Joe Girardi might seem to make some odd decisions when choosing relievers, but on the whole he’s done a fine job of managing his bullpen this year. I make this claim not based on stats, but based on what I’ve observed of the situation. Sometimes it seems he gets too cute in a LaRussa-like way, using multiple relievers to get just a few outs, but it works. Despite in April in which the bullpen had a 6.46 ERA, the Yanks pen currently sports a 3.94 ERA, fifth best in the AL. Of the things the Yanks have to worry about, the bullpen doesn’t appear to be one of them.
*If you don’t believe in advanced stats or believe that everything in baseball is self-evident, then you won’t want to continue.
Advanced stats support Girardi’s bullpen management. Jeremy Greenhouse at The Baseball Analysts looks at each team’s bullpen using a number of WPA-based figures: WPA/LI, Clutch, and pLI. The Yankees rank atop the league in Clutch from the bullpen, and up near the top in WPA/LI. In that neat Google Motion chart embedded in the article, you’ll see the Yanks dot hanging out by itself on the right. They’re easily the closest to the ideal position: top right.
Further supporting the argument is pLI, which is the Leverage Index for each player. The relievers with the highest pLI are mostly those with the highest WPA. The higher pLI means that these pitchers have the best chance of picking up WPA, since WPA fluctuates the most in high leverage situations. But because we’re seeing a high WPA, it means that for the most part the Yanks relievers did their job. Mariano Rivera was used in the highest leverage situations, 1.71 pLI, and has the highest WPA. Phil Hughes is second with a 1.43 pLI and a 2.4 WPA. Had these two failed more often, their WPA would not be as high.
Two names appear misplaced on the list: Phil Coke and Al Aceves. Aceves has the fourth highest pLI on the team, 1.07, but has a WPA of 1.92, third best in the bullpen. It would appear that Girardi should move Aceves up into higher leverage situations, and perhaps he would have if not for his late-season role as Joba’s caddy. Because Aceves was facing hitters in the fourth, fifth, and sixth innings — sometimes with a run deficit — his pLI dropped. Under normal circumstances, Aceves would move up into higher leverage situations. I expect him to regularly appear in the seventh inning this October.
The other is Phil Coke. It’s hard to get a good read on the rookie lefty’s season. He’s had great stretches and he’s had poor ones. He’s had games where he’s given up six runs, and he’s had appearances where he throws nothing but strikes. The bad appearances have hurt his WPA (as they should), leaving it at 0.88, yet his pLI is 1.29, third highest on the team. Girardi trusts Coke in big spots, and for the most part he comes through. It just seems that when he doesn’t come through, the results are beyond disastrous.
I’ve always thought that ERA is a terrible indicator of effectiveness for relievers. One bad outing can kill your ERA. See Phil Coke’s blowup against the White Sox. He’s clearly pitched better than his ERA this season. The biggest issue is that relievers work with such small samples. It’s why we see so much volatility from year to year in reliever performances. While WPA is a great narrative tool, I think it can also bring some insight into a reliever’s value. Relievers work in the context of a situation, unlike starters, who help create the context. The game is usually over halfway unfolded once a reliever appears, so he’s mainly pitching in the situation. WPA captures the situation. I’d like to see a bit more work go into this.
Pitching staffs get shortened in the postseason. The Yankees will only need three starters — if they’re willing to start CC in Game 4 of the ALCS on three days’ rest — in the first two rounds. They should also only need three relievers, maybe four. They’ll carry more, but chances are we won’t see pitchers not named CC Sabathia, A.J. Burnett, Andy Pettitte, Mariano Rivera, Phil Hughes, Alfredo Aceves get the ball very often, if at all. These are the guys who have performed best for the Yankees this season, and they’re the ones that will decide the team’s fate.
Given the circumstances, Girardi has done a good job not only of distributing the innings out of the pen, but of putting the right pitcher in the right spot. Sometimes he might make the wrong call, and when he does the fan base is quick to jump on him. Those instances tend to stand out in our minds, though, giving them a bit more weight. When we take a step back and look at the season as a whole, Girardi comes out ahead. Thankfully, other than a few lefty-righty matchups, he won’t have to think much about who to pitch and when. The Yanks relievers have done their job of proving it to him this season.