I've mentioned Drew Pomeranz more than is probably necessary over the last several posts, but there's something so compelling about the potential of a starter who converts to a reliever. Pomeranz’ transition isn't a new phenomenon. Andrew Miller went from being a bad starter to a shutdown reliever, as did Wade Davis. The logic is pretty simple: moving to the bullpen allows a pitcher to throw closer to their max-effort, and they don't need to mix their pitches as much, so they can heavily rely on their best pitches. According to research by FiveThirtyEight's Nate Silver, the average starter working in relief (for an appearance of five or fewer batters) sees their fastball velocity increase from 91.7 to 93.6 MPH, and their strikeout rate increase from 19.9% to 23.9%.
One reason more starters don't move to the bullpen is that a league-average starter is more valuable by WAR than a good reliever. Josh Hader was the third-best reliever by fWAR in 2019, but he was only worth 2.3 wins. Starters who were worth around that much include Reynaldo López, Sandy Alcantara, and Brad Keller. In looking for the next Drew Pomeranz, I came up with a few general criteria:
They should have been worth less than two WAR.
They should have at least one really good pitch that they don't use as much as they could if they were a reliever.
They shouldn't be highly-valued, either as a prospect or by their contract. Jon Lester could make a good reliever, but you're not going to move him to the bullpen.
Using these criteria and Baseball Savant, I've identified two starting pitchers from 2019 who could be worth trying as relievers in 2020.
Junis has spent the last two seasons in the Royals’ rotation. By fWAR, 2019 was actually his best season. Unfortunately, he still had a 4.82 FIP and a 5.24 ERA, and it doesn't take much research to figure out why: Junis’ fastball is really bad. He allowed the eighth-highest xWOBA on his four-seam fastball of any starter in 2019 at .444, and it has really low spin while coming in at 91.7 MPH.
However, Junis has a secret weapon in his slider. According to Baseball Savant, Junis’ slider has a movement profile similar to that of Mike Clevinger and Trevor Bauer, which is good company for slider movement. He gets the 25th-most active spin percentage on the pitch, and allowed just a .217 xWOBA on his slider in 2019, the 16th-lowest among major league starters (tied with Gerrit Cole).
Despite this, Junis only threw his slider 29.6% of the time, 24th-most among starters. Chris Sale threw it 38.4% of the time, and six relievers threw their sliders more than 60% of the time. If Junis threw his best pitch as much as some relievers and his fastball played up a bit, it's not hard to imagine him playing a role in the bullpen of a contending team.
Eickhoff was part of the Phillies’ return in the Cole Hamels trade. He had a good 2015, and a pretty good 2016, but had a 5.71 ERA in 2019 and, as a result, is currently a free-agent at the age of 29.
Eickhoff had a good curveball this year by Fangraphs’ pitch values, and the underlying Statcast numbers back that up. He was in the 82nd percentile in curveball spin and in the 84th percentile in fastball spin. However, his four-seam fastball was crushed in 2019—he allowed the tenth-highest xWOBA on his four-seam fastball of any starter at .439. In contrast, he allowed just a .239 xWOBA on his curveball, 37th-lowest in the majors, right below Mike Clevinger and Yu Darvish.
Even though Eickhoff threw his fastball less than in 2018, he still threw it 36.7% of the time, more than any other pitch. Meanwhile, while his curveball rate of 31.5% was high, 11 curveball-heavy relievers threw their curveball at least 40% of the time.
Eickhoff has lost over 2 MPH on his fastball since 2015, and he could make up most of that if he got the standard velocity boost from moving to the bullpen. His fastball still wouldn't be great in all likelihood, but paired with his good curveball, I think Eickhoff could be an effective reliever.
Other starters I considered for this list but decided not to include:
Michael Wacha had a very bad year as a starter, but still has a great changeup. However, he doesn't seem to have much to pair it with—he allowed a .400 xWOBA or higher on his curve, cutter, and four-seamer in 2019.
Julio Teheran has good spin numbers on both his fastball and curveball, but he was successful enough last year that I would expect him to get a starting job somewhere.
Chase Anderson has a couple of pitches that rate quite well, but he was also good enough as a starter last year to continue as one.
Dylan Bundy has struggled off and on and hasn't met his top-prospect potential. While I think there are underlying indicators that show he could be better next season, it's hard to imagine him providing more WAR as a reliever than he did as a starter this season (2.5).
Several of Chris Bassitt‘s pitches also have great underlying spin and movement numbers, and I think he'll be better as a starter next year anyway, or a great reliever, especially if he continues the trend of increasing his four-seamer usage (.277 xWOBA allowed) and decreasing his sinker usage (.341 xWOBA allowed).
I believe most of these starters could be good relievers (perhaps with the exception of Wacha), but according to traditional WAR models it's hard for me to see them providing more value as a reliever than as a starter. It's important to remember that even though using starters out of the bullpen has become a popular playoff strategy, it's not guaranteed to make a pitcher more valuable over the course of a full season.
P.S. Here's something interesting—the biggest moves of the offseason have been made by the Braves and White Sox, and all five have been announced directly by the teams themselves instead of through reporters. I'll be interested to see if this is a new trend that will continue.