Clayton Kershaw is, according to the PECOTA projections, supposed to be the best pitcher in baseball this year. This is hardly a surprise. He was the best pitcher in baseball last year. By ERA+, he’s also been the best pitcher in baseball over the last two years, the last three years, the last four years, the last five years, the last six years, the last seven years, and with enough innings to qualify, the seven-year veteran has been the best pitcher in baseball over the last eight years.

The projection is something pretty familiar for Kershaw: A 2.23 ERA, 237 strikeouts in 224 innings, a 19-9 win-loss record—numbers that would give him another Cy Young Award should he be in the running against pretty much anybody other than 2014 Kershaw. The 5.8 WARP would fit right in within 0.3 wins of each of his three best seasons and a small regression from last year’s performance.

We know, though, that this is just a single line reading of what’s really a (multidimensional) continuum of possible outcomes. Specifically, it’s the median. He’s 50-50 to be above this point in each of the metrics, and he’s 50-50 to be below.

Whereas the median projection looks downright mundane given what we’re used to from Kershaw, it was the tail—and not even the extreme tail—that stood out. Kershaw’s 90th percentile projection is almost unfathomable. The raws benefit from Dodger Stadium and for it always taking a while to adjust our eyes to the light of a new run-scoring environment, but the adjusted stats tell you it isn’t just your eyes. A 1.46 ERA, the lowest since Bob Gibson in 1968 and the second-lowest of the live-ball era. An 8.8 WARP, behind only 1999 Pedro Martinez as far as seasons in the WARP time (1950-present).

And this is 10 percent that he reaches or exceeds this. This isn’t some far off six-sigma best-case scenario. Ten percent is the chances of picking a Mark McGwire plate appearance from 1998 and seeing a home run. It’s greater than the chances of the Dodgers winning the World Series based on their fair odds implied by the betting lines.

So, I was all set to use our Kershaw Day to gush further on the 90th percentile what-could-be until I noticed something else.

The 10th percentile might be just as incredible.

10th percentile


90th percentile





















Kershaw’s nightmare scenario—short of the actual nightmare scenario involving a visit to… actually you know what, I won’t even say it—is between the guys who finished fifth and sixth in the NL in WHIP last year. It’s basically Madison Bumgarner’s 2014 ERA (2.98) and even if those were just ballpark tricks—which they’re not—it’s higher than James Shields’ 2014 WARP.

Sure the tails on that tail are extreme. There’s a non-zero chance he has a Justin Verlander season. In all seriousness, there’s always the injury concern.

But had he never signed a contract extension and hit free agency this offseason, his 10th percentile would have been worth close to nine figures on the market—maybe well into the nine figures, given his age.

Which is more impressive? That’s a matter of personal taste. To me, it’s still the 90th. I think it’s two things. First, that so much of my fear is placed in the injury bucket that I don’t really consider that 10th percentile as interesting. And second, that even though he had a 1.77 ERA last year, and even in a good ballpark and run environment, there’s such a leap mentally to the low-1s that the whole line becomes something out of the realm of reality.

But which one do you think is likelier? PECOTA says that an ERA below 1.46 and an ERA above 3.02 are equally likely. What say you?

Take this short poll; results coming later this week.

**Starts are held at 32, so as not to bog down the 10th percentile with various injury scenarios. This is for full-season performance only.

Sources: Play Index, Bovada

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
The amazing thing is that last year his ERA would have been 1.47 if we discount his one two-thirds of an inning blow up where he gave up seven runs. That should put into perspective that he really could get to the 90th percentile.
On the other hand, if you add in his playoff starts his ERA 'balloons' from 1.77 to 2.14.
In today's era dominated by pitching, seems more likely that he would reach the 90th percentile. However, not sure how much more difficult reaching 90% would be if the strike zone does indeed shrink.
It's all the same thing - the projection, median, upside & downside - and it's all mind-blowingly impressive.
This is pretty nuts.

The percentile thing matches pretty well with our eyes. According to PECOTA, there is the same chance that Kershaw exceeds his 90th percentile or performs worse than his 10th. You were split almost down the middle with 53 percent saying UNDER 1.46 ERA was more likely and the rest saying OVER 3.02 ERA was more likely.

I asked in the survey for your fan allegiance, and it's no surprise that the Dodgers were overrepresented with 45 Dodgers fans in 139 replies and those 45 voting 39-6 for the 90th. Put them back at a more representative percentage, and the balance is tipped the other way.

In other words, baseball fan gonna baseball fan, I guess.

Felix Hernandez, by the way, was 70-68 for his 90th percentile in the follow-up question.