Short Work: A Look at This Year’s Postseason Starting Pitching
The American League Championship Series between the Rays and Guardians was a study in contrasts when it came to starting pitching, particularly as the series extended. Tampa Bay’s starters lasted five or more innings in all but one of the six games, with Drew Rasmussen and Freddy Peralta combining for three starts of at least six innings; the former went seven innings and the latter 5 1/3 in the the final two games. Cleveland’s starters reached the five-inning mark just twice and never went deeper than that; over the final three games, the team’s starters — ace Parker Messick and reliever Shawn Armstrong, who was deputized to open Games 4 and 6 — totaled just four innings. Given that imbalance, it was hardly a surprise that the Rays rallied from a 2-1 deficit to win those three and take the series.
The World Series might present a similar contrast. The Rays and the Yankees are the only teams whose starters have averaged at least five innings per turn during this postseason, and Tampa Bay knocked off New York in a Division Series sweep earlier this month. Meanwhile, the Brewers are one of five teams whose starters have averaged less than four innings per turn; only twice did a Milwaukee starter even last five innings, with Logan Henderson reaching that mark in Game 2 of the NLDS against the Padres and Robert Gasser doing so in Game 3 of the NLCS against the Dodgers. Jacob Misiorowski — the likely NL Cy Young winner, who even before the playoffs had already exceeded his total workload from last year by more than 30 innings — burned through 88 pitches to complete four innings in his NLDS start and 94 to go 4 1/3 in his NLCS start. For the Brewers and Rays, those divergent approaches have worked. Tampa Bay went a combined 7-2 in the ALDS and ALCS; Milwaukee, 7-1 in the NLDS and NLCS.
After a season in which he ranked third in the AL in FIP (2.81) and WAR (5.0) and fourth in ERA (2.67), Rasmussen began his postseason by no-hitting the Yankees for 7 2/3 innings in the Division Series opener. In three starts totaling 21 innings, he’s delivered a 1.29 ERA and a 2.58 FIP. Peralta, who struggled during his four months as a Met before righting himself somewhat as a Ray following a deadline deal, has put up a 1.13 ERA and a 3.72 FIP in 16 innings as a starter, on top of which he added two-thirds of an inning out of the bullpen in ALCS Game 4; that cherry on top of his two strong starts helped him win series MVP honors. While their third and fourth starters, Nick Martinez and Griffin Jax, haven’t been as effective, the Rays’ postseason rotation ERA is good enough to crack the top 10 dating back to the start of division play in 1969:
Brooklyn-based Jay Jaffe is a senior writer for FanGraphs, the author of The Cooperstown Casebook (Thomas Dunne Books, 2017) and the creator of the JAWS (Jaffe WAR Score) metric for Hall of Fame analysis. He founded the Futility Infielder website (2001), was a columnist for Baseball Prospectus (2005-2012) and a contributing writer for Sports Illustrated (2012-2018). He has been a recurring guest on MLB Network and a member of the BBWAA since 2011, and a Hall of Fame voter since 2021. Follow him on BlueSky @jayjaffe.bsky.social.
I don’t think MLB cares, but if they want starters to work deeper into games then getting rid of some of the off days is the easiest way to encourage that. The equation for the manager is always “is the starter still my best option?” When the entire bullpen is rested pretty much every game, the answer is usually “no” once you get to the third time through the order, which is usually around the 5th inning. A playoff team will have three or four guys who can cover 4 or 5 innings if they are fresh. If some of those guys are unavailable, that math changes.
It would be interesting to know how much these figures would change if the pitchers chosen for inclusion were the guy who threw the MOST innings in each game, rather than the guy who threw the first pitch. Removing an opener after one or two innings isn’t necessarily an indication that the plan hasn’t worked: it IS the plan. So it feels like potentially some of the length-type stats presented here conflate unsuccessful starts by true starters with openers doing exactly what they were intended to do?
I spent a couple hours this morning poring over a log of every 2026 postseason appearance to date in an effort along these lines, trying to classify each start by whether it was a true start (taken by a player who served as a starter during the regular season), an opener/bulk game (taken by a player who was generally reliever during the regular season OR was clearly not going to stick around long regardless of his regular season usage pattern), or a bullpen game.
There’s a fair bit of gray area, much of it caused by the Braves’ starts taken by Ray Kerr and Dylan Dodd, both clearly relievers during the regular season and with three of the five longest starts; I counted them in the general opener/bullpen game camp, with AJ Smith-Shawver credited as the bulk guy in Dodd’s start and Grant Holmes as the bulk guy in Kerr’s first start, as I believe he’d have entered earlier had Kerr not been absolutely cruising. I did not count Didier Fuentes as the bulk in Kerr’s first start because he didn’t enter until the 6th inning and came within one out of closing.
(There’s room to quibble with those choices, particularly if I expand the study to include prior years for comparison because I think it might be helpful to have more objective criteria but i’m not sure “guy who threw the most inning,” as you suggest, is the exact way I’d go.)
With that, for the 70 games through the LCS round, my tally was 57 true starts (including some early exits by Hunter Brown, Aaron Nola, and Parker Messick), 13 Opener/Bullpen starts, and 9 bulk guys following in their wake (the other four games were “true bullpen” games).
The True Starts group averaged 4.72 innings, with a 3.25 ERA and 3.43 FIP. The Opener/Bullpen group averaged 1.69 innings, with a 1.64 ERA and 3.33 FIP. The Bulk group averaged 3.22 innings, with a 4.34 ERA and 4.24 FIP. Considering the True Starters and Bulks together yields an average of 4.52 innings with a 3.35 ERA and 3.51 FIP — still quite effective. Alas I don’t have a basis for recent comparisons as this was time-consuming enough to tease all of this out, but if the flow of the series dictates, I may go as far back as 2020 to see what we can learn.