s actually going to win standpoint, it may be hard to beat Greinke’s sub-2.00 ERA. Even as wins have slowly lost clout in the awards races, ERA — as flawed as it may be — still plays a huge role. Greinke currently owns an earned run average that’s 0.76 runs below Kershaw’s. And that’s without thinking about Max Scherzer or Jacob deGrom and their excellent seasons. It’s too hard to say what the electorate will look like: some BBWAA members would be very open to using FIP, xFIP, cFIP, DRA or their components, in their voting, while others would never dream of it. The actual interplay remains to be seen, but as an academic exercise, we can make the case for Kershaw pretty easily.

It starts with FIP and xFIP and Kershaw’s historic pace regarding those components of pitcher performance. We know that pitching is not just about those two fielding-independent numbers, and we know that regressing home runs allowed in the way that xFIP does isn’t necessarily even the right way to measure past performance. But those numbers do a nice job reflecting that subset of the job. Based on the things xFIP cares about, this is arguably the second-best season of the past 14 years.

If you crave a more thorough analysis of the batted ball, Kershaw is posting a .284 BABIP, which is actually 12 points worse than his career average. This is the first time since Kershaw’s rookie year that he hasn’t had an ERA well below his xFIP. And his HR/FB% (11.6%) is above average for the first time since his rookie season despite a track record of being below 8%. Kershaw has a 2.34 ERA even though his batted-ball results are much worse than what we’ve seen from him during his career.

There are two ways of looking at that. The first calls on you to believe Kershaw’s batted-ball skill is going away — or at least that it has gone away for this season. That’s possible, of course, but given what we know about BABIP, HR/FB% and sample size, you should at least be a little skeptical. Now if you want to take the position that those hits happened and Kershaw should get all the blame, Greinke is going to be your guy. Kershaw trails Greinke significantly in RA9-WAR. But Kershaw has a small lead in FIP-based WAR. The other way to look at this is Kershaw is simply being betrayed by factors outside his control.

No one will legitimately argue that Kershaw hasn’t had a very good season, but there will be plenty of disagreement about who is having the better year among the two Dodger aces. Greinke has the ERA, Kershaw has the fielding-independent numbers. We’re left at a classic pitcher evaluation impasse.

To help litigate, I called upon actual attorney and Baseball Prospectus author Jonathan Judge to shed some light on the differences in contexts both pitchers have faced in 2015. BP houses Deserved Run Average (DRA), which utilizes mixed models to control for lots of contextual factors that our versions of WAR don’t. For our purposes, I’ll just point out that Kershaw has a DRA- of 49 and Greinke’s is 44. The two big factors that shrink this particular ERA gap en route to DRA are framing and right-handed park factors. (Judge will be speaking at Saberseminar this weekend and then plans to release these components publicly so you can compare the differences yourself soon.)

While Kershaw and Greinke play for the same team, the combination of batters faced, umpires and catchers have saved four runs for Greinke relative to Kershaw this year. And when it comes to the parks they’ve pitched in, Kershaw has lost three runs relative to Greinke due to the right-handed friendliness of those parks. Take the exact values with a grain of salt, but those two combine with minor differences elsewhere to amount to nine runs of contextual difference in Greinke’s favor. That’s basically the entire run prevention difference between the two pitchers.

For all of the earlier season “What’s wrong with Kershaw?” talk, he finds himself in a very real Cy Young race, while also chasing an esoteric, but impressive record. Getting to that 51 xFIP- will be tough, but the fact he’s within reach with a month and a half left speaks to his remarkable season.

It seems odd to talk about improvements for a pitcher who already had three 7+ WAR seasons to his name before his age-27 season, but here’s a pretty remarkable fact: Kershaw’s velocity is up to its highest level since 2009:

And the harder fastball — along with the two other insanely good pitches that he features — is allowing him to induce less contact, by about 3% than his previous years.

It would be hard for me to argue — given his excellent fielding-independent numbers, the contextual factors outside his control and this increased ability to limit contact — that Kershaw is anything short of the pitcher he was a year ago no matter what his ERA says.

He’s chasing Schilling’s xFIP record and Greinke in the Cy Young race. Neither is terribly meaningful in the grand scheme of all things baseball, but both illustrate the sustained greatness of the game’s best pitcher. And if you’re looking for another meaningless fact to monitor, the last time two pitchers from the same team finished first and second in the Cy Young race was in 2002. The pitcher with the record-setting xFIP, Curt Schilling, finished second. He lost to teammate Randy Johnson in large part because of a 0.91 ERA gap. The vote may go differently this time around, but it will depend on how Kershaw performs in pursuit of this particular slice of history.