The Mariners have had a different Diego Castillo

Every time the Rays make a trade, it’s regarded as a win even before the paperwork is completed. Never mind that they traded away Willy Adames for a pair of relievers in Drew Rasmussen and J.P. Feyereisen, or that Jake Odorizzi became a better starting pitcher after leaving Tampa Bay. That doesn’t fit the narrative! So when the Mariners traded J.T. Chargois and Austin Shenton to the Rays for Diego Castillo, immediately, there had to be something that the Rays would unlock in Chargois, and Shenton had to be an under-the-radar prospect. Bollocks! I think it has more to do with Castillo being arbitration-eligible next year, making him too pricey for the stingy Rays.

Over his career, Castillo has posted a hearty 19.9 percent strikeout minus walk percentage, but his career 69 ERA- and 90 FIP- suggest that he’s either benefited greatly from the Rays’ park or defense, or that there’s hard contact suppression inherent in his batted ball profile. There’s truth to all of this: Tropicana Field has historically skewed towards pitchers, and DRS and UZR suggest the Rays have been a top-ten defense since 2018 — so those have played a role — but also, his xwOBAcon has reliably been well below league average.

There’s something to that! While he allows barrels at a league-average rate, Castillo has proven to induce weak contact more than the vast majority of pitchers: his 66.5% weak percentage ranks in the 95th percentile. But since coming over to the Mariners, Castillo has not only been getting hit as hard as ever, but he’s also missing bats at an all-time low, too:

This all comfortably fits the narrative. Perhaps it’s not just that they were trying to cut costs. It could very well be that there was something amiss that the Rays saw in Castillo that was worthy of capitalizing on his then-value. The Rays traded their trusty reliever away, and then he imploded.

Consider a representative release point from 2019:

And then from August 19th, before he went on the injured list:

The difference is striking. In 2019, Castillo’s shoulders were almost exactly at a 45-degree angle, relative to the ground. In his most recent game, though, his shoulders are much more parallel to the ground, and he’s dropped his release point six inches since 2019. That’s coincided with a dip in velocity — he’s gone from 98 miles per hour in 2019 to 96 mph in 2020, and finally 95 mph in 2021.

A loss in velocity is often a sign of an arm injury, and a lowered release point is one way of compensating for an elbow or shoulder that isn’t 100%. But also, it could just be that his optimal release point for throwing hard is a higher one. I find the argument that Castillo has been hurt all along to be uncompelling — especially because this would have been detected in a physical when he was traded. What I do find compelling, though, is that Castillo hasn’t quite been the same pitcher with the change in arm slot.

Castillo’s pitch movement, over the years:

When not accounting for gravity, his sinker has become even heavier than his slider. But also, his slider is now averaging arm-side movement. Not glove-side movement! Arm-side movement. These are both direct results of a change in arm slot and a reduction in velocity.

Of course, these things only matter insofar that they have begun to affect his outcomes. We can’t be for certain that any one thing is the culprit for Castillo’s decline — as with most things, I think it’s a combination of issues, perhaps the most compelling being seam-shifted wake (SSW), and pitch location.

The two are related, but we’ll start with SSW. First, Castillo’s 2020 spin direction at release and at the plate, from the hitter’s perspective:

At release, the spin direction of Castillo’s 2020 slider was, on average, 10:15 on a clock face, ranging from about 8 to 2 o’clock, typical of a pitch with heavy gyroscopic spin. By the time it got to the plate, though, its average had shifted to 9 o’clock, roughly spanning from 6 to 12 o’clock. That means that, on its way to the plate, the spin direction of Castillo’s slider had shifted significantly to his glove-side, and his sinker to his arm-side. That means that his pitches are spinning significantly differently by the time they reach the plate than they were upon release. To a hitter, that means that SSW causes them to end up in unexpected locations, too. Sliders end up more glove-side, and sinkers more arm-side.

That’s changed in 2021:

You can see why Castillo’s slider has begun to spin to his arm-side: its spin direction has shifted significantly at release. On average, his slider is now spinning directly over the top at 12 o’clock, spanning from about 9 to 3 o’clock. It’s become much more of a slider-cutter hybrid than the previous iterations of his slider.

Then there’s his location. First, his slider location:

Not much is changed here, other than that Castillo has started to locate more to his glove-side. If anything, I might argue that he’s locating his slider more optimally.

Then there’s his sinker location:

Castillo has shifted his sinker location glove-side as well. Given that this happened between both pitches, I’m inclined to think that this is related to his dropping release point, whether it’s that his pitches are moving differently, or that his lowered release point has made him more susceptible to his glove-side by yanking the ball across his body.

For a pitch that Castillo throws two-thirds of the time, his slider is awfully important. His sinker has always been unexceptional. Like the departed Chargois, Castillo throws his sinker just enough to keep hitters off of his slider. But now that he’s lost the feel of his slider, both his sinker and slider have suffered the consequences, and he’s neither dominant by whiffs nor contact management.

If Castillo returns from the IL looking similarly to how he had in mid-August, perhaps he was hurt after all. I’m not opposed to the notion that I’m full of shit and reading too much into this. Perhaps more likely is that the Mariners used a phantom IL stint as an opportunity to get Castillo’s mechanics back to where they were, and that may mean that, upon his return, vintage Castillo is in store. It seems clear that Castillo is better from a higher arm slot than a lower one, and his success will probably be dictated by whichever arm slot he brings back from his rehab outings with the minor league affiliates.