April 15, 2011
Relying on Relievers
The Heath Bell Curve
On Opening Day, the Padres and Cardinals dueled to an 11-inning finish, with the Pads edging the Cards in extras on the strength of a pair of RBI singles by Cameron Maybin and Nick Hundley off of Redbirds reliever Bryan Augenstein. Augenstein isn’t a bad pitcher; extra-inning games often devolve into staring matches between bullpens to see which will falter first, and something always has to give. However, when Padres starter Tim Stauffer exited the game after the sixth inning, the Padres, unlike the Cardinals, were handing the game over to the stronger part of their pitching staff.
Most teams, particularly when pitted against a strong starting pitcher, aspire to wear him down in order to knock him out of the game and reveal the soft, gooey center of the other team’s bullpen. Even if properly executed, that strategy can backfire against San Diego. In their second game, the Padres bullpen held the Cardinals scoreless for three innings after starter Clayton Richard yielded three runs in six innings. In the opener of the next series, the bullpen held the Giants scoreless for three more innings. When the Pads opened up a series at home against the Dodgers and Richard’s next start was delayed by rain after a single frame, the bullpen chipped in with ten innings and allowed just two earned runs.
That isn’t to say that San Diego’s bullpen has been perfect. On April 6, Pat Neshek and Cory Luebke combined to give up four runs after Tim Stauffer was chased from the game for doing the same. But in a comparison between the Padres’ starters and their bullpen, the bullpen clearly comes out looking like the better bet. Heath Bell, Luke Gregerson, and Mike Adams are among the best relievers in baseball. Mat Latos is the only starter on the team who is clearly above average, and he opened the season on the DL.
Given the workload they have been asked to bear and the skill with which they’ve borne it, it may not be surprising to learn that Padres relievers have thrown a higher percentage of their team’s innings than any other. At 42 percent through the first 11 games, the Padres featured one of just five bullpens to have thrown more than 40 percent of its team’s innings. The others—the Mets, the Yankees, the Royals, and the Orioles—all share the same characteristic asymmetry between the quality of the starters and the quality of the bullpen.
The best teams tend not to have this usage pattern. Last year, the three teams with the best records used their bullpens far less heavily. The 2010 Phillies secured the best record in baseball with a bullpen that threw just 29% of the team’s innings. The Rays had the second-best record, and their bullpen threw 31% of their innings. Even the Yankees—who so far in the young 2011 season have called upon their bullpen more than all but two other teams—used their relievers to record only 33% of their outs last year.
None of this should be especially surprising. After all, starters usually throw about twice as many innings as relievers, so if you’re going to have an imbalance between the two, better that the rotation be the stronger of the two. But last year wasn’t entirely average. In fact, last year’s league-wide bullpen usage rate of 32.9 percent was the lowest since 2005 and the second-lowest since the strike in 1994, as the table below illustrates.
While last year was a recent low-water mark for the season as a whole, it started—as many seasons do—with a relatively high rate of bullpen usage. Excluding the unusual case of 1995, when most starters had not pitched in some time, the average dip between the first 11 games and the season as a whole is 1.6 percentage points. (There is a weak correlation between league-wide bullpen usage rate in the first 11 games and the same rate for the rest of the season.)
The league-wide rate so far through teams’ first 11 games—34.3 percent—is similar to that of a roughly average full season. But if that mark follows typical patterns of historical decline, the final yearly rate may be much lower, in line with the very low mark observed in 2010. If that turns out to be the case, the high usage rates for teams like the Padres, Mets, and Yankees could be even more worrisome.
Reading the Lineup Cards
For some teams, a combination of rainouts and blowouts can produce a skewed result in the early going. We ought not to rest too much on the consequences of any 11 games, no matter when they happen. All of these teams could well end the season with a much lower usage rate. But there are some lessons to be learned here.
Let's take the Yankees. Their search for a fourth and fifth starter has been very closely watched, and even their supposedly more dependable options have struggled of late. Beyond CC Sabathia, they have four guys who cannot be counted on to give them six innings every time out. By contrast, the Yankees spent much of the winter assembling a very expensive bullpen. They gave $8 million to Pedro Feliciano (now out indefinitely pending shoulder surgery), $30 million to Mariano Rivera, and $35 million to Rafael Soriano. If the contracts play, the Yankees will end up giving a lot of their innings, along with their dollars, to the bullpen, and that has not generally been a winning strategy.
That’s not because it isn’t useful to have a good bullpen; it undoubtedly is. The reason why teams that rely on their bullpen very heavily tend not to win is because of what that usage pattern says about their starters: the other teams whose bullpens have thrown more than 40 percent of their innings thus far started Mike Pelfrey, Luke Hochevar, and Jeremy Guthrie on Opening Day. The Yankees (and perhaps the Padres—you never know) will probably still make the playoffs. But history indicates that their current path would be an inefficient way to do it.