The scene: it’s the bottom of the ninth inning, and the team we want to win—let’s call them the Frockies—are ahead by two runs. Married to rigid bullpen usage, Frockies manager Salty Schwartz brings in the team’s Closer. The opposition—we’ll call them the Vintners—couldn’t chip in to the Frockies lead in the eighth inning. The reason why is no mystery. The Frockies Set Up reliever is also their best reliever—he handled the Vintner lineup easily. While the Closer is no slouch, he’s either the second or third best reliever on the team. The Vintners proceed to tie up the game—to extra innings we go. The Frockies’ rigid bullpen usage might have cost them the game.
This transparent description of Wednesday night’s game in Milwaukee gets at question that might pop up from time to time this season. In a world where Walt Weiss and the Rockies employ the targeted shifting of infielders, bat their best hitters second, hit the pitcher eighth, and take advantage of platoons, why not also go all in with bullpen optimization?
Bullpen management is another site where managers tend to use arms sub-optimally from a sabermetric point of view. That point of view is very simple: managers should use their best relievers in situations that have the highest leverage. A team’s closer is usually the best reliever—why use him in the ninth inning with a two run lead and against the seven through eight hitters rather than in the seventh inning when the lead was one run, a runner was on base, and the opposition’s two through five hitters were coming to the plate?
The Rockies are in a strange situation in this area because their best reliever, Adam Ottavino, isn’t the closer. Those who argue for bullpen optimization usually clamor to take the best reliever out of the ninth inning. Weiss is already in a position to use Ottavino flexibly. The same goes for the player who just might be the Rockies second best reliever, Boone Logan. Neither is constrained by the ninth inning.
And yet, sometimes the ninth inning does provide the highest leverage situation. Sometimes the Closer is tasked with shutting down the opposition rather than formally recording three outs en route to a three run victory. That’s the point of eliminating the Closer role and just using the best relievers in the highest leverage situations.
Not only that, but we can’t discount the importance of roles for players. In the 2015 Baseball Prospectus Annual, Russell Carleton has an essay titled "I Was Promised Flying Cars and Managers who Use their Closers Correctly." The essential point is this: managers know about bullpen optimization, just like we have the technology for flying cars. But that doesn’t mean either is necessarily a great idea, and both might cause more problems than solve. "Those of us who come up with crazy ideas for teams to implement," Carleton writes, "need to first think about the ecosystem in which the players actually live. What works on paper might not work out in the wild." Still, any crazy idea can work with the buy in of the players, and players can buy into any crazy idea if they are persuaded it will work.
The reveal: the story of Salty Schwartz and the Frockies was rhetorical. Walt Weiss did everything right according to bullpen optimization on Wednesday night. It just so happened that those actions dovetailed exactly with conventional bullpen usage. Adam Ottavino entered Wednesday’s game when the Brewers had a 14.7 percent chance of winning. He was set to face the top of the order and the three best hitters in the Brewers’ lineup: Carlos Gomez, Jonathan Lucroy, and Adam Lind. Ottavino retired the first two, walked Lind, and struck out Aramis Ramirez, at which point the Brewers win expectancy was 6.6 percent. When the Rockies failed to score in the top of the ninth, LaTroy Hawkins faced an opposition with an 8.1 percent win expectancy. He came in to face Khris Davis, Gerrardo Parra, and Jean Segura. In the ninth inning, the leverage was lower and the opposing bats less intimidating.
Maybe Walt Weiss will optimize bullpen usage by way of convention. LaTroy Hawkins is still effective. He doesn’t miss bats or strike many batters out, but that’s never been his game, and he knows it. Hawkins provides stability for the rest of the bullpen by shouldering the ninth inning role, which will be a mixed bag of high and low leverage innings. Doing so allows Walt Weiss to be flexible with everyone else. In particular, Hawkins is the anchor that can allow Weiss to strategically deploy what has the potential to be a lethal trio of arms in Ottavino, Logan, and John Axford.
We’re less than a week into the 2015 season, but bullpen usage, like much else at this moment, is working just fine.