Mets' Devin Williams makes adjustments to repertoire in effort to change luck from Yankees' days

Mets relief pitcher Devin Williams throws during a spring training workout on Friday in Port St. Lucie, Fla. Credit: Newsday/Alejandra Villa Loarca
PORT ST. LUCIE, Fla. — There’s not much you can’t quantify in baseball.
There are ball-tracking systems and wearable biometric monitors, formulas that tell you what did happen, what should have happened and what has the greatest probability of happening in the future.
And still, one truism stands above the others. It comes to the fore when Mets pitching coach Justin Willard is asked about Devin Williams’ difficult 2025 season.
“He had some bad luck early on,” he said. “That’s baseball. It’s just part of the game we play.”
Williams may have had a nightmare season with the Yankees — or at least the first half of one — but there’s little doubt that baseball was unkind. He had a 4.79 ERA compared with a 3.02 expected ERA. His batting average on balls put in play was .296, which indicates a bit of bad luck, and players were able to touch up that vaunted “Airbender” changeup, even when he threw it outside of the strike zone.
But none of that really matters when you’re getting booed off the mound at Yankee Stadium or losing your closer role after years spent as one of the most dominant relievers in baseball. And so one of the focuses this year has been to coax “luck’’ in another direction: After abandoning his cutter last year, Williams plans to bring it back in 2026 and also add a traditional gyro slider, which he ditched for the harder cutter in 2022.
Adding both “is valuable in itself,” Williams said. “But I think it’s going to allow me to protect my changeup and fastball and give guys a different look and a different shape to look out for.”
It’s about more than just expanding his repertoire.
Ever since he allowed Pete Alonso’s go-ahead home run in Game 3 of the 2024 Wild Card Series, there’s been speculation that Williams has been tipping his pitches (last year he said he didn’t believe that to be the case).
Additionally, the arm angles for a fastball and a changeup differ. Although Willard said it’s usually small enough to be imperceptible, “sometimes he flirts with that bubble.”
“Generally, he’s in a really good spot,” Willard added, noting that there’s only so much guesswork to be done when a hitter has at least a 50% chance of getting it right.
Make no mistake, Williams’ “Airbender” is still a very good pitch — it boasts exceptional movement, both horizontal and vertical, with a high-spin screwball action that’s difficult to impact. But hitters adapt, and Williams has to do it, too.
“I think a little bit of it is just predictability,” Willard said. “I think when you have a pitch named after you — you have the Airbender, you have the ghost fork, you have these things — hitters are generally going to game-plan for that because it’s so good. It’s just understanding how do we build the arsenal to maximize that pitch? It’s still a really good pitch, but how do we complement the rest of the arsenal to make that even more effective?”
There’s a difference of more than 10 mph between Williams’ fastball and changeup; his cutter historically has been about 4 mph slower than the fastball.
So ideally hitters will “recognize that something hard [is coming] and think it’s fastball,” Willard said. “Then it comes in on their hands.
“Any doubt you can put in the hitters’ minds, they’re like, ‘What the hell? What was that? That’s not what I was expecting at all!’ That allows us to get in a better spot and continue to be on the attack.”
The cutter’s elevation also helps mask the changeup. The hope, Willard said, is that hitters say, ‘“Oh, OK, this is up. I think it’s going to be a changeup and it’s not. It’s breaking in on my hands,’ and I get weak contact for a ground ball.”
It’s a logical approach meant to tackle a sometimes illogical sport. Last year, it felt as if Williams couldn’t catch a break, so this year, he’s trying to make his own.



