The Drop in Yoenis Cespedes’ Launch Angle

Yoenis Cespedes has had a roller-coaster year. Hamstring, quad, hip, and heel injuries have cut weeks off his season. On the field, his overall performance isn’t inspiring; his walk rate is down, his exit velocity is diminished, and his wRC+ has dropped by 16 points. The outfielder’s hitting has fluctuated between dominant and poor, with his worst month coming in July: in 89 plate appearances, Cespedes launched just a single home run and recorded an isolated-power figure that was 43 points below league average.

Small windows of playing time can bring big performance swings, but Cespedes’ power drought wasn’t a product just of bad luck in a limited sample. Consider the chart below, which uses the LOESS method to smooth through Cespedes’ launch angle over the course of the season. Batted balls are ordered from the first (his first BIP on Opening Day) through the most recent (his last BIP yesterday). Horizontal bars are included to show his average launch angle in each of the four calendar months that make up the slugger’s season.

Cespedes’ average launch angle of 24.6 degrees in April was among the steepest in baseball. After sitting out May to recover from injuries, Cespedes returned for the next month and averaged a similarly high angle. He ended June with a .929 OPS, so his overall production didn’t signal anything out of the ordinary. But the real story is told by the smoothing curve, which shows how Cespedes was changing as a hitter. In June, his launch angle began a drop that accelerated into a plunge. By the latter part of July, he bottomed out at 12 degrees, a mark more fitting for a line-drive hitter than a slugger. His angle has climbed a bit higher in August, but it remains far below April’s range.

Such pronounced in-season launch-angle changes are unusual — Cespedes’ 29.2-degree standard deviation ranks within MLB’s 86th percentile — and there’s reason to think that they reflect real changes in his swing plane. The following table breaks up a few of his key stats for two periods: April/June (when Cespedes’ typical launch angle was above 20 degrees) and July/August (when his center has remained below 20 degrees).

The April/June pitchers who faced Cespedes were slightly fly-ball friendly, and the July/August pitchers have been slightly ground-ball friendly. But these figures are close enough to league average that such a sweeping change in launch angle would be unexpected. Meanwhile, Cespedes’ angle has dropped on both high and low pitches — by 7.9 and 8.2 degrees, respectively — so his changes appear to go beyond contacting pitch locations that are tough to lift. A more glaring result emerges when his launch angles are broken down horizontally: in the low-and-in quadrant where Cespedes thrives, his average angle has fallen 13.9 degrees in July/August. That’s his largest drop out of the four areas of the zone.

Cespedes’ revamp appears to be real, and the question that follows is whether it was disadvantageous. His poor July production hints that the shift was detrimental, but we can dig deeper. By running LOESS smoothers through Cespedes’ launch-angle/run-value combinations, we can determine his optimal launch angles on balls in play in each of the past three seasons.

Whereas his performance across different angles was very similar in 2015 and 2016, it has changed this season. The current-year curve is shifted to the left, which signals more success at lower angles but worsened hitting at higher angles. That’s led to a different range within which Cespedes performs at his highest level. In 2015, his production was maximized at a 21-degree angle. In 2016, his ideal center was slightly steeper, peaking at 23 degrees. And this year, that optimum has dipped to 18 degrees. The confidence intervals around these curves indicate that his true peaks may vary slightly, but it’s pretty safe to say that Cespedes is at his best right around the 20-degree mark.

Given this information, it would be logical to find that, in his excellent 2015 and 2016 seasons, he channeled that ideal area as often as possible. But, as the probability density curves below illustrate, his swing wasn’t centered at 20 degrees.

In 2015 and 2016, his most common launch angles were 13 and 15 degrees, respectively. So even though those swing planes led to stellar wRC+ figures of 135 and 134 in the two years, his launch angle wasn’t high enough to maximize production. The lower angle has probably helped Cespedes maintain low strikeout rates for a power hitter, but at a cost to his productivity on balls in play.

Perhaps Cespedes opened the 2017 season with the high launch angle to close that gap and optimize his swing. But the steep angle that worked for him in this year’s early goings wasn’t actually ideal, just like the low July angle wasn’t ideal, either. Instead, Cespedes is best served by generally aiming to swing near the 20-degree mark. By centering his swing in that area, he’ll maximize his value on BIP and be closer to rejoining the ranks of the league’s elite hitters.