Strategy in the game of baseball has always evolved. As it pertains to pitcher usage, that evolution has been particularly swift over the last few decades. The four-man rotations of the 1950s and ’60s morphed into five-man groups in the ’70s and early ’80s. The 200-inning season has become increasingly rare in more recent seasons. Bullpens have become more specialized and diverse. In part due to the frequency of injury among pitchers in today’s game, teams and individual players have become more curious about prevention and efficiency.
My best guess, and I am hardly alone in this thinking this, is that the future of pitching-staff organization will eventually look much different. The structure will perhaps begin with tandem starters as part of four- or three-man rotations, or the rotations will be extended to six-man rotations. Perhaps there will be a battle of ideas and practices.
While 25-man rosters limit creativity, the new 10-day DL has allowed teams to experiment, and the Dodgers — as was forecast before the season by many — have best taken advantage of the truncated disabled list. The Dodgers entered the season with the greatest number of quality rotation options in the game, and that depth came with plenty of talented but risky options like Rich Hill, Scott Kazmir, and Brandon McCarthy. The Dodgers were the first natural test case and Sports Illustrated’s Tom Verducci procured an interesting quote from a Dodgers official while reporting on the club’s pitching strategy in 2017:
“There’s no team that has the kind of depth we do,” said one club source. “This team is built to win 95 games on the strength of depth carrying us over six months. We should get to 95 wins. But the year comes down to this: Clayton, Richie and Julio being healthy and ready to go to start playoff games. That’s it. So if it means they throw 170 innings instead of 200, that’s fine. They’ll actually be better for it.”
The disabled list has always been loosely governed. It’s always been used as a roster-manipulation tool in addition to a place to earnestly store injured players. And the 10-day DL certainly hasn’t ended that. As Eno Sarris noted last week, days spent on DL this year are up about 50% compared to the 2011-15 five-year average.
The positive spin on this situation is that maybe, once the dust settles, we’ll see some reduction in days lost. Players can take a 10-day breather in a situation where they would have previously attempted to return too early. Maybe a little bit of preventative rest will reduce the amount of catastrophic injury. Trips up, days down might be the slogan.
But that’s a maybe. In the meantime, we’re left a very real explosion of unavailable players. And a few teams that are perhaps superior at manipulating that rule change and new situation, whether due to resources or superior preparation.
And perhaps thanks to resources, and superior preparation, the Dodgers are leading the way in a practice that is surely to be copy catted — if it works. And it is working: the Dodgers lead baseball in fewest runs allowed per nine innings by starting pitchers (3.59 runs). They’re one of three teams with starting rotations allowing fewer than four runs per nine innings. Sure, having Clayton Kershaw helps. But the Dodgers ranked fifth last season in the majors in runs per game from starters (3.94), they allowed 3.67 in 2015, second in MLB, and 2014 (3.81), when the Dodgers ranked 10th and the game was still in a modern dead-ball era.
To date this season, the Dodgers have never pitched more effectively in the Andrew Friedman Era, and Kershaw is just getting his slider back.
And while a number of factors could be at work, this is also the first year of the 10-day DL, and the Dodgers are leading MLB in something else: the number of starts made by starters on more than four days of rest. Consider the leaders in “Long Days Rest”, tracked by Baseball Reference:
In 2014, the Dodgers ranked 11th in starts made on four days rest or more (85). The league average was 83. In 2015, Dodgers starters made 83 such starts. The league average that season was 84. Last season, the Dodgers jumped to second (103), and the league average jumped to 87. This season, they’re on pace to shatter that mark, with 130 such starts, while the league average will grow to near 90. In 1990, the league average was 55. In 1970, it was 45.
And the Dodgers aren’t just giving their starters more rest between starts; they’re also limiting them to 87 pitchers per start, tied for the third fewest in the majors.
The Dodgers represent an interesting experiment this year: never in the history of game will a staff have been as well rested throughout a season — or, presumably, entering the postseason.
While the Rich Hills and Kenta Maedas and Alex Woods are receiving more rest and trips to the DL, Kershaw is also receiving more time between starts. He’s already made six starts on more than four days rest, according to Baseball Reference. The most such starts he’s made in a season was 17 in 2011.
The Dodgers are creating a de facto six-man rotation. If this experiment works, if the team plans for it and has the depth for it, perhaps it will become a new model going forward.
In theory, a four- or three-man tandem rotation makes a lot of sense. Starting pitchers are less effective each time they work through a stating lineup, and a team can more often create platoon advantages by flipping to an opposite-handed tandem starter. (And in the NL, for as long as DH doesn’t exist, pitchers will less often bat.) But in speaking with executives, coaches, and players on the subject, the problems with the idea include how it would stretch a staff, the limited roster spots available, and when the plan blows up on a given day. If a pitcher fails to log his innings, it really stresses a roster.
So while the piggyback rotation is ideal in theory — and could perhaps become a reality with a combination of expended rosters and a 10-day DL — perhaps it’s the six-man rotation, or some sort of 10-day DL-enabled variety, that will eventually win the day.
If a fresh Dodgers rotation rolls through October, it will be a model to be copied.