Has James Paxton flipped the switch?

When the Yankees decided against making a big move for a pitcher at the trade deadline, they subsequently put pressure squarely on the pitchers they had in-house. Without external reinforcements on the way, the onus was on the team’s struggling starting pitchers to provide internal improvement. There would be no other way to fix what was ailing the Yankees’ pitching staff.

If that group of underperforming starters was to step up, they were likely to take one of two distinct paths. Either they would stick to what they were doing and hope some combination of better luck and better execution would lead to better results, or they would make adjustments, big or small, in an effort to improve.

James Paxton has made just two starts since the trade deadline, but it’s clear he has chosen the latter path. The Yankees need their big offseason acquisition to carry the rotation now more than ever, and he may have already made the adjustments necessary to perform at or near his full potential.

Ahead of Wednesday night’s start in Baltimore, Paxton spoke with Lindsey Adler of The Athletic (subscription required). He discussed his pitch mix, and his secondary pitches in particular. In short, Paxton throws his four-seam fastball the majority of the time, and complements it with a cutter and curve. Prior to this season, Paxton used the curve as his top secondary, but in his debut campaign with the New York, he flipped the script, relying on his cutter more heavily.

It’s too simple to say that that reversal among secondaries directly caused Paxton’s subpar first-half. This isn’t a straightforward case of a pitcher using a bad pitch more often than a good pitch, as both Paxton’s cutter and curve have historically been excellent offerings. Per Statcast, for his career Paxton has limited hitters to a .223 wOBA with his curve, and a .228 wOBA with his cutter. Go a level deeper, and he’s generated a .216 expected wOBA with the curve, and a .221 expected wOBA with the cutter.

Clearly, Paxton hasn’t suddenly been using an inefficient pitch more often in featuring his cutter more prominently. Perhaps, then, the way his cutter plays off the rest of his repertoire has had a negative impact on his performance. That’s exactly what Paxton hinted at in speaking with Adler, stating that the velocity separation between his cutter and his fastball isn’t great enough to really fool hitters.

He’s right on the matter, as his fastball has averaged just about 95 mph in 2019, with his cutter coming in at 88 mph, and his curve averages 81 mph. There might be more to it than that, however. The way Paxton’s pitches tunnel could affect how they ultimately challenge opposing hitters.

Baseball Prospectus hosts a suite of “pitch tunneling” figures that attempt to capture the ways in which pitchers deceive hitters. They calculate release point differential, which measures the average variation in release point between pairs of pitches. Using release point differential, we can determine whether Paxton’s fastball and his curve look more similar out of his hand, or whether his cutter and curve do.

Turns out, it’s not close. Using BP’s database, I pulled all pitchers that threw a fastball followed by a curve at least five times to a right-handed batter (Paxton rarely throws curves to left-handers). The average pitcher in that sample posted an average release differential of 2.5 inches. Paxton’s release differential came in much lower at 1.7 inches. The trend held when I reversed the query; Paxton has a release differential of 2.1 inches when throwing a curve then a fastball, compared to a sample average of 2.6 inches.

Paxton’s release differential shoots up when he pairs his cutter and fastball. When he throws his fastball, then his cutter, to a right-hander, they average 3.1 inches of separation at release point. Reverse the sequence and he posts a 3.0-inch release differential.

Can a couple inches of separation and a few mph of velocity separation really swing a pitcher’s season? It’s impossible to tell for sure, but anecdotally, Paxton does seem to generate better results when he’s emphasizing his curve rather than his cutter. For one, his fastball seems to perform much better when it’s played off his curve. In seasons in which he’s used his cutter more (2016, 2019), Paxton’s allowed a combined .366 wOBA with his heater. In seasons in which he’s used his curve more (2014, 2015, 2017, and 2018), his fastball has generated a combined .306 wOBA.

Perhaps just as relevant are his last two starts. On August 2nd against the Red Sox, Paxton went six strong, yielding just two hits and two runs and striking out six. He used his curve 19 times, his third-highest usage rate with the pitch this year. On Wednesday, he used his curve 30 times, by far the most he has this season. He yielded just one run over 6.2 innings, striking out seven.

Zooming in on his fastball again, over his last two starts, as he’s upped his curve usage, his fastball has yielded a .202 wOBA and generated whiffs on 28% of swings. In all other starts, he allowed a .406 wOBA and generated whiffs on 26% of swings. Small samples, to be sure, but it’s still heartening to see Paxton’s fastball has played much better recently as he’s changed his pitch mix.

Can Paxton swapping his secondaries really be a panacea? Can he maintain the form he’s shown so far in August all the way down the stretch and into the postseason? We don’t have the answers. Surely Paxton doesn’t either, or he wouldn’t have let his ERA balloon towards five in the first place. But he’s making adjustments, and they’ve paid off so far. That’s all the Yankees can ask for as they pray their rotation sorts itself out before October arrives.