Risk Aversion and the Perils of Worrying About Middle Relief
The need to be risk averse runs deeply in baseball. Not surprisingly, it often leads to questionable decisions by managers and other decision makers. The incentive structure under which these people must work probably makes this unavoidable. In general, the handling of pitchers is where risk aversion inbaseball is the most profound. The one occasional exception is that the conventional wisdom around pitch counts is not 100% crystallized, but once that occurs, the risk aversion will be strong there as well.
Currently, within the world of pitching the nexus of conventional wisdom and risk aversion is most potent around the question of middle relief. A manager who goes through three pitchers while his team gives up four runs in the 6th-8th innings is seen as managing aggresively; if that same manager lets one hapless reliever absorb the four run blow, that manager is seen as lazy. Similarly, bringing a closer in for the 8th inning to face the toughest part of a team’s lineup is seen as risky and tampering with the natural order of things, while leaving that closer on the bench because an inferior pitcher, or pitchers, lost the game in the eighth will rarely earn a manager criticism or threaten his job security.
Middle relief is such as ubiquitous concern throughout baseball that during August and September, somebody who knew nothing about baseball could call in to a baseball talk show, join an online baseball chat, go to a sports bar to watch a ballgame, or talk baseball anywhere in the US and sound like an expert simply by beginning every conversation with “I’m worried about our middle relief”, and then simply nodding sagely, or whatever the online equivalent of that would be, during the rest of the conversation.
Saying “I’m worried about our middle relief” is the baseball equivalent of saying “It will all come down to turnout” when discussing a political campaign, or making a joke about the French when discussing international affairs. It is what you say, when you don’t really have anything useful to contribute to the conversation. Nonetheless, many teams continue to spend money and talent to solve this problem, not because it is the biggest problem they face, but because it is the easiest, and least risky to address.
Middle relief, or the bridge to the closer, as some are inclined to call it, is obviously important, but it this must be examined in context. Middle relievers are, by definition, the weak half of the pitching rotation. No closer ever became a middle reliever simply because his skills were more needed there. While not unheard of, it is very rare for an effective starter to be switched to middle relief. One of the reasons so many teams have so many problems with middle relief pitching is because these pitchers are not supposed to be great. The good middle relievers get moved to more important roles after a season or two. For example, in 1996, Mariano Rivera had one of the greatest years ever by a setup man-a close cousin of the middle reliever. He was promoted out of that role by spring training the following season. I don’t know of any Yankee fan who regrets that decision, but the Yankees middle relief has not been as good since.
The problem of middle relief, which is not so much a problem but a condition which draws inordinate attention because it is a safe issue about which management can worry, is often addressed in ways that weakens teams. One common solution to the middle relief problem is to simply add more middle relievers under the surreal theory that if the first five guys out of the bullpen aren’t doing the job than giving the work to a sixth or seventh will solve the problem. This approach is not harmless because adding pitchers means depleting the bench. A team with 12 pitchers usually does not have one legitimate pinch hitter, while a team with 13 pitchers, particularly in the American League leaves, themselves with a skeletal bench of one middle infielder, one catcher and one outfielder. Thus, in the name of the elusive to solve, but safe to articulate, problem of middle relief, teams leave themselves with no choice but to let their worst hitters bat in clutch situations late in the game.
Another common solution to the problem, particularly in the stretch run, is to trade for a veteran middle reliever. Teams do this, not because it is wise, but because the incentive structure encourages this. General managers who trade away prospects so solve the middle relief problem are often gone by the time the prospect they have traded away becomes a success with his new team, but those who do not make these moves are seen as standing pat and not doing enough in the heat of a pennant race.
It seems to me that there are a few easier ways to solve the perpetual problem of middle relief, but they involve taking risks and run counter to the incentive structure for baseball’s decision makers. The first is to cut back on the number of middle relievers and rather than hope to find the magic reliever from the minors simply accept that the 6th-8th innings will be tough and add extra players who can be used to bolster the offense and add non-pitching depth. It is the rare team for whom the 8th, 7th, or even 6th, pitcher out of the bullpen is more valuable than a good pinch hitter with some power, yet this is the tradeoff many teams make.
The second approach is also somewhat more counterintuitive. If the middle relief is bad, the answer might be simply to upgrade offensively at one position. Replacing a weak hitter with a stronger one for the last months of the season will often have the same effect on the ration of runs scored to runs allowed a minor upgrade to middle relief. An offensive upgrade, however, is usually easier and can be done with more certainty. A team that did this would likely come under criticism for misdiagnosing the problem, so it would be an unusually risk accepting approach. However, it does not take a Ph.D in mathematics or to figure out that if you are being outscored, there are two ways to solve the problem-by either allowing fewer runs, or scoring more runs. The sixth and seventh innings can be a lot easier if the team has two, rather than one, run lead to protect; and bolstering the offense can help make this happen.
My point is not to suggest that middle relief is not important, but that it is a problem that may be solved by tinkering with parts that are already there, getting lucky or scoring a few more runs. Obviously, if a team has good middle relief it is a big advantage, but dramatic steps to improve middle relief frequently do not work and may have unforeseen, but easily foreseeable, consequences on the team more generally. Moreover, there are other ways to skirt the issue of poor middle relief, notably by scoring more runs. The baseball people who understand these perils of the middle relief narrative, and become a little bit more risk accepting than their competitors will have a real advantage.