Puzzles

There’s been some interesting research done recently on pitcher release angles, horizontal and vertical (HRA and VRA), in quantifying a pitcher’s ability to locate their pitches well, as well as better explaining individual pitch performance.

In the research I’m referencing, Michael Rosen from Fangraphs found a relationship between a pitch’s HRA and inducing swings, identifying that the closer a slider’s HRA was to a pitcher’s fastball’s HRA, the more swings pitchers tended to get on their slider. Similarly, Eli Ben-Porat found that mirroring the VRA of a gyro slider to the fastball’s VRA has a relationship with a pitcher’s ability to generate swing-and-miss on their gyro slider.

It’s been my understanding that the higher a pitcher’s arm slot is, the more difficult it becomes to throw a sweeper, or more generally, generate horizontal movement on a breaking ball. Yet, there are plenty of examples of major league pitchers with high 3/4 or over-the-top releases that throw a sweeper (Aaron Civale, Kutter Crawford, Griffin Jax). Many drop their arm slot in doing so, but seemingly never by very significant amounts. After reading about quantifying release angles, I guessed that pitchers with higher slots probably weren’t as likely to be able to replicate a similar HRA or VRA between their fastball and sweeper as pitchers who fell into more of a low 3/4 arm slot classification. Using arm slot data created by Trevor Thrash, we can gain some insights into this question.

For all pitchers with at least 50 4-seam fastballs and 50 sweepers thrown in a given year from 2021-2024, I found there to be a moderate correlation (r = 0.48) between HRA differential and arm angle at release, but not in the way I expected to. The data showed that the lower the arm slot, the larger the differential in HRA was. Instead, I found VRA to be the release angle metric more aligned with what I was expecting to see, with a strong correlation (r = 0.54), where a higher release slot correlates with a larger VRA differential. This correlation grew to r = 0.61 when removing pitchers who had more than a 5 degree difference in arm angle between their 4-seam and sweeper.

For pitchers in the sample, there was also a moderate relationship (r = 0.36)  between VRA differential and Whiff%, with smaller VRA differentials leading to higher sweeper whiff rates. Interesting, but I might eventually opt to use a similar methodology to Ben-Porat when he looked at only gyro sliders in order to better measure whether similar trends from his findings applied to sweepers.

Caleb Thielbar is an interesting case study when considering everything detailed above. Thielbar met the criteria qualifications for every year used in this study, but this season is a bit of an outlier for him when it comes to the VRA differential.

So far in 2024, his Sweeper is averaging the lowest amount of horizontal break of all the seasons, and without any gains in velocity, has had the highest Whiff% of the four. Is it a result of better locations? Maybe. But Fangraphs Location+ doesn’t think so. It’s possible that closing the VRA gap has helped Thielbar so far this season.

To recap, yes, it’s possible for your pitcher with an over-the-top release to throw a sweeper. But at what cost?

*2024 data is through 6/15/2024*