Suppose you are a baseball manager, and your team’s usual closer is unavailable for tonight’s game. What do you do? If you are like 90% of MLB managers, then you basically just move everyone up a slot: Your usual 8th inning guy becomes your closer for the night; your 7th inning guy pitches the 8th; and so on.
But this is wrong. It misunderstands what makes roles valuable, and treats players like interchangeable parts. It is a fundamentally anti-labor attitude. Allow me to explain why…
Every once in a while, a baseball team will try to go without a traditional closer, opting instead for a more flexible bullpen strategy. Whenever this happens, people will point out how the 9th inning is not always the most high-leverage inning—indeed, the most important outs you get from the bullpen are often in a tie game in the 7th, or to stop a rally in the 8th, or some other situation when you wouldn’t normally use a closer. So saving your best reliever for the last three outs is irrational.
This theory makes sense on an intuitive level, but the teams trying this strategy almost always end up abandoning it for a traditional closer. It’s happened so many times that it’s now rare for teams to even try the “closer by committee” strategy. The 2019 Chicago Cubs tried to piece together the end of games with Pedro Strop, Steve Cishek, and the rest of their bullpen, and their relievers overall had decent numbers, finishing 8th in ERA and 6th in batting average against. But after a number of blown saves, the team broke down and signed Craig Kimbrel midway through the season (of course, Kimbrel ended up being terrible too, and the Cubs missed the playoffs).
So why doesn’t it work? I can think of two reasons, one good and one bad—but I suspect it’s mostly the bad one that keeps the status quo in place.
Let’s start with the good reason: Defined roles are good for players. Relievers tend to pitch better when they have a clear sense of what is expected of them, and can plan accordingly. Knowing that you are likely to come in with runners on base, or to face a string of lefties, or to get the last three outs of a game, etc., allows you to prepare, both physically and mentally, in a consistent way, and adapt your skills to those specific needs. A guy who’s likely to come in with runners on can focus on getting strikeouts, whereas a reliever likely to be used for multiple innings will focus on developing secondary pitches. So even though it might seem overly rigid to lock guys into roles, it may be the best way to optimize their talent.
I like this idea because it fundamentally rejects a kind of “talent essentialism” that plagues closer debates. The logic of “you should use your best reliever in the most high-leverage situation” falls apart if there is no “best reliever”—only the reliever most suited to the situation. By this logic, a closer is simply another role that a reliever has adapted to, and not the King of the Bullpen in the way that is traditionally depicted.
There is a useful lesson here that applies to all kinds of labor, specifically that success is more often attributable to structural or circumstantial forces than innate individual abilities. Talent disparities obviously exist, and people have different skill sets, but talent is often a matter of people finding roles that suit them and collaborating effectively, and not just individuals actualizing those innate abilities. This could also explain why closers seem to have more stable, long-term success than middle relievers, since the “closer” role is the most clearly defined role, that is consistent from team to team, and therefore closers experience the least change in their role from team to team and season to season.
But there is a less charitable explanation for the persistence of closers: They make a manager’s job easier. A “closer by committee” requires the manager to constantly make ad hoc decisions about bullpen usage. They have to decide how high leverage a situation is, compared to any situation they are likely to face in future innings. Without a standard formula, every decision can be easily questioned if it doesn’t work out. But having clearly defined roles allows the manager to default to a standard sequence. If this doesn’t work, then people generally blame the players, not the manager. After all, he just pushed the buttons he always pushes.
Which brings us back to the hypothetical I started this with: the Unavailable Closer situation. The way almost every manager handles that scenario—sliding everyone back an inning—suggests that they are mainly concerned with making their own jobs easier. After all, this keeps all the roles the same; only the faces occupying them change. So a manager’s decisions about how to handle the last ten or so outs of the game are roughly similar to what they would be if the closer were available.
But if the roles existed for the benefit of players, then a manager would handle the Unavailable Closer situation differently. In that scenario, the goal would be to get those outs from the bullpen while keeping everyone in as close to their usual role as possible. This would be much harder for a manager, who would have to make decisions on the fly, based on context and feel, and those decisions could be questioned if they don’t work out. But it would be much less disruptive to the pitchers in the actual game.
As it happens, the Yankees faced the Unavailable Closer problem in a game last month against the Red Sox. Aroldis Chapman was unavailable, and the team took a 3-1 lead into the bottom of the ninth, at which point manager Aaron Boone brought in Chad Green. Green was, in many respects, the natural “backup closer.” He had the most saves of anyone on the team besides Chapman, and typically pitched the 7th or 8th inning of close games. But this was only the fourth time, out of 40 appearances, that he’d entered the game in the 9th, and in two of those previous occasions he’d gotten a loss. This is not to suggest that Green is not tough enough to handle the 9th, or that he has some mental freeze about it, but that’s simply the not the role he’s thrived in this season.
In other words, Green is a very good pitcher, but his skill set doesn’t translate naturally to the closer role. He’s typically great (35 of his 45 appearances this season have been scoreless) and he’s very good at stranding inherited runners (fourth in the AL). On the other hand, he’s not especially good at toughing it out when he doesn’t have his best stuff: Of the 47 baserunners he’s allowed this year, 22 have scored. That’s not exactly what you want out of a closer…
Indeed, it was immediately clear that he was in trouble against Boston: Two of the first three hitters hit hard line drive singles. Normally, Boone could have replaced him, but the guy he would normally go to in a jam like this was already on the mound…
Compare this with the other option Boone had: leaving in Luis Cessa, who had pitched the 8th and retired Boston’s 3-4-5 hitters on five pitches. Letting Cessa finish the game would have been a little odd—Cessa had typically been used in low leverage situations this season, or in extra innings when other relievers had been exhausted. But he was having the best season of his career, entering the game with a 2.82 ERA, and Boone had already used Cessa unconventionally by bringing into a close game in the 8th, and Cessa had responded well. All that was left was for him to do something he was quite used to: Come out for a second inning (something he’d done in 10 of his 28 previous appearances) and retire the bottom of the order. If he got into any trouble, then Boone could have used Green in a situation that was more common for him: cleaning up a jam created by someone else.
Of course, it’s possible even that would not have worked. Sometimes guys just don’t perform, and I don’t want to overly fixate on one bad managerial move. But watching it all unfold was depressing, because it felt like the team was following a script written for the manager’s benefit, regardless of what the players did. They were merely interchangeable cogs to be plugged into prefabricated molds, instead of individuals with specific skills and personalities.
This happens all the time, not just in bullpens, and not just on baseball teams, and not just in sports. Workers everywhere are asked to become cogs fitting into preexisting machines, not because it’s good for them, or even because it’s efficient (the Yankees lost the game, by the way), but simply because it makes life easier for their bosses. This just proves that old baseball adage: Worker autonomy wins ballgames.
I feel like you need to grapple with the Andrew Miller run from 2016. Wasn't that an instance of a manager deploying his most talented player in an atypical way in order to maximize his contributions and the rest of the pitching staff?