Imagine pouring yourself a glass of milk without knowing its expiration date.
Such is the gag-inducing reality of relief pitching in the majors. At any moment the pitcher can turn sour -- or already has, and you just don't know it yet.
Many teams find themselves in the market for relief help. Teams are always searching for relief help for a simple reason: There are not enough good relievers to go around.
If your team is lucky enough to get one of the good ones, pat yourself on the back and get back to looking, because almost no reliever is good from one year to the next.
A principal way that Baseball Prospectus evaluates relievers, for example, is a statistic called Adjusted Runs Prevented. This statistic takes you a step past ERA, whose usefulness is at the mercy of inherited runners, to paint a more accurate image of how effective a relief pitcher is.
Of the Top 100 in ARP at this year's All-Star Break, 53 had not been on the list once in the previous four seasons. Essentially, more than half of the Top 100 relievers in baseball at midseason had come out of nowhere.
In addition, looking back at the 2005 Top 100 relievers, 68 have dropped off the list half a season later. Only 32 pitchers -- averaging out to one per team -- were in the Top 100 in both '05 and '06.
Think about what this means. There are roughly 200 relief pitchers in Major League Baseball at any given moment, so anyone not in the Top 100 could be pegged as a below-average reliever. That averages out to three decent relievers per team with a couple handfuls left over.
So, of the 2006 Top 100 relievers, about two thirds are not as productive as the average No. 3 reliever less than a year ago.
It's important to emphasize that this is not just a change at the bottom of the list. The 2005 Top 100 includes only five of the 2006 Top 10, nine of the Top 20 and 26 of the Top 50 -- about half the group each time.
Cliff Politte of the White Sox was baseball's No. 4 reliever in ARP in 2005, a key cog for the World Series champion. This past weekend the White Sox designated him for assignment. In ARP he had fallen to dead last among 570 relief pitchers.
Who knows -- we may not have seen the last of him.
This year's turnover in the Top 100 is not a fluke. Over five seasons, from 2002 to '06, 304 relief pitchers have appeared in the Top 100. Of that group, 193 -- again, about two thirds -- made the list in just a single season. Fifty-five did it twice, 34 did it three times, 17 did it four times and five have been on the list for five consecutive years: Francisco Cordero, Jason Isringhausen, Mariano Rivera, Scot Shields and Billy Wagner.
Finding a relief pitcher you can count on is like a dartboard toss. Track records are almost completely useless. When you think about it, it makes sense -- with only a few exceptions, relief pitchers are failed starting pitchers, showing at some point in their careers that they lacked the repertoire and/or consistency to survive a series of lengthy appearances. A move to relief pitching is a way of making lemons out of lemonade -- temporarily.