I have an annual valuation problem on RotoWire. After running the projections for all of our pitchers and then turning those projections into values, the closer prices inevitably come up short, considerably so in some cases, when compared to what they'll cost at the auction table. In many cases, they'll even come up short compared to other relievers that aren't their teams' respective closers.

The problem derives from the volatility of the closer's role, plus the fact that so much of that value from the closer comes from just one stat, the save. In any given season, a third of the closer jobs will have changed hands, and many other teams will have used multiple guys to pick up saves at some point. Yes, the elite relievers will provide a boost in ERA and WHIP, even with their limited innings, and occasionally a handful of closers will strike out enough batters to be helpful in that category. But more often than not, closers are neutral at best in those categories, because they throw just 60-to-70 innings. And rarer still is the relief pitcher that can help you in the Wins category, let alone a closer that can do so. Meanwhile, relievers are extremely volatile in terms of their performance – Mariano Rivera is a freak of nature. There's not a set of "Mariano Rivera's" – there's just the original, when it comes to repeatability from year-to-year. More often than not, you're lucky to get two-to-three solid years of performance to bank on, let alone four or five.

The problem gets exacerbated by our projection schedule. We do our first set of projections for the season back in December to get them in time for our annual magazine. In many cases, teams haven't signed their putative closers yet and elite relievers haven't signed, as was the case with Rafael Soriano and Brian Fuentes. Once both pitchers signed with teams with incumbent closers, we had to drastically alter their respective projections. Even now, there are teams that we really are just guessing about how they'll use their bullpen, like the Rays. Other teams have resolved their closer situation, at least for now, like the Pirates, who named Joel Hanrahan their closer over Evan Meek. But just how much stability does he have in that role, given his struggles when given the opportunity with the Nats in the past? Can we reliably say he's going to get 35-to-40 saves? Are we fudging reality by giving every incumbent closer full saves, just so the valuation formula and the cheatsheets list those pitchers accordingly?

Consider Francisco Cordero, the Reds' incumbent closer. He has had 34 or more saves the last four years, but his performance last year suggests that he might not be the most stable closer for 2011. His strikeout rate dropped to a career-low 7.31 batters per nine innings, while he walked 4.46 batters per nine, an unacceptably high rate. He's in the final year of his four-year deal with the Reds, a team that's expected to contend again in the NL Central. There's no contractual reason to stick with him if he flops, and there's more of a sense of urgency for the team if he struggles. His manager, Dusty Baker, has been very patient with him in the past, but if the Reds are sitting at .500 in June and Cordero has blown 4-to-5 saves, a change could happen. And if that's the case, the Reds have two excellent alternatives, including one that throws as hard as 105 mph in Aroldis Chapman. The other, Nick Masset, actually bailed Cordero out in a couple of save situations last year, most memorably in Chicago when Cordero lost the strike zone. Given those set of facts, it's hard to give Cordero a full 35-to-40 saves in his projection, let alone a favorable ERA or WHIP projection. Strictly going by the values generated from his projection, Cordero is not a top-30 reliever – not even close.

But at the same time, do we want to have the dollar values and cheatsheet reflect what's going to happen on Draft Day? Or should we emphasize skills over roles? I engaged in a conversation with RotoWire Managing Editor Chris Liss on this topic yesterday, and while I've conceded that we do have to make some of those changes to make it closer, a day like Monday happens. Four teams had major news regarding their respective closer jobs:

- The A's Andrew Bailey walked off the mound in a spring training game in Arizona against the Indians, complaining of elbow pain. This comes on the heels of Bailey's season ending early last year with an arthroscopic elbow surgery. He was slow to get going this spring, though prior to Monday the news was good. The late news on Bailey is that he has a forearm strain, but no structural damage. Bailey right now won't need surgery, but it seems certain that he won't be ready for Opening Day. Incidentally, this news broke 30 minutes before the RotoWire Staff League reserves and minor league draft. One of my competitors, blog contributor Dalton Del Don, snagged Bailey's replacement (at least, his most likely replacement – more on that in a second), Brian Fuentes, one spot before me in that draft. The A's situation illustrates the other problem with projecting relievers – Fuentes is the leading candidate to replace Bailey, but he's far from the clear cut only solution. The A's spent good money to bring in Grant Balfour this offseason, and they have three other quality relievers that have saved games for them over the last two years – Brad Ziegler, Michael Wuertz and Craig Breslow. Any one of these five relievers theoretically could step in while Bailey is out. Good luck getting an accurate projection on the number of saves from each reliever in the Oakland bullpen.

- The Rangers' Neftali Feliz reversed his recent opinion about his role and now said that he prefers starting over closing. The Rangers have been stretching him out in spring training as a starter (which is what he was in the minors), but holding out the possibility that he still closes. As a closer, Feliz is possibly a top-five reliever. But how well will he hold up as a starter? When will the Rangers ultimately decide his role? How do you accurately project that performance? Essentially you have to use two sets of projections – one as a closer, one as a reliever. Unfortunately, that doesn't really work on draft day, especially if your expected dollar value outcome diverges materially pending his role. Meanwhile, the Rangers don't have a clear successor in place, and manager Ron Washington has already expressed his desire for the organization to bring a veteran from outside the team to step in as the closer if Feliz starts.

- The Nationals' Drew Storen had another bad outing, and the team suggested that they'll use a closer committee to begin the year, one that might not include Storen, and might not even leave him in the majors. If Storen does get sent to Triple-A, how long does he stay there? What will his role be upon his return? Does the team use an actual committee, or does manager Jim Riggleman give the ball to either Tyler Clippard or Sean Burnett first, see them succeed, and keep handing them the ball in the role? What if the replacement flourishes even when Storen is ready to return – could Storen get Wally Pipp'ed?

- The Phillies' Brad Lidge was diagnosed with biceps tendinitis. He and the team have downplayed the injury, saying that he could pitch if this were the regular season. But he's already been scratched from a scheduled spring training game Thursday, and of course his season was marred by elbow problems in 2009 and delayed last year, before he was finally whole. Not only is Lidge's availability threatened by this injury, but perhaps his performance will be too. If that's the case, all of a sudden you have a closer who actively hurts you in the ratio categories again.

So ... what do we do about this problem? I'm inclined to stick closer to the status quo – when valuing closers, put more of an emphasis on the pitcher's skill than the role. There has to be some concession to draft day value, but my inclination is not to increase the value of all closers, but instead to value the premium closers more. I'll pay that extra dollar for Rivera or Brian Wilson, rather than paying the sticker price for Cordero or the other closers. It's a risky strategy in some respects, as I could get left without a closer, but the residual effect is that I'll also have more draft day dollars to spend on more reliable qualities.