A Brief Investigation of Spiked Curveballs

This post is all about curveballs; specifically, those that are spiked. Of course, unavoidably, all types of pitches get spiked at least once in a while, but curveballs likely end up in the dirt more often than most pitch varieties.

That might be the case for several reasons.

For one, curveballs, by definition, are intended to move downward — north to south — and miss barrels by dipping beneath a batter’s bat path. The downward rotation of the baseball, invariably subject to gravity and additionally aided by the magnus effect, embellishes a curveball’s downward arc. This characteristic arc informs how the pitch is used: a curveball looks to be a strike at its apex, and bends out of the strike zone toward its zenith.

Sometimes, due to how a curveball is gripped, a pitcher’s finger or knuckle simply catches on the baseball, and the pitch is yanked downward unintentionally. Pitchers are human, after all.

Finally, curveballs might be spiked more often due simply to being thrown low in the strike zone on average. The histogram below illustrates the vertical height of curveballs thrown by qualifying starters in 2021.

As you can see, the vertical height of curveballs thrown by qualifying starters last year are quite normally distributed. The bottom of the strike zone is, give or take, 1.5ft, or 18in. The average height of a curveball thrown by qualifying starters last season was 1.8 feet, or just a couple inches above the (generally held) bottom of the zone.

This downward trend continues with the development of a plate appearance; as the count deepens in a pitcher’s favor, curveballs get thrown lower on average. Below is the average vertical height of curveballs, grouped by the number of strikes on a hitter, thrown by qualifying starters in 2021.

What about spiked curveballs, though? How often are they occurring and who benefits from them most? To account for their occurrence, I defer to Baseball Savant. For that same group of pitchers last season, all curveballs that were defined by Baseball Savant as being either blocked swinging strikes or simply blocked balls I have taken to be “spiked.”

Of all the curveball variations thrown by qualifying starters last season, 6.74% had pitch descriptions as either blocked balls or blocked swinging strikes. Of those pitches, there was roughly a 2:1 ratio of blocked balls to blocked swinging strikes. In all, hitters gave chase on 32.93% of those pitches, a higher percentage than this writer might have expected.

Interestingly, there was not too terribly much of a height discrepancy between those spiked pitches that were swung at versus laid off of. The mirror histogram below depicts the vertical height of each type of outcome.

Blocked balls had a median height of .455 feet while blocked swinging strikes had a median height of .610 feet. The difference between those two median heights is less than 2 inches. So, roughly 1 in 3 spiked curveballs translate to a swinging miss, despite not being terribly different in height relative to spiked balls that were laid off.

Finally, who benefitted most from employing curveballs — of the spiked variety — in 2021? Below is a table of all qualifying starters in 2021 that threw more than 200 curveballs according to Baseball Savant.

Several pitchers got a lot of mileage out of spiked curveballs. Germán Márquez got the highest percentage of his curveball swinging strikes by way of spiking them: of the 104 swinging strikes he induced, 34 (or 32.7%) came by way of a spiked curveball. On the other end of the spectrum, Sean Manaea, who didn’t induce many swinging strikes on curveballs in 2021, had even less success with spiking it. Of 54 swinging strikes on curveballs, just 1 came by way of a spiked curveball.

Who had the highest percentage of success among only spiked curveballs, though? The far right column covers swinging strikes that were blocked as a percentage of all blocked balls. Here, Corbin Burnes reigns supreme: batters gave chase on nearly 61% (!) of his spiked curveballs last year, nearly double the average percentage among this group. His teammate, Brandon Woodruff, wasn’t far behind though, inducing swings on over half of his spiked curveballs.

So, roughly 1 in 15 curveballs were spiked by qualifying starters last year, and the range of outcomes/success on those pitches was incredibly varied, but on average 1 in 3 resulted in a strike. Going into 2022, it will be interesting to see if that success in baiting batters can prove sustainable or not.