Over at Fangraph's, Jack Moore has begun examining the league's historical approach to closer usage. This is a very interesting (and quick!) read, well worth the time.
In the Sabermetric community, closers are often targets of spite and sighs. Reasons for this include: A) closers usually get exorbitant contracts (for relievers) yet too-often pitch in low-leverage situations (the textbook "save situations," which sometimes may be as easy as merely getting a pitch over the plate so that the shortstop or pinch-hitter can wildly tomahawk it into the ground for an easy out) and B) managers and traditionalists too often evaluate closers on their ability to "save" a lot of games or produce a high "save" percentage -- never mind that the closer may actually be doing a poor job at possibly the second easiest spot in the bullpen (according to leverage, *cough cough* Fernando Rodney *cough*).
Anyway, I think Moore does a good job to examine how we've reached this point of over-value. Give it a read!