How do the 2012 and 2013 Orioles bullpens compare?

I've asserted in several of my recent posts that the 2013 Orioles are a better team than the 2012 squad, even though they're clearly getting inferior results. In one of my most recent posts, I explored how the team was getting much better hitting and only marginally worse pitching, but with a worse record overall to show for all of it. I want to break that down a step further today and talk about the Orioles' bullpen woes specifically.

A huge part of the 2012 Orioles' historic efforts in one-run and extra-inning games was a shutdown bullpen. In going 29-9 and 16-2, respectively, in such contests, the key bullpen arms of Jim Johnson, Darren O'Day, Pedro Strop, Troy Patton and (later in the season) Tommy Hunter and Brian Matusz seemed capable of coming up with the absolute biggest outs whenever they were needed. You can't win so many close games without a stud bullpen, and that's exactly what the 2013 Orioles had.

The "close game" argument and the "lucky" argument go hand in hand. The same folks predicting that the Orioles would regress to the mean in close contests predicted all along that the team's bullpen would regress to a more normal performance. Relief pitcher performance from year-to-year is one of the toughest variables in baseball to predict. Outside of a once-in-a-lifetime talent like Mariano Rivera, or even once-every-five-years-or-so talents like Troy Percival, Trevor Hoffman and (so far) Craig Kimbrel, teams just don't get a lot of relievers that they can count on for filthy performances year in and year out. If they were that good, the logic goes, they'd be starters.

The predictors of regression have had a mixed bag. The Orioles are indeed worse off in one-run games, but they haven't regressed to anywhere near the overall talent levels that various projection systems have saddled them with (the team was supposed to play .500 ball or lower this year, remember -- so even if it seems like a disappointing season, they're still outperforming expectations). But let's put aside all of the general discussions of luck, talent, management and regression and focus just on the team's relief pitcher corps. It quickly paints a stark picture of what might have been this year (all stats through Sunday's play):

Taken in aggregate, the numbers make it clear that the 2013 bullpen has been a bit less lucky, but also gotten hit a lot harder. These numbers also don't account for leverage, of course, and we really don't care about the relative fortunes of Randy Wolf and Dana Eveland versus T.J. McFarland and Josh Stinson. But I'm willing to assert that the splits would be as bad -- if not worse -- if we drilled down solely to Johnson, O'Day, Hunter, Matusz, Strop and Patton. Just take a look at Baseball Reference's "Close & Late" stats for the two bullpens:

Same story, but far, far more pronounced -- luck is at play, but there's an element of getting straight-up rocked, too. And the team is giving up pivotal home runs like they're going out of style.

Bullpen regression is a pretty hard thing to plan for. It's easy to predict that a pitcher like Pedro Strop, who clearly never quite knew where the ball was going, was a ticking time bomb, and plan around him. It's harder to imagine, in the same year, that Darren O'Day might forget how to pitch to lefties (.657 OPS in 2012, .945 in 2013), Jim Johnson would blow nine saves (and counting), Tommy Hunter might rediscover the art of yielding home runs (2 HR as a reliever in 2012, 10 and counting in 2013), and Troy Patton would flat-out get worse -- all in the same year. A stray trade for Francisco Rodriguez isn't going to make up for those kinds of problems.

The Orioles are still a promising young squad with a core of hitters and starters in or approaching their prime. Hell, they may even sneak into the playoffs this year despite their many frustrating tendencies. But regardless, the approach to the bullpen in the coming offseason will be an interesting one to watch. The Orioles have a less-than-stellar recent history with high-priced free-agent relievers (need I mention Kevin Gregg, Mike Gonzalez, Danys Baez, Steve Kline and company?), but it might seem dicey to roll with the current crop and just assume that they'll fluctuate back toward excellence.

I've always advocated -- and continue to advocate -- picking up as many cheap relievers as possible and making the best ones earn their job, year in and year out. In fact, the Orioles did just that in 2012. If you'll recall, the Orioles had such a surplus of relief talent in spring training in 2012 that they let invitee Pat Neshek walk after a scoreless spring, and go on to have a ridiculous (286 ERA+) season for the Athletics. It may have been their very complacency in 2013 that sentenced the team to its bullpen regression, with no clear alternatives to readily slot in for Strop or Patton, the most obvious candidates for regression coming into the season.

Dan Duquette has been excellent at under-the-radar reclamation projects throughout the Orioles roster. Hopefully, he can turn his offseason focus in this regard to the bullpen, and pick up a few surprise performers for the 2014 campaign. If the team can hold on to even a normal percentage of its close contests (unlike this year, where it's dropped well over half of them), they'll be able to contend next year and beyond. They just shouldn't plan to count on the same guys to hold those leads.