The Bullpen Puzzle

So the Red Sox were widely known to be targeting relief pitching this offseason. Along the way they picked up some hitters, but the 'pen remains a big mess.

Meanwhile, the two relievers most linked with the Sox — Scott Downs and Matt Guerrier — have accepted offers elsewhere. Not coincidentally, both those contracts are guaranteed three-year deals, and the Sox have been reluctant to offer the third guarantee year.

It's not hard to see why. Here is a list of the best relievers in the Theo Epstein Era (130 ERA+ or better, an admittedly crude measure of reliever effectiveness, in at least 25 appearances) in chronological order:

2003

• Byung-Hyun Kim 
• Mike Timlin

2004

• Keith Foulke
• Scott Williamson
• Curtis Leskanic
• Ramiro Mendoza

2005

• Mike Timlin
• Mike Myers

2006

• Jonathan Papelbon
• Javier Lopez

2007

• Jonathan Papelbon
• Hideki Okajima
• Mike Timlin
• Javier Lopez
• Manny Delcarmen
• Brendan Donnelly

2008

• Jonathan Papelbon
• Manny Delcarmen
• Hideki Okajima
• Javier Lopez
• Justin Masterson

2009

• Jonathan Papelbon
• Ramon Ramirez
• Hideki Okajima
• Takashi Saito

2010

• Daniel Bard

How many of the 17 relievers listed here performed at a high level for three consecutive seasons? Three. Papelbon, Okajima, Lopez. That's it, and really, Lopez is an example of why ERA+ probably isn't the best stat to use. He was notorious for somehow ending up with a good ERA despite not being very good at throwing strikes, and more advanced linear-weights measures indicate Lopez was in fact only an above-average pitcher for two seasons (2007-08) and excellent only in 2008.

Only Okajima and Papelbon, then, have posted three consecutive excellent seasons as a reliever in the past seven years, with Mike Timlin managing three nonconsecutive great years and an overall impressively consistent career. But that's why — other than Kim, Foulke and Papelbon — they're middle relievers, right? If they were good enough to start, they'd be doing that. If they were good enough to close, they'd be doing that. They are the fringiest of the fringe pitchers, and as a result, they are just hanging on.

So a three-year deal is not a particularly appetizing prospect for a middle reliever — especially not one, like Downs, who turns 35 in March and sported the lowest strikeout rate of his relieving career in 2010 or one, like Guerrier, who is 32 and saw his own strikeout rate drop to 5.3 per 9 in his fourth consecutive year of 70-plus appearances.

The Sox apparently sill have an offer out on Brian Fuentes, who will turn 36 in August and has yet to be anything but a closer. Transferring to middle relief might help hide the fact that he has stopped being as effective against right-handers, though I'm not sure anything can hide for much longer the fact that he posted a bounceback season in 2010 thanks to an unsupportably low BABIP (.217 on a career line of .286, whereas in 2009, with a .296 BABIP, he posted a 1.40 WHIP).

So if relievers are commanding three-year contracts, and the Sox are not giving them out, what's a team to do? I hate to say it, but the throw-crap-at-the-wall approach has worked before, much as it's generally derided.

For example, the '07 bullpen was awesome, thanks to the surprising debut of Okajima, the development of Delcarmen, the trade of David Riske for Lopez, one final gasp from Timlin, and a handful of good performances from Donnelly before he lost most of the season to injury. Along the way, the Sox got a decent performance from castoff Kyle Snyder, while enduring the J.C. Romero, Joel Pineiro and Eric Gagne experiments.

The 2004 bullpen was similarly built from castoffs and spare parts, yet performed admirably throughout the season then slammed the door on the Yankees en route to the Promised Land. The 2003 "closer-by-committee" bullpen, a failure mostly because of the manager running it, stabilized with the arrival of Kim and was so good in the playoffs that it made the infamous Pedro Martinez decision even more incomprehensible.

In fact, 2010 is an example of a bullpen that didn't use that approach: Papelbon, Okajima and Ramirez were coming off good seasons, Delcarmen was recovering from an injury, and Bard was the hotshot rookie coming off a great debut. The rest of the pen was the usual mix of talent — young guys like Dustin Richardson and Robert Manuel, old guys like Joe Nelson and Scott Schoenweis, prospects like Michael Bowden and Felix Doubront — with the usual mix of results (some good, mostly bad), but the top five arms were holdovers from previous seasons who had performed well. That four of them were deeply disappointing is unfortunate, but doesn't really speak to the merits of the castoff-gathering approach — rather, it does for sure speak to the riskiness inherent in asking middle relievers to repeat their performance from year to year, never mind three in a row.

It seems likely that Epstein's best solution to the bullpen problems lies in two places: The return of Jonathan Papelbon to performance closer to his 2009 than his 2010, and the emergence of a player we probably haven't even discussed in this post.