What Else You’ll Need Besides Spin Rate

Soon, we’ll get games on television again. I know it’s hard to believe, because it seems like it’s been so long, but it’s true. It’ll only be spring training, but it’ll be baseball and it’ll be great.

Along with these televised games, we’ll hear commentators discussing key statistics. It’s very possible that, due to the rising popularity and availability of Statcast, we’ll hear about increases in spin rate when it’s relevant. That’s great! But there will be some context we won’t hear, and that context will be important.

One bit of context comes from the concept of useful spin. We report total spin, but that’s just the spin imparted on the ball. Total spin. But spin has to be considered in the context of arm slot. Imagine trying to put backspin on a ball. The more upright your hand is on release, the more of your spin is “harvested” as backspin that’s useful to the movement you’re trying to produce; the more your arm drops down, the more that spin gets converted into useless sideways spin.

It looks like spin is inherently useful when you look at charts like this one, and I’ve heard from batters before that higher spin pitches are harder to pick up. But we haven’t yet publicly nailed the usefulness of spin by itself, at least not from the research I’ve seen. We’re working on it.

Another piece of context for spin: it’s related to velocity. It generally goes up with velocity, something Trevor Bauer discovered when he created the Bauer Unit — spin divided by velocity. You’d want to know if the pitcher who increased his spin also increased his velocity, for sure.

In fact, when you look at the pitchers that added significant spin (+10%) from 2015 to 2016, all of them but one also increased their velocity, as well.

If you agree with Arik Florimonte’s assessment that it’s more 50/50 and we shouldn’t count all of those 1.01s as velocity increases, fine. We can do this mathematically, too. And we can switch over from velocity to effective velocity, which takes into account the point from which a player releases the ball and adjusts the velocity for that release point.

The relationship of four-seam fastball spin from one year to the next is pretty strong. The r-squared is .816, meaning that last year’s spin explains about 80% of the variance in this year’s spin. If you limit it to only pitchers that have thrown 100 pitches in both years, that r-squared jumps to .851. That’s immediately stronger than any results-based pitching metric in terms of year to year stickiness.

Harness spin to effective velocity, though, and your year-to-year correlation could improve, given the correlation between velocity and spin. I found the average spin for every non-decimal effective velocity between 85 and 96, and then indexed every pitcher’s spin against that average. Justin Verlander, king of spin, produce a spin rate that was 15% better than the league average for his effective velocity bucket in 2016. Look at how strong the relationship is between this effective-velocity indexed spin rate from 2015 to 2016.

Now last year’s effective-velocity indexed spin rate explains 83.8% of the variance in this year’s number.

So which pitchers actually improved from 2015 to 2016? By this second methodology, there are nine pitchers who increased their spin more than 5%, but it’s a different set of pitchers than those cited above. These are the true outliers, although really only the first two satisfy our original requirements for significant change (10%).

Returning to the original point of this post, we find that, if a pitcher is producing different spin rates, there are more questions you have to ask to really understand the whole context. Did he change his arm slot? Is he harvesting as much of that spin as he was in the past? Did he improve his velocity? Did he change his mechanics? Is he releasing the ball in the same space? And, even simpler, what does that spin rate change look like against league averages?

It’s okay if we don’t get the full deluge of facts every time spin rate comes up, though. There’s an ongoing battle between context and brevity when it comes to producing good sports television. Too many columns, too many stats, too much context, and you lose the viewer. Too little explanation around the story you’re trying to get across, though, and it loses some power. The number without context becomes a meaningless abstraction.

So spin rate changes? Sure. I want to know about that. I’ll also want to know a few more things.

[A data snafu with the first pull led to incorrect numbers. I’ve updated the post to reflect this fact. Thanks for pointing it out!]