Requiem for the knuckleball
There are a lot of uniquely fun things about baseball, but my favorite is the knuckleball.
The knuckleball breaks the key rules of pitching. Pitching well is about throwing fast, with a lot of spin, to precise locations, and mixing different pitches so the hitter doesn’t know what’s coming. A knuckleball is thrown slowly, with no spin, without knowing where it’s going, and rarely throwing other pitches — and, like a glitch in the game, it can still work.
For decades the knuckleball sat in an optimal range for fun: it was slightly less effective than normal pitches, so only a few pitchers threw it. If it was much less effective, nobody would ever throw it. And if it was more effective, most pitchers would throw it; it wouldn't be a novelty, it would just be what we think of as a normal pitch.
But that equilibrium was very fragile. Pitchers have used better technology and training methods to improve other pitches, leaving knuckleballs in the dust. (Another advantage of knuckleballs, that you can throw lots of them without tiring your arm, has been obsoleted by larger pitching staffs.) Only two knuckleballers have pitched in the majors this decade — Matt Waldron became the second last weekend, though even he threw only 20% knuckleballs — and both were demoted after one game.
This is my best attempt at calculating games pitched by a knuckleballer over time, and it's not encouraging:1
It's now been a full decade since a new knuckleballer made the majors and stuck around for more than a cup of coffee (Steven Wright, 2013). I’d bet at even odds that we never see another true knuckleballer have a substantial career again.2
How could the knuckleball be saved?
You could change the baseball itself. Knuckleballs are believed to move unpredictably because of how air moves across the ball’s seams (though apparently recent research has called that into question). Balls with higher seams would make the ball move more, so knuckleballs would be more effective; all other pitches would also become more effective, but perhaps the relative effect would favor knucklers.
However, don’t expect this to happen anytime soon. That change would reduce scoring, because all pitches would be harder to hit and because balls with higher seams wouldn’t travel as far. Hitters are already struggling to keep up with advances in pitching, so this seems like a non-starter.
Perhaps rules that severely limited the number of pitchers on a roster could help knuckleballers, who can pitch more innings without getting tired, but that seems unlikely to me, and in any case I don’t think there’s a strong pro-knuckleball lobby within the game.
So enjoy the last few knuckleballs you see, because they might be the last.
Knuckleballers identified from these two lists (1, 2); games pitched data from Fangraphs via the pybaseball package. This analysis isn’t perfect — it counts all games from “knuckleball pitchers”, even though some only switched to throwing knucklers later in their careers; on the other hand, it probably excludes some knuckleballers not on those two lists. Code for the full analysis is here.
Defined something like, pitching in 10+ games for multiple seasons