I like the Rays.
They've been the inspiration behind a couple of my pieces, whether that be about Tyler Glasnow allowing a 458 foot homer or how to evaluate relievers' salaries.
Every move the Rays make contrasts traditional baseball decision-making. Baseball goes left; the Rays opt for the right path (literally and figuratively). Baseball spends millions on free agents, the Rays trade for their 15th competent outfielder. Baseball convention states if your starter is rolling, you leave him in, the Rays...... well, you know what happened.
Their success can be attributed to their outstanding analytics team, the group that dictates much of their in-game moves. Their staff, which is routinely considered the best in the game, takes into account every minute detail of a baseball game and optimizes it, whether that be by creating new positional alignments, revamping the idea of a starter, or filling up their bullpen with pitchers with different release points.
A couple of weeks ago, I was perusing Twitter and noticed this article by MLB.com's Mike Petriello.
Pitcher List's Alex Fast later sent out this tweet, prompting me to explore a question that's been on my mind: is it important to stock your bullpen with pitchers who throw from unique release points?
https://twitter.com/AlexFast8/status/1320855440377421824
Common sense would dictate so. It makes plenty of sense, logically-speaking. Besides, it jives with the most accepted rule in baseball: Always Trust the Rays.
Their relief corps enjoyed massive success in 2020, with their fWAR, ERA, and FIP all finishing top-5 in the Majors in their respective categories. As shown in Fast's GIF, their bullpen featured a ton of pitchers that had funky release points.
But, can the success of their bullpen be directly attributed to the diversity of their pitchers' release points? I wanted to find out.
I collected every reliever's data from the 2015-2020 seasons and sorted them into their respective teams. (I excluded all relievers that changed teams midseason). I then tallied each pitcher's release point (RP_X) and added a weight that was based on his innings that season. After all, Drew Gagnon's 12 innings of work for the 2018 Mets shouldn't count as the same as Aroldis Chapman's 57 frames for the 2019 Yankees.
I then added up each pitcher's weighted RP_X's, calculated the standard deviation of each team's bullpen's release point, and plugged it into my correlation matrix.
I set my dependent variable as the standard deviation of each bullpen's release point. If a team's standard deviation was higher, it showed that they had more pitchers with different release points.
Rather surprisingly, I found my initial guess to be quite inaccurate. There was no evidence of a strong correlation, or really, any correlation between my dependent variable and a bullpen's overall performance.
It's worth noting that the 2020 team with the highest standard deviation was the Rays. No surprise there.
Despite my earlier assumption, it's apparent that loading up a bullpen with pitchers who throw from unique arm slots doesn't have an impact on a bullpen's overall performance.
I now needed to do something that terrified me: explain why the Rays are wrong.
Did the Rays just do their math wrong?
I doubt it.
Did I? Now, there's a much higher chance of that happening.
After a couple of days of reflection, I realized a major caveat in my study, the implication of which is major. I was conducting it too broadly, and instead, should've taken a more granular approach.
Having a lot of pitchers who throw funky represents the only one slice to the big eight-piece pizza pie.
There's a fantastic scene in Rian Johnson's 2019 murder-mystery, Knives Out, where the protagonist detective (played by Daniel Craig and his 10/10 southern drawl) speaks about a murder through the metaphor of a hole inside of a doughnut, before ending his speech by stating, "We must look a little closer. And when we do, we see a doughnut's hole has a hole in its center."
That's what this study represents. It's a hole inside of a doughnut, but if we look a little bit closer, we're able to look into the center of this doughnut's hole.
And when we do, it makes sense why the Rays' bullpen pitched so well in 2020. It's not only because they filled their bullpen with relievers with funky release points, it's because of the way they used them.
Having a lot of relievers who all have different release points doesn't help a team out, but having the ability to switch in and out pitchers with unique release points certainly does. And that explains why the Rays stocked their pen with these pitchers. By maximizing their chances at giving opposing batters contrary release points, they're able to play favorable matchups.
https://twitter.com/mike_petriello/status/1315989914790563840?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1315989914790563840%7Ctwgr%5E&ref_url=https%3A%2F%2Fwww.mlb.com%2Fnews%2Frays-pitchers-unique-arm-angles
Let's say you are a Major League hitter facing Major League pitching (and if so, good luck to you sir). The first time through the order, you're facing Tyler Glasnow, a tall right-hander who releases his pitches at 12 o'clock. For your second plate appearance, Aaron Loup, a veteran southpaw who throws from a funky arm slot, gets the assignment against you. Your third time at the dish, Ryan Thompson, a side-arming frisbee-throwing righty, is staring you down sixty feet, six inches away.
Good luck swinging at pitches that are coming from three completely different arm slots.
The Rays have a knack for upending traditional baseball strategy. Simply stocking up your bullpen with pitchers with different arm slots isn't enough. What's more important is how a team deploys them.
Always Trust The Rays.
Opening Picture Credit: Alex Fast of PitcherList and Mike Petriello of MLB.com.