What If a Baseball Team Used a Different Pitcher Every Inning?

Much ado has been made about how modern relief pitching has changed the game, some even going as far as to say it has broken baseball. We’ve seen bullpens getting the call earlier and earlier, but with how markedly better pitchers perform in one inning, who can blame them?

What if this was taken to an entirely new level? What if a team built its strategic foundation on the idea that it would send out a new pitcher for every inning? What might that look like?

Roster Construction

The rule of thumb for approaching the 25-man roster is to have five starters and seven relievers. Now we could certainly try to make this work with twelve pitchers, but seeing as I have no idea what toll this will take on their arms, we’re going to opt for 14. This leaves us room on the roster for a backup catcher, infielder, and outfielder. Among non-pitchers, positional versatility and durability will be especially prized.

As for the pitchers themselves, we have no need for aces or elite closers. Of course we want to get good pitchers, but I think we would be wise to save our payroll and spend heavily on batting. We will primarily be targeting promising youngsters before their first big contract, with an aging vet thrown in here or there if the price is right.

Concerning whether we should be targeting starters or relievers, I think it’s an open question. The workload demanded would be somewhere between what is expected of either. Relievers would be more used to coming in for brief stints, but seeing the success that starters have had in relief stints, it would be foolish to write them off entirely.

The other thing to consider is that pitchers with extreme situational strengths and weaknesses will shine in this system. If you’re bad against lefties, guess what? You don’t have to pitch against lefties anymore. This would allow us to find gems that simply wouldn’t work on other teams.

Rotation Management

Trying to rotate 14 pitchers an inning at one time is an unprecedented challenge, but not an impossible one. With a couple of guiding principles, we should be able to arrive at a logical approach:

• For every game, the manager should know what nine pitchers are available to him.
  • There will be five pitchers slated to rest each day.
• The workload of each pitcher should be as similar as possible.
  • The resting pitchers will be rotated through fairly evenly, resting at least once every three days.
• One of the resting pitchers should be available for a potential extra inning appearance.
  • Of the five resting pitchers, one should not have pitched both of the previous two games.
• The manager should have full control to change the order in which the pitchers appear.
  • If he wants a certain pitcher to face the top of the order their third time up, he should be able to move that pitcher from the sixth to the fifth or the seventh depending on how earlier innings went.
• Mid-inning pitching changes will not be allowed.
  • We’re already going to be on a tight leash for pace of play. Plus introducing countless more decision points is likely to cause overthinking. Keep it simple; trust the process.

With all this in mind, we should be able to put together a rotation. The foundation of this will be three four-man platoons that take turns pitching and resting along with two pitchers who alternate being slated to play and being on call for extra innings. Bearing in mind that the manager can rearrange the pitchers mid-game as much as he likes, here’s a quick look at what a six game rotation might look like:

What Would a Season of This Look Like?

Great question, thanks for asking. The real answer is I have no idea. If each pitcher shares the load evenly, we’re looking at each guy throwing 104 innings in 104 games. No pitcher has appeared in 104 games since Mike Marshall in 1974, so the scarcity of rest days may have a larger effect than we’re accounting for. We’re gonna give it a shot anyway.

Below are listed the top fourteen pitchers by 2019 Wins Above Replacement (WAR) per 200 innings pitched that are making less than $1M in 2020. To get their simulated WAR for a season under this system, I took their WAR/200 IP from last season, adjusted to 1/14 of 162 nine-inning games.

Then I attempted to capture the bump in performance from converting to one-inning appearances. Nate Silver of FiveThirtyEight observed an increase in strikeout rate from 19.9% to 23.9% when pitching for one inning, so I’ve applied that jump to the simulated WAR. This jump may actually understate the increase in WAR since it is relative to a 0% strikeout rate, not compared to a replacement level player, but I’m going to chalk up the fewer rest days as a reason to temper that a bit.

(I know this list is completely unrealistic. Just get the 14 of the best young pitchers. There’s also only one lefty, but sometimes, it’s just fun to look at some video game-esque numbers. It gets way crazier later, just bear with me.)

Holy Buster Posey, that’s an outlandish season! Not only would it be far and away the greatest team pitching season of all time, if the rest of the lineup was populated with replacement level players, the team would be the sixth best in baseball in 2019.

But we’re not getting replacement level players. We pinched our pennies on pitching so we could open the wallet for bats, but just as a Moneyball challenge, let’s complete this roster without exceeding the lowest payroll of 2019, the Rays’ $64,178,722.

We’re going to assume our team sees an average number of plate appearances, 6,217, and that they are evenly divided amongst our 11 batters plus one for pitchers. (This is trying to capture that some of these players may get injured, need rest, serve as pinch hitters for pitchers, etc.) That means we are looking at 518 PA for each batter, so let’s go find the best 2019 WAR/PA lineup we can get without breaking the bank:

The projected record for a team with 108 WAR is 156-6. I think this different-pitcher-every-inning thing might be worth a shot.