Are Stuff And Location The Same Thing?

Half-baked thoughts on release angles

A few months ago, I started to wonder why some four-seam fastballs have more induced vertical break (iVB) than other fastballs. I ran a linear regression and learned that the vertical release angle — in other words, the angle that the ball is pointed out of the hand at release — has a stronger relationship to iVB than release height, velocity, or any other variable that I could add from a Statcast .csv. I wrote about all of this (and more!) in a new post for FanGraphs headlined “It’s Release Angles All The Way Down,” but there’s an idea I left out of that piece and want to briefly explore here.

This idea, roughly, is that stuff and location can’t be separated. Here’s why.

Holding release height and arm angle constant, fastballs thrown toward the bottom of the zone have more induced vertical break, or carry. Fastballs thrown toward the top of the zone have less carry.

The theory of why, as I write about in the FanGraphs post, is that throwing toward the bottom of the zone allows pitchers to get “behind the ball,” allowing them to throw with more backspin and therefore more carry. My research on sweepers and sinkers is preliminary, but there also appears to be a relationship between horizontal break and release angles — getting more “around the ball” appears to allow for more sidespin and therefore more break.

This means that the act of aiming and throwing to a specific location changes how the pitch moves. Which leads to a weird question: Can stuff can be actually separated from location? If the location that a pitcher aims the ball impacts the pitch’s movement properties, can stuff and location be untangled?

The implication of “stuff” models is that they can be untangled. But a fastball has a “good shape” in part because of its location — not just because certain locations are better than others, but because the location itself tells you about the way the pitch moves.

These are some of the reasons why I think teams might be moving away from stuff models and toward biomechanical models. As I wrote in the article:

This biomechanical information — not just release angles, which have a strong relationship to both command and stuff, but everything downstream of these release angles — can potentially be leveraged to build better predictive models, identify hidden draft prospects, spot injuries, and accelerate the player development revolution.

Stuff and location, ultimately, are downstream of a pitcher’s biomechanics. Ball flight informs pitching outcomes, but biomechanics inform ball flight.

I’m not sure if any of this makes sense, but I’d love to hear what you think.