It's been somewhat forgotten in the euphoria of the Phillies improbable late charge to the 2007 division title, but through four months of last season, the team's biggest problem was unquestionably the bullpen. Injuries to Tom Gordon, Brett Myers, Ryan Madson and others, stomach-turning incompetence from the likes of Antonio Alfonseca and--I still can't really believe it--Jose Mesa, and the always-present danger that Charlie Manuel would decide it was time to see what Anderson Garcia could do all rendered perilous the late innings of games through early August.

Then everything fell into place. In 20 September appearances, J.C. Romero put up an ERA of 0.00. Brett Myers, Tom Gordon and Geoff Geary were all almost as good. Even the likes of Francisco Rosario and Clay Condrey came up huge down the stretch. The team's biggest weakness was, when it mattered most, its greatest strength.

What does this mean for 2008? My guess is that the bullpen will be neither as bad as it was through most of 2007, nor as strong as it was in that season's final month.

The cast of characters looks to be similar to the group that completed last season, with the obvious difference of a new leading man. November's acquisition of former Astros closer Brad Lidge allowed the Phillies to return Myers to the rotation, where he presented a better (and cheaper) alternative to any starter on the free-agent market. Over his career, the 31 year-old Lidge has put up cartoonishly great numbers some years (in 2004, his 157 strikeouts ranked 15th in the NL... in 94 innings!) and Jorge Julio numbers (tons of strikeouts, but an ERA over 5) others.

The concern about Lidge is his psyche, which purportedly suffered permanent damage when Albert Pujols took him deep to win Game Five of the 2005 National League Championship Series. (Usually forgotten in this telling is that Lidge's Astros won the next game, and the series.) His home runs allowed are a little high for a closer, and OFJOAB won't help with that. But if ever a manager seemed well suited to buck up a closer's shaky confidence, it's Charlie Manuel. My guess is that Lidge will be just fine, turning in a great last season before free agency and leaving the Phillies with two nice picks in the June 2009 draft. Baseball Prospectus' PECOTA system projects the following line for Lidge in 2008: 4-4, 3.54, 17 SV, 60.2 IP, 1.27 WHIP.

J.C. Romero and Tom Gordon are slated to be the setup men for Lidge. Both guys come with risks: age and injury with Gordon, inconsistency for Romero. Though he contributed more than his share of nightmare moments in a rough 2007 (3-2, 6 saves, 4.73 ERA), Gordon has been so good for so long--you have to go back to 1996 to find his last ERA over 4--that the questions with him are all physical. He's 40 and pitching with a slight tear in his shoulder--not good for a guy whose signature offering is the hammer curve. The first news out of camp is good on Gordon, but he'll have to be handled with care. BP projects him for 2-2, 4.26, 6 SV, 40.2 IP, 1.37 WHIP.

Romero did more than anyone last year to burnish Pat Gillick's "Still Got It!" post-2001 Mariners reputation. Picked up off waivers from the Red Sox last June, the 31-year old lefty put up an absurd ERA+ of 373 for the Phillies thanks mostly to his uncanny power to direct batted baseballs into gloves (15 hits allowed in 36 innings). Unfortunately, that's pretty much the definition of "unsustainable." Romero was rewarded last winter with a three-year, $12 million contract, a deal that represented fair market value yet still looks like one the Phillies will regret. Every season in MLB, one or two relievers put up great seasons you couldn't have gotten odds on in Las Vegas; generally they revert back to pumpkins soon thereafter. That said, it's not like Romero's been a bum his whole previous career; with the Twins from 2002-2005, he put up three well above-average seasons in four. If that's the guy the Phillies picked up, he'll be fine. PECOTA forecasts a 2-2 record, 4.45 ERA, 2 saves, 45 IP, 1.59 WHIP.

The fourth guaranteed reliever is righty Ryan Madson. The 27 year-old gets little love despite an outstanding 2007 season cut short by injury: Madson pitched to a 3.05 ERA and 151 ERA+,  holding opponents scoreless in his final 8 appearances (12.1 IP). A meme has emerged that Madson isn't cut out for late-inning work, but he came in for the 8th or 9th inning in seven of those last eight scoreless appearances. The fastball/changeup specialist hasn't convinced PECOTA, which foresees a 3-3, 4.33 campaign in 56 innings for 2008.

Off-season pickup Chad Durbin is theoretically competing for the 5th-starter slot, but it's much more likely he ends up in the bullpen sharing middle-relief duties with Madson. After a mostly terrible first few years in the big leagues, Durbin did adequate work for the 2007 Tigers, putting up a 4.18 ERA in 28 innings as a reliever. (As a starter, he pitched to a 4.88 ERA in 99.2 IP). BP sees a 3-4, 5.31 season in 59.2 IP (including 6 starts--presumably not very good ones) for his 2008.

The two remaining bullpen spots are up for grabs as we sit here six weeks before the first game of the season (!). My guess is that the last two slots go to a hard-throwing strikeout guy and a second lefty, respectively--and if nobody overwhelms, Clay Condrey, who fits neither description, comes back to rack up more service time. Candidates for the former include Francisco Rosario, Joe Bisenius, J.D. Durbin, and Rule 5 pickup Lincoln Holdzkom. Here are their PECOTA projections:

Rosario: 2-2, 4.43, 43 IP, 1.43 WHIP
Bisenius: 3-4, 5.74, 56 IP, 1.79 WHIP
J.D. Durbin: 3-4, 5.26, 65.2 IP (9 starts), 1.59 WHIP
Holdzkom: (no projection)

The lefties include ex-Pirate Shane Youman, Rule 5er Travis Blackley, and Fabio Castro; Mike Zagurski, the rotund southpaw who pitched with the team last season, will begin 2008 rehabbing from the hamstring tear he suffered late last year, but we'll throw him into the mix too.

Youman: (no projection)
Blackley: 5-7, 5.88, 101 IP (15 starts), 1.61 WHIP
Castro: 3-4, 5.17, 61 IP, 1.63 WHIP
Zagurski: 2-2, 4.64, 36.2 IP, 1.51 WHIP

And finally, Condrey: 2-2, 4.76, 44.1 IP, 1.51 WHIP

Adding this all up, it looks like a middle-of-the-road bullpen. Lidge should be a solid closer; strikeout rate is key, and he's got one of the best. Between Gordon, Romero and Madson, an expectation of 1.5 good seasons (someone will miss half the year or so) seems reasonable. The numbers for the guys who will fill bullpen slots 5 through 7 look ugly, but this is where the variability is the greatest: any of Rosario, Castro, Bisenius, Zagurski or the also-rehabbing Scott Mathieson could break out and suddenly find the command they need to be solid relief assets. The key, as always, will be to minimize the damage allowed through the first three months while Manuel and the coaching staff figure out who is and who isn't reliable. Early reliance upon the lower-ceiling but more predictable guys like Chad Durbin and Condrey could be a smart way to proceed.