A little more than a decade ago, a pitcher by the name of R.A. Dickey threw a kind of knuckleball that baseball had essentially never seen. Now Roki Sasaki might be taking it a step further.
Rather than sitting in the 60s or low 70s, Dickey threw his knuckleball upwards of 80 mph — and for strikes. When the “angry knuckler” worked, it confounded hitters so much that it won Dickey the 2012 Cy Young Award, making him the only knuckleballer to ever win pitching’s highest honor.
We are not saying that Sasaki is throwing a knuckleball — but the reality isn’t as far away as you would think. Sasaki is throwing a pitch that, like a knuckleball, can break in multiple directions pitch-to-pitch. And he’s doing it by throwing the ball with less spin than any non-knuckleball pitch in baseball.
We’re talking about Sasaki’s splitter, a pitch that is absolutely living up to its reputation as one of the best in the world in spring training so far.
Sasaki took the mound for the Los Angeles Dodgers for the second time in Arizona on Tuesday and essentially did what he did in his first outing, throwing four scoreless innings and 41 pitches while striking out two and allowing one hit against the Cleveland Guardians.
In his first start, Sasaki threw his splitter 18 times, generated eight swings and got seven whiffs, per Baseball Savant. In his second start, he threw the splitter 11 times, generated four swings and got three whiffs. Ten whiffs on 12 swings (an 83.3% whiff rate) is a comically dominant number that is guaranteed to go down once the games count, but it also reflects that not a single person knows what to do with this pitch yet — not even MLB’s cameras.
Roki Sasaki’s splitter is not of this earth
In both of Sasaki’s starts, MLB’s Statcast system — an advanced pitch-tracking system that records velocity, movement and spin rate, among other things — has repeatedly failed to immediately identify Sasaki’s splitter from his other pitches, most notably his slider.
In theory, it’s easy to tell which pitch is a splitter from the extremely low spin rate, but the challenge is that this thing’s movement is so varied that the velocity and movement profile bleeds into slider, changeup and curveball territory. Baseball Savant initially identified Sasaki’s splitters as all three of those other offerings before correcting midgame.
Here are back-to-back splitters Sasaki threw at poor Tyler Freeman, the first of which broke eight inches glove-side for a swinging strike — slider territory — and the second of which broke only one inch for a called strike. That is not a normal delta between pitches, and it gets more ludicrous when you learn that one batter earlier, he threw another splitter that broke eight inches arm-side.
That’s not normal! The only pitch that is supposed to be able to break eight inches one way, then eight inches the opposite is a knuckleball, and Sasaki has managed to insert that into a mid-80s, controllable package. In two games so far, it looks like a unicorn pitch and a supremely weird one at that.
The reason for Sasaki’s unbelievable movement is mostly the lack of spin. He averaged a comically slow 518 rpm on the splitter in his previous start and then 575 on Tuesday. For perspective, the median splitter rpm for all pitches last year was 1,326 and the lowest rpm on a splitter in a single game last season was 577 rpm (Emmanuel Ramirez of the Toronto Blue Jays). Meanwhile, Matt Waldron’s knuckleball averaged 244 rpm in 2024. Sasaki’s splitter is closer in spin to a knuckleball than the average MLB splitter.
Even compared to other spitters, Sasaki’s is atypical in its movement. In both his starts, its average vertical drop was larger than that of any splitter in MLB last season, and again, it broke in multiple directions pitch-to-pitch. Compare that to Ramirez, whose splitter last year moved consistently to his arm side like, you know, pretty much every other MLB splitter.
What we have here is a pitch that, even if identified by batter in mid-air, they still don’t know where it’s going.
Roki Sasaki pretty much has no ceiling
One of the main criticisms of Sasaki as a prospect before he signed with the Dodgers was that he was essentially a two-pitch pitcher with his fastball and splitter — he made some progress with a slider last year — and it’s difficult to succeed as one of those in MLB.
That outlook underestimated Sasaki, who, it was announced Tuesday, will make his MLB debut when he starts Game 2 of the Tokyo Series next week. Rather than knock him as a conventional pitcher with too few pitches, it might make more sense to think of him as a knuckleballer who can also mix in an upper-90s fastball. He throws a pitch with so many potential destinations that hitters have to be waiting to have a chance, and he also happens to be perhaps the hardest-throwing pitcher on the planet. It’s like if Dickey had Craig Kimbrel’s fastball.
That’s why every team wanted Sasaki this winter and why he probably would’ve received upward of $200 million if he’d been allowed to be a true free agent, like Dodgers and Japan teammate Yoshinobu Yamamoto was last season.
If there’s any reason for pessimism after Tuesday, it’s that Sasaki averaged only 96.3 mph with his fastball, nearly two full ticks below his average last week. There could be a reason for that — innocuous or concerning — but it was also just one start in which Sasaki was clearly still effective.
