Clayton Kershaw’s somewhat new weapon in his fight against predictability

On Tuesday evening in Game 1 of the World Series, Clayton Kershaw made one of the biggest starts of his life. It was a bad matchup on paper: the Rays offense was very good against left-handed pitching this season, and was exceptional against off-speed pitches, which Kershaw throws more than half the time. Kershaw responded with a start which could have ranked among the best of his career had Dave Roberts not elected to pull him early for extra rest after the offense gave the team a large lead.

Kershaw threw 78 pitches in his six innings of work, striking out eight Rays and allowing just one run on two hits. After a wobbly first inning, he looked unhittable for most of the game until Kevin Kiermaier smacked a hanging slider into the right-field stands. Of the 38 swings the Rays took against Kershaw on Tuesday, 19 were missed. A 50% missed swing rate is the best Kershaw has ever had in any start, both in the regular season and postseason.

The major criticism after most bad Kershaw outings is that he’s too predictable. After all, he throws three pitches and follows a pretty rigid pattern with them. That was indeed the complaint after Kershaw’s mediocre start against Atlanta in the NLCS: that the batters were ready for Kershaw the third time around because he didn’t vary his approach enough in the first two trips.

There is some truth to that, but interestingly enough, Kershaw did not adjust one area of alleged predictability against the Rays: the location of his fastball. Against Atlanta, 64% of his fastballs were to his glove side, the side he typically targets. Against the Rays, that number rose to 71%. 57% of Kershaw’s fastballs were below the belt against Atlanta, that number was almost identical when facing the Rays. The Rays were also just about as aggressive in their approach against the first pitch, and Kershaw was just as insistent on throwing a first pitch glove-side fastball as ever. The Rays just didn’t do anything with them.

However, there’s more than one way to be unpredictable, and Kershaw did just that against the Rays. Here was the first hint, an 84mph slider to Brandon Lowe:

Before that pitch, Kershaw threw 14 other sliders, ranging between 87 and 90 miles per hour, a pretty typical look for him. His fastball velocity was normal before and after this pitch, so this slider velocity wasn’t impacted by health, it was something intentional. Kershaw’s next pitch to Lowe was an 85mph slider, which he barely avoided swinging at.

Later in the game, Kershaw threw an 82mph slider to Joey Wendle, which he fouled off. It was right down the middle. And on his last slider of the game, again to Lowe, he threw this 81mph pitch:

Like the 82mph slider before it, it was thrown in a bad spot, but Lowe seemed unable to time it. This is where the unpredictability comes into play, if all of Kershaw’s sliders were in the same velocity window, maybe Lowe would have made a better swing on the mistake.

The slower sliders also moved differently than the regular ones. Below is a plot of every pitch Kershaw threw in his Game 1 start, plotted with velocity and vertical movement:

As you can see, there are three groupings, as you’d expect for a pitcher who throws three different pitches. On the lower left are the curves, on the upper right are the fastballs, and just below them are the sliders. However, the slower sliders are leaking into the curves – the particularly slow ones have a lot more sink than the hard cutter-ish ones. If the movement is distinct, batters almost have four pitches to worry about, not three.

This is not the first time Kershaw has experimented with throwing his slider more slowly. Here’s one from a start against the Diamondbacks in September:

This experiment didn’t even start in 2020. Here’s a slower slider from last year:

In 2020 Kershaw has thrown 14 sliders below 84mph. In 2019, Kershaw threw 17 of those pitches. You can see when he started experimenting with it last season on this graph of pitch velocity broken down by game. The lower bars on the red line represent the velocity of Kershaw’s slowest slider for each contest:

The timing of the more frequent appearance of the slower slider last season also corresponds with his hottest stretch of that year: July and August. In his best 10 start string of 2019, from June 8th to August 16th, Kershaw had a 2.08 ERA with 74 strikeouts and 16 walks in 65 innings. This season, Kershaw had a 2.16 ERA with 62 strikeouts and 8 walks in 58-1/3 innings. Kershaw’s strong 2020 season can also be attributed to his extra velocity, but he had stretches like this in the extremely recent past.

The slower slider isn’t why those two stretches were good or why they were so similar, but perhaps a sign that Kershaw has more confidence in his stuff and is more willing to experiment with it. It brings to mind the times when Kershaw lowered his arm slot on particular pitches to left-handed batters a few years ago.

Another thing that brings those arm-slot dropping fastballs to mind is how Kershaw is actually using the slower slider. All 14 of Kershaw’s sliders below 84MPH this season have been thrown to left-handed batters. Interestingly, this is an adjustment from last season, when only 8 of 17 such pitches were thrown to lefties. It’s unclear why Kershaw now feels the need to only throw this pitch to his fellow southpaws, but for now they’re the only batters with something extra to think about.

Overall, then, the slower slider is only a minor weapon in the battle against predictability. Perhaps using it a bit more frequently against the Rays was in response to Freddie Freeman beating him on the third time around in the NLCS. Either way, Kershaw has now thrown five of his 14 slow sliders this season in the playoffs, so it will be something Rays hitters will have to look for in Game 5.