Skip to main content
ABC News
Starters Are Getting Shelled — And Getting Pulled Early — This Postseason

Sunday night’s American League Division Series game between the New York Yankees and Cleveland Indians was a true pitchers’ duel — the kind of taut, low-scoring affair we tend to think of when we conjure up mental images of postseason baseball.

It was also completely out of place in this year’s playoffs.

Through Sunday’s games, this is the highest-scoring postseason through the Division Series1 in 15 years, with the typical score resembling the Nationals’ 6-3 win over the Cubs Saturday night (or the pair of 8-2 victories the Astros hung on the Red Sox to open their series) much more closely than the Yankees’ 1-0 Game 3 margin. Starting pitchers have borne the brunt of the damage, and — coming on the heels of the Postseason of the Reliever — managers have used this as an excuse to hook them quicker than ever. All signs point to this being a bullpen-dominated postseason like no other.

For all the comparing of rotation strengths we love to do going into the playoffs, few starters have fared well in October so far. According to our game score metric, which quantifies a pitcher’s performance in a given start according to a variety of stats, starters are collectively suffering their fourth-worst early-round performance since 1995.2 They’re also allowing their worst ERA and easily pitching the fewest innings per start of any postseason in that span. Not even Corey Kluber and Chris Sale, probably the two best pitchers in baseball this season, were immune to downright awful starts in the postseason’s first week.

But more significantly, starters are also being pulled out of games much earlier than usual, even after accounting for their rocky performances. To judge this, I tracked each starter’s performance in all Division Series and Wild Card games since 2009,3 breaking the game down into two chunks — the first three innings were one unit, and everything from the fourth inning onward was another. Not surprisingly, there’s a pretty reliable relationship between how few runs a starter yields in his first three innings of work and how many batters he gets to face over the remainder of the ballgame, which we can use to set up an expectation for how long each start will last, controlling for performance. So far this year, starters are being allowed to face about two and a half fewer batters after the third inning than we’d expect based on how many runs they allowed through three:

Starters are getting the hook early this postseason

Difference between actual and expected batters faced from the fourth inning onward, for Division Series and Wild Card games

2017 24 11.5 1.9 7.3 9.8 -2.6
2016 34 12.4 1.7 9.3 10.3 -1.0
2012 44 12.0 1.2 11.6 11.5 +0.1
2011 38 12.2 1.4 11.2 11.0 +0.2
2009 26 12.3 0.8 12.7 12.4 +0.3
2010 30 12.0 0.9 12.5 12.1 +0.4
2015 42 12.3 1.7 10.7 10.3 +0.4
2013 40 11.7 1.2 12.1 11.5 +0.6
2014 32 11.7 0.9 12.9 12.1 +0.8
Average 34 12.0 1.3 11.2 11.2

Expected batters faced from the fourth inning onward based on runs allowed through the first three innings.

Source: ESPN Stats & Info

This trend was already starting to emerge last postseason, when starters were getting a slightly quicker hook than expected. And it got a bunch of attention on the game’s biggest stage, as Cleveland’s Terry Francona and the Cubs’ Joe Maddon each drew praise from analysts for deploying their bullpens creatively en route to the World Series. So perhaps it shouldn’t be surprising that this year, almost every playoff manager seems to be emulating Francona and Maddon. In 2017 to date, relievers are tossing an incredible 52 percent of the available innings in playoff games — which, if it holds up, would be the first time they’ve cracked half the innings in a single postseason since at least the beginning of baseball’s expanded playoffs in 1969.4

That’s not to say starters’ roles are being completely diminished in the modern postseason. Some, like Boston’s David Price on Sunday, are being asked to enter the game as relievers, where they stand to be more effective anyway. One side effect of managers having such an all-hands-on-deck approach in every postseason game is that any pitcher — starter or reliever — could be asked to play an important role in any given game. The kind of appearance that used to be reserved for legendary pitchers in must-win games, such as Pedro Martinez’s six-inning relief masterpiece in Game 5 of the 1999 ALDS or Randy Johnson’s scoreless inning and a third in Game 7 of the 2001 World Series, might become the norm.

But if the early games of this postseason are any indication, teams will have to weather a rough first few innings from their starters before settling in for the bullpen chess match. For better and for worse, it’s just another way in which 2017-style baseball is starting to look very different from the game we all grew up watching.

Check out our latest MLB predictions.


  1. Including Wild Card games since 2012.

  2. Division Series and/or Wild Card.

  3. The earliest season for which ESPN’s Stats & Information Group has data broken out by inning.

  4. I estimated this for seasons before 2009 by using the postseason pitching data in Sean Lahman’s database, parceling out innings by role in proportion with how many of a pitcher’s postseason appearances came in starts or in relief.

Neil Paine was the acting sports editor at FiveThirtyEight.