Relief Pitchers Perform Better Than Starters

With the Rangers and Cardinals each capturing their respective league championships over the weekend, one central theme of the 2011 postseason has become obvious – how important a dominating bullpen can be in October. Both the Rangers and the Cardinals were let down by sub-par starting pitching, but were rescued by shut-down work from their relief corps, and the depth of both bullpens allowed each manager to have a quick hook with struggling starting pitchers.

Throughout both series, but especially the NLCS, this was treated as something of a surprising development. We were constantly being told that the failure of the Cardinals starters to get beyond the fifth inning was a significant problem, and that if Tony LaRussa didn’t start to get more innings from his rotation, the Cardinals were doomed. Of course, the Cardinals never did get much from their starters in the NLCS (they combined to pitch 24 1/3 innings in six games) but won the series pretty easily anyway, outscoring Milwaukee 43-26 in the process.

The reason for the Cardinals success – besides an offense that put up seven runs per game, which never hurts – was the quality and quantity of work that LaRussa was able to extract from his bullpen; Jason Motte, Fernando Salas, Mark Rzepczynski, Octavio Dotel and Lance Lynn allowed three runs in 24 2/3 innings. The Cardinals essentially shifted innings from their starters to their relievers, and given what we know about the relative performance of starters and relievers, this was absolutely the right call.

Here’s the side by side comparison of starters and relievers for 2011:


Starters: 7.5%
Relievers: 9.4%


Starters: 17.8%
Relievers: 20.6%


Starters: 10.0%
Relievers: 9.0%


Starters: .293
Relievers: .287


Starters: 71.6%
Relievers: 74.3%

Starting pitchers issue fewer walks, but are worse at everything else despite being selected for the rotation due to their higher overall talent levels. The difference in strikeout rate for relief pitchers more than covers the extra walks, and then the lower BABIP and HR/FB rates help relievers keep those guys on base rather than letting them score. Not surprisingly, relievers posted an ERA of 3.69 this year, while starters had a 4.06 ERA as a group, even though starting pitchers are unquestionably better overall pitchers than relievers.

This isn’t a one year aberration either – due to the structure of relief work (shorter outings leading to increased velocity and situational match-ups, especially), teams almost always get better performances out of their bullpens than they do out of their rotations. What we’ve seen from Texas and St. Louis isn’t so much an outlier as it is the wise usage of the better asset.

During the regular season, getting quantity of innings from your starting pitching is necessary, because teams simply play too many games in a row to ask their relievers to carry this kind of workload. In general, even with a seven man bullpen, teams are only asking their relievers to pitch about 33% of the regular season innings. However, by capitalizing on the built-in travel days and the need for only for only four starting pitchers, the Cardinals handed 54% of their NLCS innings over to the bullpen, while the Rangers asked their bullpen to pitch 48% of the ALCS innings.

This marginalization of each team’s rotation gave both the Rangers and Cardinals significant advantages, as they leveraged better performances out of their bullpens than they could have hoped to get from their starters. Or, really, even than a team like the Phillies could have hoped to get from their starters. What the Cardinals bullpen did in the NLCS was essentially equivalent to giving their team three starts of eight innings each while only allowing one run per start. If a team’s rotation did that, it’d be obvious that they were going to win the series as long as their offense was even marginally competent.

For both Texas and St. Louis, their starting pitchers aren’t terrible, but they’ve recognized that they are more likely to keep their opponents off the scoreboard if they get to their bullpens as early as possible. This correct realization is the biggest single reason why these two teams are going to play for the World Series title, and the winner will likely be decided by which team’s relief corps performs better over the next week.

Given the relative costs of building a great rotation versus a great bullpen, don’t be surprised if this is the new blueprint for playoff success going forward. You can spend a fortune trying to gather a few high quality starters and keep them healthy, or you can spend a lot less and have a wide variety of bullpen arms available to bail out lower quality starters early in playoff games. The latter plan has worked to perfection for both Texas and St. Louis, and I’d expect that more and more teams will begin to build their rosters this way going forward, in order to take advantage of the natural fact that relief pitchers just perform better than starting pitchers.

Dave is the Managing Editor of FanGraphs.

Newest Most Voted
Inline Feedbacks
View all comments
12 years ago

How much of this is due to the fact that the lineups were set up for starters and cannot be adjusted for relievers mid game.