Through the symbolic but not mathematical halfway point, everyone in the Yankees starting rotation has had an interesting thing in common. None of them have been good at throwing curveballs. Fangraphs offers a metric called "Pitch Values," which attempts to quantify how many runs a certain pitch has saved or cost a pitcher. According to this metric, Yankee starters are the worst in the league at throwing curveballs, with a cumulative score of 10.1 runs below average.
To be fair, CC Sabathia and Michael Pineda don't actually throw curveballs. CC's high 70's slider can look like a curveball, and has been a solid weapon for the big man. Pineda and Masahiro Tanaka also both throw very good sliders, so it's not as if Yankee starters simply don't have breaking balls to use.
Early in the season, Fangraphs' August Fagerstorm noticed that the Phillies were making use of the curveball, with everyone in the starting rotation featuring one. While the Phillies have come back down to earth since April, Fangraphs' pitch value metric still has the Phils rotation as the best in baseball at throwing curveballs.
One possible explanation for the Yankees' ineptitude with curveballs can be explained with Statcast data. Here are the average spin rates of the Yankees' starters in 2016:
|Player||Curveball Spin Rate|
|Masahiro Tanaka||2244 RPM|
|Nathan Eovaldi||2274 RPM|
|Ivan Nova||2335 RPM|
|2015 League Average||2308 RPM|
Nobody in the Yankees' staff throws a curveball with a particularly high spin rate, and none of them have a particularly high arm slot. In an earlier post, I argued that pitchers with lower arm slots need higher spin rates on their breaking balls, since they can't depend on gravity as much as pitchers like Clayton Kershaw who throw over the top. With this in mind, the Yankees' trouble with the curve could be related to the collective inability to generate spin on the pitch.
While the jury is still out on whether a pitcher can increase the spin rate of his four seam fastball, there is reason to believe that a pitcher can add RPM's on his curve. Here is a scatter plot showing qualified pitchers' curveball spin rates and fastball spin rates, with Phillies starters in red and Yankees' starters in blue:
*R-squared = 0.21, min 200 fastballs and 200 curveballs, except for Yankees and Phillies starters
There is a decent correlation between fastball spin rate and curveball spin rate, but there are also some clear outliers. Obviously, there are some mechanical adjustments that pitchers can make to add spin to their curveballs.
It might not be a huge problem, but it is noteworthy whenever an entire pitching staff can't effectively throw a certain kind of pitch. Curveballs are considered to be more platoon-neutral than sliders, and the mostly right-handed Yankees rotation has the eighth-worst ERA against left-handed hitters. Newcomer Chad Green has had sizable platoon splits at both Triple-A and in his brief major league career, while Luis Severino's slider had yielded an .861 OPS in the majors.
One possible problem for Severino is his inability to change speeds. His changeup, which is his slowest secondary pitch, is thrown at an average 88.6 mph, roughly seven miles per hour slower than his average fastball velocity. In theory, everyone on the Yankees' pitching staff could stand to gain from throwing a curveball.
With the abundance of available game film, more qualified people within the Yankees organization should try to figure out what allows Phillies starters like Jeremy Hellickson and Aaron Nola to generate so much spin on their curveballs. As inconsistent as the Yankees' rotation has been, it features several young arms with plenty of time to improve.
*Data is courtesy of Fangraphs, Baseball Savant, and Baseball America.