Pat has a good review of the main hurlers in the Bucs' 2006 bullpen here. The bullpen was a relative strength for the Bucs this year, and the team controls every good pitcher in it next year. They also have a number of young arms (Jonah Bayliss, Josh Sharpless, Jesse Chavez and possibly Brian Rogers and Dave Davidson) who played most or all of 2006 in the minors but who may be able to contribute to the big-league pen in 2007.
So, what should the Pirates do? Well, I've already said that I think they should trade most of the top performers in their bullpen, but today I'd like to explain a few things about why I think that what I think that should mean.
Reliever performance is highly variable from year to year. Baseball Prospectus once showed this by looking at the top relievers over a period of several years and seeing how they performed a few short years later. It wasn't pretty - most top relievers experienced a steep decline relatively quickly, whether due to aging, injuries or plain old ineffectiveness. For every Mariano Rivera or Trevor Hoffman who's rock solid year in and year out, there are several Eric Gagnes, Brad Lidges and Octavio Dotels. I don't think BP looked at this, but I suspect this is at least as true, and actually probably even more true, for relievers who aren't elite performers like Rivera.
For anecdotal (and painful) evidence that reliever performance fluctuates a lot from year to year, I present to you the 2002 and 2003 Pittsburgh Pirates. The '02 Pirates had a very good pen. Mike Williams, Scott Sauerbeck, Brian Boehringer and Mike Lincoln all tossed at least 60 innings and had ERAs below 3.50. The Pirates won 72 games that year.
In the 2002-2003 offseason, the Pirates hoped Brian Giles's bat, a decent group of young starters, a good group of extremely cheap free agent signings, and their bullpen would help the team return to .500. So they kept all the main players from their good 2002 bullpen and added Julian Tavarez.
We all know what happened next. Giles was very good before being traded in August; the decent young starters were, well, decent; the cheapo free agents were much better than anyone could have anticipated. And yet the Pirates won only 75 games because the bullpen couldn't get it done. Williams was absolutely horrible and he ended up being traded to the Phillies for Frank Brooks; Boehringer was a complete disaster; Sauerbeck was worse than usual; and Lincoln was awful. The Bucs' bullpen ERA went from 13th in the majors in 2002 to 27th in 2003.
The current core of relievers the Pirates have is better than Williams, Boehringer, Sauerbeck and Lincoln were, but they're hardly sure bets to be good next year, much less whenever the Pirates could potentially be competitive. Mike Gonzalez is great but injury-prone. Salomon Torres is in his mid-30s. Matt Capps is promising, but he almost just set a record for appearances in a season by a rookie. And John Grabow just isn't that good.
The Bucs would do well to trade as many guys as possible. They should also try to trade Damaso Marte - Marte will essentially cost the Bucs $2.75 million next year if they exercise his option, and as WTM points out, he should have trade value for someone. The Pirates should know that, too - they got Giles, Freddy Sanchez and Jack Wilson in exchange for three of the pitchers on Marte's comparables list, and that's even before considering Marte's excellent stuff. (Okay, that's not entirely fair - Jeff Suppan was also involved in the Sanchez trade, and Ricardo Rincon was younger than Marte when he was traded for Giles. But I think you can see what I mean.)
Because the Pirates have good relievers in the majors and a number of youngsters ready to fill in, they really do stand to gain this offseason by trading from a position of strength. They would be especially wise to do so because of the extreme volatility of reliever performance. We'll see whether the Pirates can learn from what happened to them in 2002 and 2003. I'm guessing they learned very little.