Could a Shortened Season Resurrect the Four-Man Rotation?

As strange as it sounds right now, a day will come when baseball people start thinking about how to win baseball games again. That might not be until spring 2021, if COVID-19 forces Major League Baseball to sit out an entire calendar year. But there is still a chance it happens sometime in the next couple of months, and if that’s the case, teams will be preparing for a season unlike any they’ve played before. In addition to all of the structural changes that might be necessary for games to proceed, the simple fact of the season being considerably shorter than normal could change the way teams approach the games. In light of this, one strategy for teams to consider is a return to the four-man starting rotation.

Four-man rotations — or at least, the concept of throwing your best arms on three or fewer days rest — used to be fairly commonplace in the majors. In a piece by Russell A. Carleton at Baseball Prospectus from 2013, he found that around 40% of starts occurred on what we’d now call “short” rest as recently as the early 1970s. Soon after reaching that recent peak, however, the practice nosedived, with short rest starts happening less than 10% of the time by 1984 and continuing to free-fall until reaching a point of near-extinction over the last two decades.

The point of Carleton’s study was not to find out when the four-man rotation died out, but why it did. To that end, he came up with little statistical reasoning. Pitchers who threw on short rest did not seem to perform any worse than they’d be expected to on full rest, nor did any cumulative effects seem to wear them down faster over the course of a full season. This was true regardless of the time period, too, which means it didn’t seem to have anything to do with whether a pitcher was more or less conditioned for it. The best explanation for why teams broadly and swiftly shifted to a five-man rotation was simply fear of injuring arms, and hoping that more rest might prevent that.

Carleton’s piece didn’t take the extra step of actively calling for the four-man rotation to return, but plenty of others have. Major league teams, however, have been taking steps to limit the responsibilities given to starting pitchers, not expand them. In 2014, David Price led the majors with 248.1 innings pitched, while 34 pitchers reached 200 innings, and 65 reached 180. In 2019, Justin Verlander’s league-leading innings total was just 223, with 15 pitchers reaching 200 innings, and 33 pitchers reaching 180 innings. Over a full season, a four-man rotation could add as many as eight starts to a pitcher’s workload, amounting to 40-50 extra innings of work. It would be rather shocking to see a team take that kind of step over 162 games.

But if the season is cut considerably short? That might be a different story. As Dan Szymborski has covered here before, the gulf between baseball’s good and bad teams is narrowed with each game taken away from the regular season. With a full 162-game slate, we can be reasonably certain of a few different things. The Dodgers, Astros, and Yankees are near-locks to be playoff teams, and about a third of baseball has virtually zero chance to win the World Series. Cut the length of the season in half, and chaos ensues. A team like the Rangers goes from having almost no route to the playoffs to basically a one-in-five shot, while the powerhouse Dodgers are missing the playoffs about three out of every 10 times.

If there is a 2020 season, we know it is going to be much shorter than normal, which means each individual game is going to matter a hell of a lot more than it previously would. Under those constraints, it could get more difficult for a team to justify allowing fifth starters and swingmen to take games from their best starters. If the Nationals have just 81 games to outrace Atlanta, New York, and Philadelphia in the NL East, what’s the point in penciling in Joe Ross or Austin Voth for the same 16 starts you’re going to give Max Scherzer, Stephen Strasburg, and Patrick Corbin when you could instead ensure those three are all starting 20 games? How much could the Rangers help their playoff odds if they pitched Mike Minor, Lance Lynn, Kyle Gibson and Corey Kluber 20 times each instead of taking four starts from each of them to give to Jordan Lyles?

To figure out who might benefit most from this, I conducted a rough study. I took the presumed starting rotation in each team’s Depth Charts projections, and cut their projected WAR totals in half, to adjust from a 162-game schedule to an 81-game version. After adding up the WAR totals of those top five starters on each team, I removed the fifth starter, and increased the first four starters’ WAR by 25%, to get their total after four additional starts. These were the biggest improvements that came from dropping the fifth starter:

Now, this study presupposes the four best guys would make all 20 (or, in the case of the ace, 21) starts, which we all know wouldn’t actually happen. Any version of a shortened schedule seems likely to include doubleheaders, in which case an additional starter would be needed to pitch on those days. Guys also move in and out of the rotation as they under/outperform expectations, and injuries could have an even larger impact than normal this year.

And yet, this table almost certainly undersells what some teams stand to gain from this kind of move. Because while this study assumes the four best pitchers are always the ones taking the mound, it also assumes that the pitcher they’re taking starts from is the fifth-best pitcher on the staff. As injuries mount, however, that won’t be the case. Your third, fifth, and sixth best pitchers could be hurt, and suddenly the difference between a four- or five-man rotation could be the difference between throwing your best pitcher on a certain day or throwing your seventh, eighth, or ninth-best pitcher.

If the Cincinnati Reds, for example, suffer two injuries in their rotation, and those injuries are to Trevor Bauer and Luis Castillo, the team is probably left in a pretty bad spot regardless of how they choose to patch those holes. But if they would lose Wade Miley and Anthony DeSclafani, then shortening the rotation to four pitchers would give extra starts to Bauer, Castillo and Sonny Gray, instead of asking unproven arms like José De León or Lucas Sims to carry a significant burden.

Because of the significant difference in value between a team’s one or two or three best pitchers and those further down the depth chart, it seems probable that the number of wins gained wouldn’t just be one or two in some cases, but three or four. The table also potentially undersells the gain in wins because it completely removes the fifth starter from the equation. On many teams, that wouldn’t be the case. The fifth starter would typically hang on in the bullpen, where he would likely add some value, and in some cases, add even more value than he would as a rotation piece.

Consider, if you will, the Chicago Cubs. Their projected fifth starter, at the moment, is Tyler Chatwood, whose most recent extended venture as a starter in 2018 resulted in him walking more than eight batters per inning, notching a 5.60 FIP and 0.4 wins below replacement level. Then, pitching mostly out of the bullpen in 2019, he lowered his FIP to 4.28, and improved his WAR by one full win. Our Depth Charts projects him to be worth 1.6 WAR as a starter in 2020, but you’ll forgive me for thinking that might be a tad optimistic. Would the Cubs not be much better served, in a tight NL Central race, handing Chatwood’s starts over to Yu Darvish, Jose Quintana, Jon Lester, and Kyle Hendricks for a season that might last all of three months?

Cutting one rotation spot also makes sense from an overall roster-building standpoint. As MLB has tried to sort out what this season could look like, one thing that has come up rather consistently has been the idea of teams needing larger rosters due to what could be an increased number of doubleheaders, and the possible effects of playing neutral site games in a place like Arizona, where mid-summer heat will cause players to wear down much more quickly. Because starting pitchers virtually never appear out of the bullpen on a rest day, that means each roster spot used on a starting pitcher is essentially dead weight on days that player isn’t starting. Eliminating one of those rotation spots would free up space for an extra arm the team could actually use whenever it wanted to, instead of just once every five days.

It isn’t in the best interests of every team to use a four-man rotation. The Dodgers have seven or eight starters such that the team shouldn’t necessarily need to choose a more exclusive group than is necessary. A team like the Tigers, which is rebuilding and has a large group of arms who would benefit from major league experience, also doesn’t have much incentive to shorten the list of guys who are getting starting opportunities.

But if we do see the return of baseball in 2020, a team that approaches the season the exact same way it would if all 162 games were being played may be making a mistake. Winning an unconventional season could mean taking unconventional measures. If a four-man rotation this year would still mean a pitcher throwing fewer innings than he would in a typical season, and there is no clear added risk of arm injuries or loss of effectiveness, as Carleton once theorized, then you’ll never get a better opportunity to make the most of great pitching.