Kiyoshi Mio-Imagn Images

If you were only looking at the top-line numbers that Roki Sasaki has posted through 10 starts — a 4.59 ERA and a 5.04 FIP in 51 innings — you could be forgiven for thinking that the 24-year-old righty had made little progress since last year’s abbreviated rookie season. At times, it has seemed as though he might be better off ironing out his mechanics and approach in Triple-A or in the bullpen, where he found some success last fall after missing four and a half months with a shoulder impingement. While Sasaki’s early-season starts brought to mind last year’s struggles, a little over a month ago he made a change to his repertoire, adding a second offspeed pitch to his mix. Since then, he’s pitched more effectively thanks not only to the new offering, but also to better command and velocity.

Sasaki’s improvement has come at a particularly opportune time for the Dodgers, as they’re once again carrying on without either Tyler Glasnow or Blake Snell. Glasnow left his May 6 start after just one inning due to back spasms and still hasn’t been cleared to throw off a mound yet, while Snell didn’t make his season debut until May 9 due to shoulder soreness, then lasted just three innings before discomfort in his elbow forced him from his start. He was diagnosed with loose bodies in the elbow and underwent surgery using the same NanoNeedle Scope 2.0 procedure that Dr. Neal ElAttrache had just used on Tarik Skubal. The new version of the surgery is aimed at accelerating recovery time, but Snell has been moved to the 60-day injured list nonetheless, and can’t return until early July. Increasingly, it appears Glasnow won’t be back before July either.

Sasaki’s new pitch is a splitter, but it’s not the same splitter he threw last season, and it’s not entirely new; it bears more resemblance to the one he threw in NPB with the Chiba Lotte Marines and in the 2023 World Baseball Classic for Team Japan than it does to its immediate predecessor. Statcast has redefined the offspeed pitch he threw last year as a forkball and is tracking the two pitches separately. Sasaki throws his splitter about five miles an hour faster than his forkball, and relatively speaking, it gets a bit more rise and a lot more arm-side run. Since introducing the new pitch — which I’ll get into more below — he’s lasted at least five innings in all six of his starts, something he’d done in just four out of eight starts last year and one out of his first four this year.

Here’s a look at Sasaki’s numbers before and after adopting the new pitch, and how they compare to last season:

While those stats won’t put Sasaki in the Cy Young discussion, they’re clearly a step forward, with his strikeout-walk differential more than tripling relative to his first four starts of the season. His ERA since then has dropped by 2.33 runs per nine, while his xERA and FIP have fallen almost as much. In just about any other situation, the pitcher who put up those first two lines would have been sent to Triple-A to regroup. But the Dodgers wooed Sasaki on the basis of both their ongoing success and their commitment to furthering his development, since they couldn’t just blow him away with dollars because his age made him subject to international signing rules. They’ve maintained that he would be better off working with their major league staff instead of in the minors, and have resisted calls to return him to the bullpen or farm him out in the event of a roster crunch. Now their patience is paying off.

On Saturday night against the Phillies in Los Angeles, Sasaki pitched one of his best games of his brief stateside career. Through his first five innings, he walked just one batter (Bryce Harper with two outs in the first) and surrendered just one hit, a second-inning solo home run by Alec Bohm. He began the sixth by striking out Justin Crawford swinging at a 99-mph fastball, his seventh punchout of the night, but then allowed back-to-back singles to Kyle Schwarber, whom he’d already struck out twice, and Trea Turner. Schwarber’s hit was a 44.2-mph check-swing dribbler that eluded the grasp of third baseman Santiago Espinal, though Turner’s single was a clean one, and it ended Sasaki’s night after 84 pitches. Reliever Alex Vesia walked Harper to load the bases, but escaped the jam with a strikeout and a groundout. Alas, the Dodgers ended up losing, 4-3, as Tanner Scott allowed three runs in the eighth inning, ending a streak of 12 consecutive scoreless appearances.

Sasaki’s 18 swinging strikes against the Phillies matched his career high, set on May 17 against the Angels. Three of them came on the splitter (27.3% whiff rate), seven on the slider (50.0% whiff rate), and eight on his four-seam fastball (32.0% whiff rate). His fastball was just as much a story on that particular night, as he hit 100 mph with it twice. While he reached 100 mph 16 times coming out of the bullpen last October — during which he allowed one run in 10 2/3 innings and notched three saves — the only previous start in which he’d done so was his major league debut against the Cubs on March 19, 2025 in Tokyo, where he may have had a wee bit of extra adrenaline. Sasaki threw his seven fastest pitches of the season against the Phillies, all 99.5 mph or higher, and 14 of his 22 fastest, all 99 mph or higher. Four of those produced strike three, with Adolís Garcia, J.T. Realmuto, and Schwarber joining Crawford in that club.

Sasaki has thrown 17 of his 22 pitches of 99 mph or higher over a three-start span, underscoring the fact that he’s not only ironed out the mechanical difficulties that plagued him last season and improved his command, but that he’s in better shape, too. “It was more about my body’s functionality and overall conditioning,” he said through an interpreter after Saturday’s start. “I think those adjustments ended up helping.”

The new pitch came about when Sasaki, finally in a good place with his mechanics, told Dodgers pitching coach Mark Prior and assistant Connor McGuiness that he wanted to throw his splitter harder. During a bullpen session in San Francisco, McGuiness suggested Sasaki use a more traditional splitter grip “where the ball comes more off the end of his fingertips, rather than being lodged firmly between his middle and index fingers,” as the California Post’s Jack Harris described it.

“He threw one of them, and we were like, ‘That’s pretty gross,’” said Prior. “And then he threw a few more, and they were pretty good.”

Prior and McGuiness gave Sasaki the green light to use the new pitch in a game, and he went to town during his April 25 start against the Cubs, throwing it 48 times out of 99 pitches while temporarily shelving the forkball; the splitter count was so high because Sasaki felt the Cubs were sitting on his fastball. His outing wasn’t a great one, as he served up three homers and allowed four runs — both season highs — in five innings, but the last of those home runs came after the Dodgers had put up a sixth-run fourth inning, and even after that, an 8-4 lead allowed Sasaki to pitch into the sixth inning for the first time all season, though he didn’t record an out.

Sasaki restored the forkball to the mix for his May 2 start against the Cardinals, throwing nine of those and 41 splitters out of 104 pitches, his highest stateside pitch count. He allowed three runs in six innings in that one, two of them on Jordan Walker’s two-run homer… off the forkball (d’oh!) while walking two and striking out four. His pitch mix has varied widely from start to start. Here’s his Statcast usage graph, compressed horizontally to eliminate last year’s long injury outage, and with the color of the new pitch adjusted from a darker teal (too much like the forkball in its display) to navy blue:

The Dodgers like the new splitter because it tunnels similarly with Sasaki’s four-seamer before breaking late. Here’s a shot from Statcast’s 3-D pitch tool, showing the shapes of each pitch while highlighting the commit point, the point at which the batter has to decide whether or not to swing. You can see it a bit better on the inset I created, and again I’ve shifted the colors to more clearly distinguish the splitter from the forkball. Its commit point is the one closest to the four-seamer; follow those two pitches to the batters box and you can see how far apart they end up and understand that this is a potential problem for hitters.

MLB.com’s David Adler recently summarized the differences between the two pitches:

The original, slower, extremely low-spin variety he started out with, which can move in any direction, like a knuckleball.

A new version… which is faster, higher-spin and has more traditional drop-off-the-table movement.

Here’s a comparison of their specifications, along with the rest in Sasaki’s repertoire, including the 2023 WBC version of his splitter:

I’ve highlighted the 2026 versions of the splitter and forkball for easier comparison of their velocity, spin rate, and movement. Sasaki’s current splitter has less spin and gets more arm-side movement than the one he threw in the 2023 WBC. He was throwing harder then in general — with an average four-seam velocity of 100.3 mph — and getting a lot more spin on all of his offerings. It’s also worth noting that Sasaki has scrapped last year’s sweeper in favor of a more traditional slider (which some have referred to as a cutter, but let’s not go there today).

Sasaki’s overall chase rate has improved from 27.0% to 34.7%, with the splitter having the highest chase rate of any of his offerings at 36.2%. Here’s a look at his results by pitch type before and after adding the new splitter:

Batters haven’t been able to do much when making contact with the splitter, and by just about every measure, they’ve also fared worse against the four-seamer and slider since its addition. The slugging percentage against the forkball is distorted by the aforementioned Walker home run, which was hit 110 mph but at just a 17-degree launch angle, and wouldn’t have been out at Dodger Stadium or 12 other existing parks; note the xSLG of just .301 against that pitch since it was demoted in the pecking order. One underrated improvement has been the low average launch angles of the forkball and slider. Sasaki’s groundball rate has improved from 40.7% to 43.6% with the change, while his average launch angle has dropped from 15.6 degrees to 12.3. His barrel rate has climbed from an already-high 9.1% to 12.8%, but that’s been less of a big deal thanks to the extra strikeouts and fewer walks.

These numbers don’t suggest Sasaki is ready to be mentioned in the same breath as Cristopher Sánchez or Jacob Misiorowski in an NL Cy Young discussion. Nor would he top the list if the Dodgers were fleshing out an October rotation with Snell and Glasnow available, as was the case last year after both missed significant time due to injuries. Sasaki isn’t even the only pleasant surprise within the rotation, as Justin Wrobleski has pitched to a 2.87 ERA and a 3.41 FIP while whispering sweet nothings about spreadsheets. Nonetheless, Sasaki’s better command and reconfigured repertoire have made for a most welcome development.