In 2012, the San Francisco Giants won the World Series. During the playoffs, they used 5 different starting pitchers. Those pitchers started 160 of their 162 games. By contrast, the 2024 Dodgers used 6 starting pitchers in their successful campaign; those pitchers, however, started only 48 of their regular season games. Between these two poles lie everyone else. The biggest argument against expanding baseball playoffs is that it devalues the regular season. When a team utilizes a starting rotation in the postseason that has little overlap with the starting rotation that got them there, that argument becomes particularly apt. MLB’s marketing slogan this fall is “October Hits Different.” Well, it pitches different, too.

I focus on starting pitchers for a few reasons. First, even as the inning pitched by starting pitchers continues to decline, everyone grants that the quality of a team’s starting rotation is still a vital index of overall quality. But I think people miss the fact that the main reason that this is true is that starting pitchers eat so many innings. While that’s crucial in a 162 game season, it is almost meaningless in a playoff season littered with off days. In addition, the demand for starting pitching in the playoffs is low enough that lots of erstwhile starting pitchers find themselves assigned to bullpen duty in the playoffs, further lengthening the bullpen depth.

So in some ways the structure of the playoffs already makes a starting rotation less important. The 1968 Detroit Tigers won a World Series starting Mickey Lolich 3 times, Denny McLain 3 times and Earl Wilson once. So this isn’t completely a new phenomenon by any means. And the quick hooks and intentional bullpen games used in the modern game mean that 7- and 8-inning starts won’t happen, but I’m still old-fashioned enough to think that the more different the regular season game is from the October game, the worse it is for the regular season.

Second, starting pitchers make a lot of money. Expanding the playoffs make it possible to reach the playoffs with a worse starting staff. If you can get to the playoffs, you definitely don’t money wasted on fourth and fifth starters if that monety could have been reallocated to hitters… the people who actually win playoff games. This is just the flip side of the point above: the pitching level in the playoffs is so high, and changing so rapidly, that only elite hitters can reliably hit them. This doesn’t mean aces are overpaid; they may well be underpaid. But it does mean that giving them a bunch of starts in the regular season may well be a waste and we shouldn’t be worried when an ace goes down for a month or two, subject to their ability to return.

Third, it’s my study and I can study anything I want, so playoff starters it is.

The Data

I looked at every playoff team in the Retrosheet Database, I then deleted the teams that used 3 or fewer pitchers after 1969, largely one-and-done Wild Card teams. I also threw out 2020, because it was a stupid season. For each of these teams, i kept (a) the number of different starting pitchers used; (b) the team’s W-L record; (c) the team W-L record of the regular season games started by the same pitchers; and (d) the number of regular season games.

We’re going to judge the importance of the regular season starting rotation by the fraction of regular season games started by the guys who started in the playoffs. The overall graph looks like this:

Before the playoff era, the fraction of games started by your your World Series starters was essentially random, and largely depended on whether you used your top 3 starters or top 4 — when there were at most 7 possteason games; there are only 18 teams in this period that used as many as 5 starters in the World Series, and only one who used 6. There were some teams who essentially started their rotation while others did not. We should note the one first time when a reliever started a World Series game was not recent. In 1950, Jim Constanty, the first reliever ever to win the Cy Young Award, started Game 1 of the World Series for a depleted Phillies staff.

There was a big bump with the arrival of the LCS in 1969, as every team realized that they needed to use more pitchers to get through the playoffs. But since then, the trend since 1969 has been steadily downhill, and while there was brief bump in the 1990s, the acceleration has been pronounced in this century. (The red line is the three-year moving average.)

To separate out the “Games Played” effect, in which more playoffs games played will tend to higher coverage fractions, we can subset the data by number of playoff games played and just look at those teams that played 10 or more playoff games*:

The trend is much clearer here.

Is it just fewer pitchers used, or are they new guys?

There are a lot of reasons for low regular season coverage fractions.

  • Fourth and fifth regular season starters don’t start many playoff games
  • Pitchers get injured and return from injury in time for the playoffs
  • Pitchers get injured after a lot of regular season starts and miss the playoffs
  • Late season acquisitions, either by trade or promotion

The first of these doesn’t bother people too much. One team may have a better rotation depth than another, and that may be really important in deciding who makes the playoffs, but I think we all just accept that it won’t be a factor in the playoffs themselves. The last of these is “traditional,” and thus least likely to bother anybody. And the late-season trade has diminshed greatly with the earlier deadline.

Injuries are another matter. I have a friend who’s a big NBA fan, and the tendency of teams to rest players for a while in midseason is regarded as a growing problem. If I want to see LeBron James play, but he doesn’t road games in January and February then part of my reason to watch the regular season games goes away. I don’t think players fake injuries, but when it’s in nobody’s interest to rush players back, they won’t rush back.

What are you doing about starters moved to the bullpen?

Nothing. They don’t count. If the manager didn’t want to treat them as a playoff starter then neither do I.

What’s your point?

We have reached the stage where, for the teams going deep into the playoffs, their starting pitchers started less than half of their regular season games. The 2024 Dodgers were an extreme, but represent an ominous trend. Combined with the quick hook for starters, it is easy to foresee a time when a team starts its ace and maybe its #2 with the rest as bullpen games. I have no objection to this in priniciple — maybe it is the best way to manage a playoff team. But you can’t use this strategy in the regular season… you’ll burn everyone out before the All Star game. And as long as the strategy for regular season player management is a loser in the playoffs, you’re going to devalue the regular season. You couldn’t do this in the past, because so few teams made the playoffs if your only choices were to configure yourself for the regular season or the playoffs, you had to choose the regular season. Then you adjust in the playoffs as best you can.

The opposite strategy is now feasible: downplay your rotation, don’t worry too much if a pitcher needs a month or two off. Back into the playoffs and have a team ready to win. An associated prediction is that you’ll end up with no great regular season teams. The evidence here is very anecdotal, but there were no 100 win teams this year or last year. It may be MLBs idea of parity to get 12 87 win teams to fight to see who wins a World Series, but it isn’t mine.

Solution?

This problem actually has a pretty easy fix: eliminate travel days in the postseason. Every team now travels on charters. They can travel across country whenever they need to. The difficult travel, West Coast to East Coast, can be alleviated by playing West Coast day games on travel days and then not playing the next day until night. Or you can make an exception for that single situation.

I realize this situation has some issues, particularly with TV scheduling, and antagonizing TV networks gives this solution no real chance of implementation. But if you think the regular season is the “real” game, and the postseason is the “gimmicky” game (and I realize plenty of people feel exactly opposite!) this change will make postseason pitching management a lot more like in-season pitching management.

Altenatively, if you want the regular season to be more like the postseason, shorten the season in games and add off days. Then every team will have two starters and 13 bullpen pitchers. If you want to know what that looks like, though, look at the Braves’ season this year.

==============================================================

*For those wondering, the one example of 10 or more playoff games before 1969 was the 1962 Giants who played a three-game tie-breaking series with the Dodgers before losing in 7 games to the Yankees.