In the fantasy baseball universe, relievers have a somewhat tarnished reputation as being fickle contributors to team success. Of course, in the chase for saves and sometimes holds, team owners are willing to forgive, if not quite, forget, the capriciousness of a reliever.

Don’t believe the hype. Sure, relievers are prone to lose their job. And relievers who allow multiple runs in a single team-ERA-killing outing deserve all the scorn they get. But let’s also put some blame on the backs of those in the fantasy world who roster relievers without full consideration of their overall skills, regarding them as one-category machines pure and simple: Live by the save — die by it.

But are relievers really that unpredictable? Maybe not.

From 2005-2007, 73 relievers completed at least 20 saves. (Some relievers are counted more than once as having made 20 saves in more than one season.)  Of those 73 relievers, 60% of them followed up a 20-save season with another. That’s not a very good percentage and would seem to support those who think that closers are very volatile.

However, we can break down these relievers into two categories: Those who struck out at least eight batters per nine-innings and those who didn’t.

Among the latter group, only 35 percent followed up a 20-save season with another.

Among those with a 8+K/9 IP, the completion percentage jumps to 74 percent. A good percentage of the remaining relievers who didn’t make it to 20 saves again, such as Takashi Saito, and JJ Putz last year or B.J Ryan and Mike Gonzalez in 2007, failed to do so because of injury.

Is a 74% completion rate, with injuries, any different than the attrition rate for fantasy #2 or #3 starters? We’ll let others do that research.

In the meantime, it may be best to take a much more nuanced approach to relievers going into an auction or draft. Even if you don’t believe that relievers contribute to overall team ERA and strikeout success, the ability to maintain the closer’s job is dependent on the ability to get the job done. The takeaway may be that you lay off Bobby Jenks (5.6 K/9), Matt Capps (6.6), and Troy Percival (7.6), even though they seem to be headed into this season with closer gigs. And conversely, lock up Chad Qualls (8.7), Frank Francisco (11.8), and Carlos Marmol (11.8). These guys may have more tenuous holds on the closer position, but they have better skills to not be so damn capricious.