After their surprise run to the playoffs this year, the Mets have a lot of work to do this offseason to follow up on their unexpected success. From their rotation alone, they lost Sean Manaea, Luis Severino, and Jose Quintana to free agency — that trio combined to make 94 starts in 2024 — leaving Kodai Senga and David Peterson as the lone holdovers. Earlier this week, New York signed Frankie Montas to a two-year deal to begin restocking its pitching staff. On Friday, the next domino fell: Clay Holmes, erstwhile Yankees closer, agreed to a three-year contract worth $38 million with an opt out after the second year.
With an extensive résumé covering high-leverage innings a borough over, you might expect the Mets to plug Holmes into the bullpen behind Edwin Díaz and call it good. Instead, they’re planning on transitioning him to the starting rotation. Holmes has all of four major league starts under his belt and all of them came during his rookie campaign back in 2018. He worked extensively as a starter in the minor leagues but the year he made his major league debut was the last time he threw more than 100 innings in a season.
In their never-ending search for effective starting pitching, teams have turned their gaze to the bullpen over the past few years. Just this season, Garrett Crochet, Reynaldo López, and Jordan Hicks made the jump from the bullpen to the rotation. Last year, Seth Lugo, Michael King, and Zack Littell made a similar shift. It doesn’t always work; for every success story like Lugo’s or López’s, there’s a cautionary tale like that of Hicks or A.J. Puk. Even Crochet, for all his early success, seemed to wear down towards the end of the season, despite some extremely careful handling by the White Sox given the left-hander’s injury history.
Still, even if success in the rotation isn’t guaranteed, there’s an easy fallback option should Holmes struggle in longer outings: send him back to the bullpen as a setup man. That’s what the Marlins did with Puk after his four disastrous starts in April, and he thrived once he was returned to his natural habitat. The Mets have plenty of need for high-leverage bullpen arms, with Díaz the only truly proven option currently on their roster. Of course, if Holmes can’t hack it as a starter and is shifted back to the bullpen, it would leave a hole in their rotation, but it still limits the downside of the signing considerably.
If the rotation is Plan A for Holmes, the question is whether he’ll thrive given a much larger workload. Generally, to have sustained success as a starter, a pitcher needs a strong fastball, decent command, and at least a pitch or two to handle opposite-handed batters. A deep repertoire isn’t necessarily a requirement — López is essentially a two-pitch pitcher — but a plan to combat the platoon advantage is necessary to work through a lineup multiple times.
There’s both good and bad news for Holmes on that front. He has a three-pitch mix, with a bowling ball sinker making up the majority of his pitches and two distinct styles of slider — a harder gyro slider and a slower, Yankees-style sweeper — splitting the balance equally. Way back in his distant past, Holmes threw a four-seamer, a changeup, and a curveball, but he ditched those pitches once he shifted to the bullpen full-time. Over his career, left-handed batters have produced a .314 wOBA against him, while he’s held right-handed batters to just a .267 wOBA. When you break down Holmes’ arsenal by pitch type, you learn a bit more about his platoon splits:
Clay Holmes Platoon Splits by Pitch Type
His hard slider is essentially a neutral pitch; Holmes is effective with it no matter who he’s facing. More importantly, his sinker doesn’t possess that big of a platoon split — just 23 points, which is pretty low for a pitch that generally runs a significant split. That can probably be attributed to the crazy amount of vertical movement he’s able to generate with the pitch; when you’re pounding a pitch into the ground, it’s hard to have success against it no matter which side of the plate you’re standing on. There are some yellow flags with Holmes’ sweeper — it has a 102-point split — but that isn’t so surprising. Since the sweeper gained popularity over the last few years, we’ve realized that it’s generally much more effective against same-handed batters.
Whether his slider is enough to keep left-handed batters at bay is something the Mets will need to figure out, but Holmes has a solid foundation to work from. And if there’s an avenue to expand his repertoire, we may have already seen a glimpse of it during the postseason:
That’s Holmes striking out Vinnie Pasquantino with a four-seam fastball at the belt in Game 2 of the ALDS. You could count on one hand the number of four-seamers he had thrown over the last four years, but he fired 10 during the playoffs and earned two whiffs on five swings against the pitch. As you can see in the clip above, Pasquantino swings well under the pitch even though it was pretty center cut. He was likely expecting a ton of drop from a pitch in that velocity band, and the four-seamer simply didn’t dive like Holmes’ sinker would have. Using that riding four-seamer as a change-of-pace pitch to work off his sinker could give Holmes an extra wrinkle to help him as he works through a lineup multiple times. Reintroducing his changeup or adding a splitter could also pay dividends, though it’s unclear just how much he’s willing to tinker with his pitch mix at this point.
Of course, just taking Holmes’ repertoire and assuming it will stay the same as a starter isn’t realistic. When Eno Sarris examined the relievers who were making the transition to starting, he found that their repertoires lost around eight points of Stuff+ overall, with their fastballs losing around 12 points of Stuff+. More tangibly, López lost three ticks off his entire repertoire, while Hicks lost five. If we assume Holmes experiences a similar loss of velocity (and spin, since those two characteristics are linked), what would the results on his pitches look like? I queried Baseball Savant to see how his sinker performed when thrown 94 mph or slower (it averaged 96.6 mph this year) and and how his slider performed when thrown 86 mph or slower (it averaged 87.1 mph):
Hypothetical Starting Pitcher Clay Holmes
Unsurprisingly, both pitches performed worse when thrown slower. Thankfully, Holmes’ slider was still effective at generating swings and misses and his sinker was still pounded into the ground at those lower speeds. Knocking less than three ticks off his sinker and less than two ticks off his slider is pretty generous, but the sample sizes at even lower velocity thresholds just weren’t big enough to warrant inclusion. Suffice it to say, if he suffers a Hicks-ian level of stuff deterioration, his stay in the rotation might be short lived.
Whether or not he expands his repertoire, Holmes looks like he has many of the tools needed to succeed in the rotation. Of course, he still needs to show he can handle a starter’s workload, but Holmes essentially has two seasons to prove that he can hack it in longer outings — the opt out after the second year of the deal gives him the ability to test the market again if things go swimmingly. And if things don’t work out in the rotation, his familiar role as a high-leverage reliever isn’t the worst fallback option.
