Literary theorist Terry Eagleton once said that “deconstruction…insists not that truth is illusory but that it is institutional.” In Major League Baseball, pitching staff management is nearly axiomatic; aside from intermittent attempts at four-man staffs, six-man staffs, or bullpens-by-committee, each team typically uses five starters, has a designated closer, and assigns other roles based on perceived skillsets such as handedness or durability. However, traditional baseball staffs utilize pitchers inefficiently and tend not to be cost effective. Thus, this article will explore how to build an affordable, results-driven pitching staff.
In the ten seasons played from 2002 to 2011, Major League pitching staffs compiled somewhere between the 1411.2 innings of the 2010 Pirates and the 2007 Padres’ mark of 1484.2 innings. These staffs faced a smattering of leverage situations over the course of those 1450-or-so innings each season. These range from the lowest possible Leverage Index—henceforth referred to as “LI”—of 0.0 (i.e. pitching to the home team while having a four run advantage with the bases empty and two out in the bottom of the eighth inning) to the highest possible LI of 10.9, which occurs while pitching to the home team with a one run advantage in the bottom of the ninth, two outs, and the bases loaded. The average situation is an LI of 1. Naturally, it follows that a team should use its best pitchers in high leverage situations and its least talented arms in lower leverage scenarios. However, this often is not the case. A closer who comes in to face the home team in the bottom of the ninth with his team up three runs—a save situation, if you will—enters a situation with an LI of one. He could conceivably give up a base hit of any variety and a home run to the first two hitters of the inning before the LI reaches a high level. It almost certainly is unwise to waste a team’s best relief pitcher in a low or medium leverage scenario. Therefore, teams should utilize high talent pitchers only in high leverage scenarios, middle relievers in medium leverage situations, and mop-up men in low leverage roles. Managers also cannot be reticent to change pitchers, as adjusting quickly to changes in base-out situations is crucial for highly analytical managers. On that point, Tom Tango notes in “The Book” that relievers can warm up in as few as four minutes, so it would be feasible to quickly adjust to changes in LI. Relievers tend to be volatile commodities year-to-year, so the scrap-heap bullpen model (see: Tampa Bay Rays) can be a very cost effective approach.
Starters should be approached via cost-benefit analysis as well. First, Tango illustrates in the aforementioned “The Book” that starters tend to do progressively worse against each batter starting with the eleventh man to bat in the game. Pitchers erode in effectiveness during their third time through the lineup so much that an average batter would perform three percent better than expected. Starters, therefore, should be, as they say, on a short leash after facing the first eleven men of the afternoon. With a capable relief corps that has a couple of innings eaters, teams should be able to absorb the brunt of a four inning start every now and again. Further, as the Baseball Prospectus team eloquently argued in their book “Baseball Between the Numbers,” starters generally throw slightly more effectively on three days of rest rather than four days of rest. The starters studied in that book did not lose effectiveness in August or September because of their higher workloads; this alleviates a common worry for the average fan, pitcher, or manager, as no one wants to see their star pitcher let up down the stretch. A way to optimize cost efficiency as well as on-field production, therefore, would be to build a four man rotation rather than a five or six man rotation. Starters would hypothetically soak up more innings per year and there would be no need for a fifth starter, so long as teams built depth in AAA and carried a long reliever or two who could spot start in the event of a minor injury. This would also give teams further roster flexibility and financial flexibility to add a situational bullpen piece.
All that being said, a pitching staff could be built as follows: four starting pitchers, two situational left-handers, two high leverage relievers, two medium leverage relievers, and two low leverage, long relief pitchers. This twelve man pitching staff would be able to utilize key platoon splits, adequately adjust to any game situation, and have enough depth to withstand a poor outing from or minor injury to any of its starters. Further, a manager could hypothetically use game theory tactics and mix and match his starters to the opposing team’s strengths by substituting a starter out for one of the long relievers. If teams were to embrace situational LI, avoid the pitfalls of locking fifth starters and closers into albatross contracts, and throw conventional wisdom out the window, they would be more successful on a per game basis and over the course of a season.