Sandy Rosario, a 27 year old Dominican native right-handed reliever, had one of the most interesting off-seasons in all of baseball. Despite what appeared to be a good fastball, he was considered a fringy reliever in the scouting community, and was designated for assignment by the Miami Marlins, who he had appeared in a handful of games over three seasons. The Red Sox claimed him off waivers, but just over a month later, traded him to the Oakland A’s. Two weeks later, the A’s took him off their 40 man roster and he was claimed again by the Red Sox. He lasted a grand total of two days as a Red Sox again, and was claimed by the Cubs, who let him go less than 10 days later. The journey finally ended, still in December, when he was claimed by the San Francisco Giants.
The Giants started Rosario in the minors, but brought him up at the end of April, where he has been here to stay, pitching very well in a small sample size by both traditional and advanced metrics. So far in his career, he hasn’t really gotten swings out of the strike zone at an above average rate, but Rosario throws in the zone more than league average, and gets less contact than league average. At this point in his career, Rosario has thrown 536 pitches tracked by Pitch F/X. Release point wise, his closest comparison is David Robertson, but as you can see, it isn’t exactly uniform for his career
However, when you look at just his time with the Giants, Rosario’s release point is much more uniform, which could be because he has finally found a release point to stay, a result of taking out multiple seasons, or just a small sample size fluke with the way different parks measure release points
There appears to have been quite a dramatic change in Rosario, whether via his choice, or by the Giants prodding him. This change has moved his average release point, as well as his average pitch location
He is actually releasing the ball from a lower arm angle, and is further out from the center of the rubber. Strangely, this has made him more of a glove side pitcher than an arm side pitcher on average, and he is actually throwing the ball higher. A change in pitch selection does help explain this, as he is throwing less fastballs and changeups, and more sliders according to MLB AM tags this year than he was with the Marlins. Again, this is something you wouldn’t expect to lead to success, just like the release point change of coming further out and lower. For one, the Giants are using him pretty much as a platoon pitcher, facing right-handers two-thirds of the time (versus just 40% of the time in his short stint with the Marlins). This would explain the difference in slider usage. However, he is locating the individual pitches differently (entire graph is strike zone):
He is locating the fastball more up in the zone, and getting the slider slightly lower, and much more horizontal on average. He is getting less spin on the slider on average as a Giant (which can be a good thing with breaking pitches, though more so with curveballs than sliders), and he is getting more movement both horizontally and vertically with it. It still isn’t excellent movement, and it could have something to do with park change (movement data from park to park is very different, but it does seem to be more, which would mesh with the idea that it is spinning less (I am unaware if there are spin biases by parks, though it would be very odd if there were) and the change in release point would most likely give him different movement, even if it wasn’t better.
Becoming a slider heavy right-handed specialist can provide a nice niche for Rosario and keep him on a 40 man roster for quite some time. However, there are some drawbacks. While correlation doesn’t necessarily mean causation, Rosario has lost quite a bit of fastball velocity this season, almost two full MPH from where he was as a Marlin. It is doubtful he will ever get that back, and at his age, he is likely to lose some more velocity in the coming years. He still throws the fastball half the time, he isn’t Luke Gregerson or Sergio Romo, at least not yet. If the fastball regresses even more, it is unlikely that the changed release point and the improved slider will outweigh it, and it is hard to see him having sustained success.