August 17, 2015
Mother May I?
For the last month or so we've been going through Statcast data and applying it to player evaluation, much like the Astros purportedly did in acquiring Collin McHugh. Alan Nathan helped me introduce some work that took an in-depth look at pitch spin. We explored the concepts of gyrospin versus useful spin, a topic that Nathan wrote about previously here at BP. Last week we revealed part one of applying that analysis, wherein Tyler Lyons was identified as a breakout candidate based on his slider spin rate and opportunities for improvement in his other offerings.
This is part two of the application, this time focusing on curveballs rather than sliders. After all, McHugh's curveball was supposedly a big part of the reason the Astros fell in love with him. Of course, the application of this analysis is contingent on understanding how exactly McHugh went from riding the waiver wire to a top-50 pitcher in baseball. From last week's article:
There's this idea that the Astros looked at some data, saw that McHugh spun a good curveball, and scooped him up out from underneath the Rockies' noses, and that upon landing in Houston McHugh turned into a good MLB starter while changing almost nothing about his approach.
Analyzing curveballs with Statcast is a difficult task because, as we mentioned in our initial analysis, the PITCHf/x data does not play nicely with the Statcast data when it comes to curveballs. In order to identify pitchers who have consistent and potentially effective breaking balls we will compare their useful spin (derived algorithmically) to the total spin as observed by Statcast. Unfortunately in some cases the useful spin comes out as larger than the total spin which is, of course, physically impossible.
Unfortunately that means eliminating several pitchers whose data can't be relied upon because of this discrepancy. It would seem logical to think that pitchers with a ratio of useful spin to total spin greater than one do in fact have high ratios on their curve, but we simply can't build a compelling case without knowing the limitations of the data. For that reason, the following pitchers are not being considered in this analysis:
Richards, Hernandez, and Jesse Hahn have curveballs that are often lauded as among the best in the game. I'm not sure what would cause a discrepancy like this between the PITCHf/x data and the Statcast data, but I couldn't in good faith use these pitchers in this analysis. It's worth noting that there is likely to be something skewing all the numbers, which means that all pitchers have somewhat inflated ratios. That said, we can at least feel comfortable knowing that the rest of the pitchers have ratios that are physically possible. With that said, here are the next 10 pitchers in terms of ratio of useful spin to total spin:
There are two names on the list that fit the Collin McHugh mold. The first is Taylor Jungmann, a tall right-handed pitcher the Brewers selected out of the University of Texas with the 12th pick in the 2011 draft. For Jungmann, it's amazing the difference a year can make. Take a look at his comment from the most recent BP Annual:
The 2011 first-rounder Jungmann is struggling to adjust to his declined stuff, as he's now working in the high-80s and low-90s with a fringy breaking ball and an inconsistent changeup. One scout suggested his best days were in college at the University of Texas, and his best hope to make the big-leagues is as a back-end starter. Even then, his command must improve to make that a reality.
For Jungmann, things have already changed to some degree. His fastball has picked up some of the life it lost, sitting in the low-90s with regularity. His curveball is now among the top-spinning curves in MLB. He's seen tremendous success by using his fastball and sinker to get to his off-speed stuff. Take it with a grain of salt because the sample sizes are so small, but opposing hitters haven't been able to hit any of Jungmann's non-fastballs:
Will Jungmann be able to continue this level of dominance? Probably not. There is, however, reason to believe that he won't regress quite as much as one might expect given our expectations coming into the season. His curveball is clearly a legitimate weapon, and his fastball is back to what it was during his college years. This is only one data point, and it perhaps contradicts the findings of Matthew Trueblood last week, but information is power and this is another puzzle piece in your Taylor Jungmann jigsaw if your living room table is anything like mine.
Unfortunately for the GMs of opposing teams, Jungmann has perhaps taken his step forward already, and the Brewers are in full-on rebuilding mode. It's going to be tough for opposing teams to pry away his six years of team control from the Brewers and their soon-to-be analytics-focused GM.
The other player on the list of best curveballs who could be a breakout candidate is Twins pitcher Trevor May. Much like Jungmann, May was once a top prospect who has lost his shine while working his way through the upper minors. The Twins were able to acquire him in the Ben Revere trade, and he's since made his MLB debut.
That debut was pretty ugly. He put up a 7.88 ERA over nine starts and 10 total appearances in his first foray in the majors, last year. While May was a bit unlucky, his FIP (4.77) and walk rate (4.34 BB/9) didn't set him up for success, and the results followed in suit. This season May has pitched better, though it'd be hard not to given his near-eight ERA the year before. He's put up an ERA over four in his 29 appearances, to go with a DRA roughly half a run higher than that. This is only part of the story, because May has pitched 14 1/3 strong innings out of the bullpen this season, which brings down that ERA a bit.
While Jungmann made the jump to MLB and excelled, May made the jump and scuffled. Where does May need to improve to have the kind of success that Jungmann has enjoyed thus far this year?
May uses a five-pitch mix, though lefties see only four of those pitches. To be more specific, May throws a fastball-change-sinker-curveball (in that order) combination to lefties, while righties see fastball-change-curve-slider-sinker. May also has some strange usage splits depending on the count of the at-bat. First, the normal: His four-seam fastball usage stays fairly steady regardless of count, and his sinker is heavily used when he's behind in the count and all but disappears when he's trying to put opposing hitters away. Now the oddities:
The most important question is whether these oddities provide some insight into how May can improve. Here are a few simple suggestions.
Ditch the Sinker
Early this season May's sinker was a good groundball pitch, but he's arguably been more successful with his slider and/or changeup as his primary groundball offering. Ditching the sinker would mean throwing more of May's best pitches, which is surely a good thing.
Work Off the Change and the Fastball
Challenge Hitters Inside
The images above are from the catcher's POV, with the vast majority of May's hottest zones being away from the hitter. Righties see pitches mainly down and away, an attribute coming out of the increased breaking-ball usage for May against same-handed hitters. Lefties have a more varied zone profile, though it still leans heavily toward the outer half. Remember that May leans heavily on his fastball and change against these hitters, so this zone profile makes some sense.
That said, challenging hitters inside would help keep opposing hitters from sitting on a certain pitch in a certain location. We already know that May has some pretty consistent usage patterns, so it wouldn't be hard to figure out what pitch might be coming in a given situation. If the hitter also knows that the fastball he sees is going to be on the outer half, then his job is that much easier. Pitching inside would keep opposing hitters honest, and help May work later in the count to any given hitter.
May is clearly not enjoying the success that fellow curveball expert Taylor Jungmann is seeing. That said, he has the potential to break out like Jungmann has; it's just a matter of putting the final pieces together to make things click. May has the credentials: a high-spinning curve to go along with a solid fastball and great change. The only question left is whether the Twins, or some enterprising team, can fix what ails him. May has the tools to succeed; he just needs to learn how to leverage them.
Thanks to Alan Nathan for his help in deciphering, manipulating, and understanding the Statcast and PITCHf/x data, and for his ongoing support of this analysis