We all know what it takes for a relief pitcher to be good, the same things that make any pitcher good (in a statistical sense or otherwise). What makes relievers different from other players is that being good isn’t enough for a player to matter – for a reliever to make an impact they need to combine performance with leverage. Leverage is a finite ‘resource’ to spread around a bullpen, so some relievers matter disproportionately more than others – one reason why ‘ERA’ or ‘WAR’ don’t give you the real story of bullpen quality. Every bullpen (except maybe Minnesota’s) has a top guy who is – in theory – going to be the one that pitches in the highest pressure situations. Except for one kink – the ‘save’ statistic and hidebound managers have effectively converted this ‘relief ace’ role into that of ‘9th inning when we have a small lead’ guy. The fact that managers consciously ration leverage to their best relievers would lead us to believe that a team’s top reliever should have a disproportionate impact on the team’s wins and losses but that could be mitigated by the fact that he wouldn’t even be considered for key situations that don’t happen in the 9th inning.
Follow through the jump for a short tour past the closers of the American League;
Before we get into team by team specifics, the first things to note are the spreads: In terms of win-probability added, the most valuable closer in the AL last year was Jose Valverde with 4.17 wins and the least valuable was Matt Capps of the Twins with -1.1 for a top-to-bottom spread of 5.27 wins. Capps, of course, lost his closer role in mid season (bad closers tend to do that) causing a bit of a drop in leverage. You’re unlikely to ever see a -4 WPA from a closer, though it’s theoretically possible, because they lose their 9th inning jobs and get demoted to mop-up duty. In the NL we did have some bigger stinkers than Capps in Bobby Parnell of the Mets with -2.86 and Ryan Franklin of the Cardinals with -2.82. The bar for the last decade was set by Brad Lidge of the Phillies in 2009 with -4.54 wins – an example of what you can get from a failed closer that somehow holds onto his high-leverage role. Why did he get the opportunity to do so much damage? The year before Lidge had the highest WPA in baseball with 5.37 wins. The highest totals tend to be in the 4’s and 5’s, no closer is ever likely to get enough chances to do better than that, no matter how well they pitch.
So, the theoretical spread is about 10 wins between the best closer imaginable and the worst closer imaginable. In any given league in any given year it’s probably closer to 5 or 6 wins. That isn’t insignificant, by any means. All of these players are good relief pitchers, even if some are having bad years. You would definitely want them on your team, you just might prefer to see somebody else protect a 5-4 lead in the bottom of the 9th. The spread between bullpens overall in the AL was about 12.5 wins by WPA between the Yankees (at the top, unsurprisingly) and the Twins (at the bottom, equally unsurprisingly). That spread looks to be fairly normal, though the Yankees bullpen was unusually good and the Twins not quite historically bad. The spread between the best and worst closers in a league appears to be roughly half the spread between the best and worst bullpens overall – only 42% in the 2011 AL but often larger in other years.
Now lets take a look at who was great, who stunk and why:
Detroit Tigers – Jose Valverde
Valverde finished with 49 saves (to lead the league) in 49 chances, a 2.24 ERA and 1 measly WAR to go along with his 4.17 wins added. By contribution, he was the most valuable reliever in the league (though 4.17 wins isn’t historically high). By peripherals, he certainly didn’t appear to be the best reliever in the league. In part he pitched best when it counted (which is what a closer is supposed to do) in part he got lucky by sprinkling hits and walks and working out of trouble. Among AL closers, Valverde’s ERA was 3rd best. But… his strikeout rate was only 9th of 14 and his walk rate was 3rd worst (and the top guy, logically, lost his closer role as a result). There is nothing wrong with Valverde’s number regarding balls in play – his BABIP was low, but his career BABIP is also low. He does a solid job of getting ground balls, for a power pitcher, and keeping the ball in the yard. Nonetheless, a K/BB ratio of 2 isn’t enough to keep him in the top tier of closers, much less best in the league. If anything, Valverde’s 2011 season is a lesson in the variance inherent in a reliever’s job and perhaps the importance of luck over raw ability. That said, Valverde may actually pitch better in 2012 though he’s unlikely to get as lucky.
Jonathan Papelbon – Boston Red Sox
Papelbon finished with only 31 saves and a 2.94 ERA, 3 WAR and 3.53 wins by WPA. It would seem by those measures (aside from WAR, that tries to control for things like BABIP) that Valverde clearly had a better season than Papelbon and with that I would agree. As to who’s the better pitcher? Well, Papelbon had a highish BABIP last year – but more importantly he didn’t do as good of a job as Valverde in making those critical pitches to keep runners from scoring with a LOB% under 70. But those other peripherals? He struck out about 40% more batters than Valverde last year and walked only 1/3 as many. If Papelbon had not given up some of those key hits with runners on, he would likely have topped 5 wins by WPA for the third time in his career. If I had to guess who would pitch better in 2012, I’d put my money on Papelbon – the Red Sox may be playing with fire in not resigning Papelbon early – though I suppose that he, like every other player on the roster, is bearing the burden of blame for their September collapse.
As much as I like Papa Grande, and as great as the 2011 season was for Tigers fans and Valverde personally. I have to agree with the Snyders here – he’s a good closer, but he’s not an exceptional closer and his peripherals look more like a 1-2 WPA guy than a 4-5 WPA guy. And while luck is a big deal in relief, it’s also such a highly leveraged position that getting the best of the best means a lot. Flipping him for a Papelbon and a sandwich pick is an idea I could have gotten behind. Oh well, maybe next year – his deal will be up again in a year.
For the rest of the league, I’ll just give you the table and you can make your own judgments
A few things bear mentioning: for 5 teams, the closer was the only positive thing in their bullpen with a negative WPA total for all other bullpen slots combined. One of those 5? The Texas Rangers. Neftali Feliz was unexceptional as a closer over the course of the season, but the rest of the Rangers ‘pen was pretty foul. That is, of course, a problem they fixed at the deadline by acquiring a number of top-tier relievers for the stretch run – and the rest is history. A 6th team also had negative WPA from the non-closer ‘pen, but they don’t really count since their primary closer was just as bad. Poor Twins.
Four of those teams had a significant split during the season between two different closers, the A’s, O’s, Twins and Jays. The numbers above reflect only the guy who got the most saves, who isn’t necessarily the one who started the year in the role or the one who finished it. A couple of other teams had a different closer early on, but yanked him early in the season, like the White Sox.