When it comes to evaluating prospects or looking at roster make-up, I’ve always been a little biased against relievers. It’s not that I think all relievers are replaceable, but in the short-term of baseball history, what teams have paid for and what they’ve gotten has been grossly out of proportion when it comes to relief arms. We’ve seen the almighty save stat loose some of it’s luster, but teams generally still spend a good bit on relief arms. Somehow Brandon Lyon got 3 years and $15M from the Astros this offseason. Sure, Ed Wade isn’t the best GM in baseball but other teams are still (in recent memory) spending big money on relievers. The New York Mets are paying Francisco Rodriguez nearly $11.5M per season to close for their baseball team on a deal he signed over the 2008 offseason. K-Rod has been a dominant reliever in his career, but even the best relievers in baseball (the elite guys who are closing) only rack up 2 wins per season above replacement. And that’s just a handful of relievers. Unfortunately for the Mets, Rodriguez struggled this past year posting a FIP nearly 1 entire run over his career average this season (4.01 compared to 3.00). His control worsened, but most of his other pitching ratios are near career levels. The Mets are hoping he’ll bounce-back this season but I digress.
My feelings about relievers stem from the the following:
~ Generally speaking, anyone can relieve. The list of starters-turned-to-relievers is long. You rarely see anyone go in the other direction from reliever to starter. Therefore, the potential population to draw relievers from is enormous.
~ They just don’t pitch that much. Having a relief ace is great, but he’s only going to pitch in 60-some innings for your team. There just isn’t a lot of available PT for a reliever to influence a game. I think you could make the case for shifting back to multiple inning usage, a’la Goose Gossage, but with specialization in the ‘pen, I’m not sure that’s going to happen anytime soon.
Because of these two reasons, I have a hard time justifying paying the going rate for relievers, which admittedly has taken a hit lately. You can still find good buys (Jeremy Affeldt being one of them) but in general, paying for relief help through the FA market seems risky. Tom Tango in his seminal post on calculating WAR ran the numbers behind roster construction and found that by WAR, you should only use around 10% of your entire payroll on relievers. That’s by far the smallest percentage of any portion of your roster. It’s even lower if you consider pitcher risk.
Switching gears a little, after trying to think about how to sum up my feelings re: relievers, I then started to think about how teams manage their bullpens. We’ll often hear the phrase in the press or on the TV that a certain manger “really manages his bullpen well.” Managing a bullpen can be a catch-all term but to me it normally means when in-game a manager calls on his relievers. You want to use your best reliever in the best possible spots based on leverage — or the state of the game. We give relievers credit for working in tight late game situations. Example: Brian Wilson comes in a 1 run game in the 8th innings, 2 outs, and the bases are loaded. That’s a much higher leveraged situation than Brandon Medders working mop-up in the 4th inning of a blowout. Wilson will get credit for his hard work and Medders will not. Seems fair, right? In theory, managers will use their best relievers in the toughest leveraged situations. It makes no sense to bring in Wilson during the 5th inning when the Giants are up by 6 runs.
Therefore, I wanted to look at every qualified reliever (by FanGraphs) by 2009 and examine their FIP vs the average leverage index they pitched in. The theory remains that the higher the leverage, the better the reliever should be. Nobody wants the worst member of the bullpen pitching the 9th inning in a close game.

There seems to be on obvious relationship between pitcher quality and use. Managers aren’t going to throw a 5.00 FIP pitcher into the 9th inning and expect him to wiggle out of tight jams. You can see a few relievers that posted very high FIP scores but still worked very high 2+ leverage situations*. The players who posted poor FIP’s but were still give high leveraged situations are the Established Closer ™ types. Brad Lidge had a down year posting a 5.45 FIP but he still pitched in tough situations. Brian Fuentes was very similar. The graph also shows how good Brian Wilson was last year. He pitched in extremely high leveraged situations while posting a FIP of 2.50. His success by FIP combined with his usage adds up to a 2.4 win season. For a reliever, that’s an incredible sum. Brian Wilson’s work out of the bullpen was one of the highlights of the Giants’ 2009 season. But, as his price rises through arbitration, the Giants might want to start combing their system for their next great relief arm.
*A quick word about leverage, 1.0 indicates a neutral situation. Closers tend to work around 1.8 and guys who are mopping up will be below 1.0. Read more about leverage at FanGraphs.
Phil Hughes’ success as a reliever has been well documented but despite his sub-2.00 FIP score, he was leveraged pretty low. You can chalk that up to team composition. When you’ve got Mariano Rivera in front of you, you’re probably not going to get the high leveraged situations. Also of note is Luis Perdomo, former Rule 5 pick by the Giants. Perdomo is notable not because he performed well for the Padres, but because his average leverage index of 0.28 was by far the lowest of all our relievers. In order to keep a Rule 5 pick, he must remain on your active roster for the entire season and it’s pretty obvious that the Padres hid Perdomo in their bullpen, only letting him out when the situation allowed.
Managers seem to have an idea of who their best reliever is and how they should use him. Which makes me think that the idea of managing a bullpen is overrated to an extent. Should we give credit for a manager realizing who his relief ace is and when to use him? Even if managers are judging relievers by things like ERA or scouting reports, it seems that by and large, they make the right call. Which makes me think, is it all that hard to do?