April 7, 2011
The Closer Quandary
As the season began, Fernando Rodney’s hold on the Angels’ closer job was believed to be tenuous. Other than possessing the “proven closer” label, there wasn’t much about Rodney to recommend him for the role. His “success” as a closer, such as it was, was more a testament to how overrated the role is, not his own ability to pitch.
Despite those concerns, few would have expected him to surrender the title as early as he did: Rodney was removed from the closer role on Tuesday, after just two outings and one blown save. What was it about the one-and-a-third innings Rodney had pitched so far this year that wasn’t already apparent from the previous 398 innings under his belt? Sure, the most recent innings were worse, but anyone can pitch that poorly in less than two innings. A more impressive sign of mediocrity is being able to pitch a hairsbreadth away from replacement level for eight seasons, which Rodney had already accomplished.
Rodney is being replaced for the time being by Jordan Walden, who can probably pitch better than Rodney did in his first two appearances, but that's as high as expectations should be set for the rookie pitcher. Yes, he had gaudy strikeout totals in limited action last season, but his minor-league strikeout rates—against inferior competition, mostly at Double-A and below—don’t hint at an ability to sustain that success in the majors.
The more interesting question is this: Given the choice between such uninspiring relief options, does it really matter which pitcher is chosen to pitch the ninth and which sets him up? And is there a lesson to be learned from other teams whose closers have looked shaky to start the season?
In general, teams make the relief pitcher they think is their best the closer, and the next-best relief pitcher the setup man. So how much does the order in which those two pitches are used matter? It's not difficult to figure out. Looking at the period from 1993 to 2010, I identified a closer and setup man for every team—the closer was the pitcher with the most ninth-inning save opportunities, while the pitcher with the most save opportunities in any inning who wasn’t the closer was designated the setup man. In order to see how they performed in close games, we needed a more realistic definition of "close" than the saves rule provides. I defined a close game as one wherein the pitcher’s team is trailing by one run, tied, or leading by one or two runs.
When you look at these games, what becomes instantly apparent is that a team's top two relief pitchers don't get to pitch in close games much more than half of the time—58 percent for closers and 55 percent for setup men, for an average of 56 percent. (Setup men pitched slightly more often than closers, which drags the average down a little.) While closers pitched in a slightly higher rate of close games, they pitched in fewer games overall, so setup men on average pitched in one more close game per season than their teams’ closers. In other words, making a team’s best reliever its closer doesn't get him into tight games more often. What it does seem to do is get him into fewer games than the team’s next-best reliever. And whether employed as a closer or as a setup man, top relievers end up pitching rather frequently with the game not seriously at stake.
The modern role of the closer is an accident of history—its progenitor, Dennis Eckersley, was in the right place at the right time. A balky shoulder, among other maladies, brought Eckersley’s days as a full-time starting pitcher to an end. His manager, Tony La Russa, came up with a way to use Eckersley that suited both his talent and his limitations. Rather than taking that as a cue to tailor roles to the individual qualities of each pitcher, managers are trying to take a whole bunch of pitchers and turn them into Eckersleys. It’s a staggering display of how to learn the exact opposite of an important lesson.
I would be remiss not to mention the role the save rule plays in all this. As someone who spends a fair amount of time working with baseball stats, I often get accused of ruining the sport. That’s giving me too much credit—I've only dinged it at worst—but there are baseball stats that are making the game worse, and the save is one of them. To slightly modify a quote from economist John Maynard Keynes:
The ideas of sportswriters, both when they are right and when they are wrong, are more powerful than is commonly understood. Indeed the world is ruled by little else. Practical men, who believe themselves to be quite exempt from any intellectual influence, are usually the slaves of some defunct newspaper columnist.
The role of the closer is defined not by what is best for the team, but by what is best for the closer, who personally benefits from accumulating saves--a pursuit which is not necessarily equivalent to helping his team win. Managers bring in their closer to protect a three-run lead, which is a staggeringly easy assignment for all but the worst of pitchers, but will leave their best reliever to sit and watch as a lesser pitcher blows a tie game in the final inning, all because an arbitrarily-constructed rule tells them that it is not a "save situation," as if that term and "potential game-changing situation" are interchangeable. They're not.
All of this extra attention to making sure the best reliever pitches the ninth, and only the ninth, hasn’t led to a significant increase in how often teams hold on to ninth-inning leads—on average, teams will cling to one additional one-run lead every two seasons, compared to rates posted prior to the evolution of the modern closer role. (Probably less if you’re left to pick between pitchers like Rodney and Walden.)
Managing a bullpen is a thankless chore, and more complicated than it may appear at first glance—a manager has to win not just today’s game, but tomorrow’s and the day after that, and he also has to factor in how recent use has affected each reliever’s effectiveness. Having set roles for everyone in the pen relieves the manager of a lot of that labor, and relief pitchers seem to prefer having set roles as well. What’s more, there is a strong aversion to relieving one’s fireman with an inferior pitcher—the successors of the writers who gave us the save rule and other such contrivances lie in wait to pounce upon the manager who deviates from these norms in a losing effort. (Never mind that managers who adhere to them will often be rewarded with failure as well.) So despite what people like me say, teams are going to have closers for a very long time to come.
But because of their extra prominence, the successes and failures of closers take on a greater significance in our minds—and in the press—than they really deserve. Better pitchers than Fernando Rodney may struggle to pitch well over a short stretch of time—across town, Jonathan Broxton has a ballooning ERA, and in Chicago, Carlos Marmol has been roughed up quite a bit to start the year. They’re both pitchers with long track records of success, and while just about anything can happen in sixty-some innings, the most reasonable expectation is that these pitchers will go back to being as successful as they have in the past. Do they have to close games to help their teams win? Not at all. But if managers think that their best pitchers need to be closing games, they shouldn’t let a week’s worth of games tell them otherwise.