WAR does not love the reliever. A quick stop at Fangraphs will show that in 2013 the best bullpen in the game--that of the Texas Rangers--was worth all of 7.6 fWAR. That's quite a bit, but it also covers about a third of the team's roster, leaving each individual reliever worth relatively little. Joe Nathan, far-and-away their most productive reliever and the fourth most valuable bullpen arm in the game last year, came in as just the team's seventh most valuable player overall. Put in context of the league as a whole, he was worth about as much in WAR as Nate McLouth, Not a glowing endorsement.
The Red Sox, meanwhile, are not looking to replace small contributions. Even accounting only for players that are guaranteed not to return--Jacoby Ellsbury and Jarrod Saltalamacchia--the Sox are trying to recoup 9.4 fWAR. That's nearly two wins above what that league-best Texas bullpen managed as a whole last year. So when they ended up signing a player like Edward Mujica while there were more "important" parts of the roster to shore up, some fans may have felt they did not have their priorities in order.
There's another way to look at this, though. I don't think there's much question the Red Sox were not enamored with this crop of free agents. They had their eyes on Carlos Ruiz, but other than him it was a market distinctly lacking the sort of deals that really personified their fantastic 2012-2013 offseason. There was no way to sneak in 3-5 WAR players without really paying for them in both years and dollars.
So what's a team to do? In Edward Mujica, we may see the Red Sox hedging their bets by dipping into a slightly different strategy. In 2012, the Baltimore Orioles returned to the postseason for the first time in 15 year, perfectly reversing their 69-93 record from 2011. They did so despite producing a run differential of just +7. No other playoff team had a figure below +56, that coming from the Detroit Tigers in an exceptionally weak AL Central.
Photo Credit: Dustin Bradford
How did they do it? How did they outperform their expected win - loss record by a full eleven games? A ridiculous 29-9 record in one-run games, and a 16-2 performance in contests that went to extra innings. The Orioles bullpen may have come in fourth in WAR that year, but just looking at it, their top-five innings producers had ERAs ranging between 2.28 and 2.64. To make a long story short, when the Orioles needed someone to pitch an important inning, they never had to accept a lesser option.
There's something to be said for the idea that a bullpen does more than just what's shown in WAR. A hitter might pick up some big hits in a blowout going in either direction, contributing context-independent wins, but little actual value when it comes to improving the team's record on that given day. A good reliever, however, is mostly left out of those games. A full 60% of Koji Uehara's plate appearances, for instance, came in "late and close" situations (defined by Baseball-Reference.com as coming in the seventh or later with the batting team tied, ahead by one, or with the tying run at least on deck) where any sub-par performance can have a huge impact on he outcome of the game.
There's some level of diminishing returns with this, likely. Boston's setup crew in Junichi Tazawa and Craig Breslow were not always lavished with praise throughout the season, but they were actually quite good on the whole. Having someone like Mujica in town would be even more important if we really didn't have anyone to rely on after Koji, but it's certainly not like the Red Sox were ever really flush with relievers.
Even discounting the three-ring circus that was Bailey and Hanrahan in the early months--they alone managed to blow seven saves in just 39 games--the Red Sox were very often scrambling for answers, even in important games. Who can forget Brayan Villarreal's only stand? The trials and tribulations of Matt Thornton? Alex Wilson may yet be a good reliever, but last season he had 33 "late and close" plate appearances despite an ERA of 4.88. Clayton Mortensen, Franklin Morales, and Drake Britton combined for 13.1 "high leverage" innings by Fangraphs' numbers. And before you discount that as a small number, Fangraphs only recognizes about 150 such innings thrown by Red Sox relievers and starters alike last year. That's nearly 10% going to some really questionable pitchers.
Now, in 2014, we're looking at a bullpen with an excellent top-end that's deep as can be to boot. Koji Uehara, Edward Mujica, Junichi Tazawa, Craig Breslow, Andrew Miller, Burke Badenhop, with a final spot going to Ryan Dempster as a swing man or Brandon Workman, who really came into his own last year, should the need arise. There's not a single name in there that's really cause for concern in the way so many of last year's options were. Yes, there's always potential for injuries, but even in this the Red Sox should be better off, with the replacement options much the same as last year, just with that much more time to adjust to a relief role in the minors.
Are the Red Sox going to win 75% of their one-run games next year? No, but with a bullpen as good as this one, they might be looking at a better figure than 21-21 like they managed last year. That 97-65 record actually came in three wins beneath the 100 predicted by their +197 run differential.
No, the difference between the 2013 and 2014 editions of the bullpen is not going to be worth that 9.4 fWAR lost. Stepping away from "value," it's not going to actually win them nine more games either, most likely. But it can certainly help to offset the blow. Add in the other changes--the addition of Xander Bogaerts and Jackie Bradley Jr., for instance--and it gets a little easier to see how the Red Sox could repeat as World Series champions.