Print and Go Back ESPN.com: Gammons [Print without images]

Sunday, September 14, 2003
Bullpens always unpredictable

By Peter Gammons
Special to ESPN.com

Sept. 14

"We Don't Know (----)"
-- Sign in Theo Epstein's Fenway Park office.

The logic that applies to building a better bullpen is the same logic that applied to Kevin Costner and James Earl Jones finding Moonlight Graham, or to the first 130 pages of Camus' "L'Etranger," or, if you must, Lowell George's "Apolitical Blues."

Eric Gagne
Eric Gagne is the latest award-winner pitcher to wear Dodger blue.

It is the logic that, when the Dodgers were trying to find a closer, they traded for Paul Quantrill for that purpose, and in the deal told the Blue Jays they could have either Luke Prokopec or Eric Gagne, although they preferred to deal Gagne. That was about the time they traded Matt Herges to Montreal for Guillermo Mota, the one-time shortstop in the Mets organization who went to Montreal in a minor-league draft.

And now Quantrill sets up for Mota, who sets up for Gagne, in what may be the best bullpen of this generation. "I can't say," Dan Evans admits, "this all came together in some master plan."

It's the same logic that applies to Tim Worrell, who had seven saves in his 10 previous major-league seasons, but is among the National League save leaders in the absence of Rob Nen. Or to Joe Borowski. In 2000, having been released by the Reds in spring training, Borowski was pitching for the independent Newark Bears when Cubs GM Jim Hendry came in to see Bobby Hill, only to have Hendry reject signing Borowski, which sent him to Mexico before he finally was signed. And now he's the closer of choice for Dusty Baker and Hendry.

It is the same logic that caused Ken Williams to trade Keith Foulke for Billy Koch, only to have Foulke be the AL's best closer while Koch says he picks out Gatorade for his fellow White Sox bullpen mates -- while Tom Gordon is coming back as one of the game's best, most dominant closers.

Epstein has spent 10 months trying to assemble a bullpen. In that time, he has acquired or signed Chad Fox, Mike Timlin, Byung-Hyun Kim, Ramiro Mendoza, Scott Sauerbeck and Scott Williamson ... and still, every time a Red Sox starter leaves a game, New Englanders are convinced the end is near. For logic's sake, go back one year and ask whom you'd like Epstein to get on his budget. Then go to your Baseball Prospectus page and click on the relievers. No, there isn't anyone outside of the MIT or Cal Tech library that understands what all those numbers mean, but we do know they are an accurate reading of a reliever's season. When the 2002 season ended, Kim was the second most effective reliever in the majors, Alan Embree was 18th, Sauerbeck 19th, Williamson 23rd ... and Mendoza's adjusted runs-prevented number was 5.1, as opposed to Gordon's --1.9 (in case you haven't noticed, Mendoza has given up more hits than Pedro Martinez). And Fox was in the top 20 in 2001, a season in which Embree was among the 10 worst relievers in baseball, as was Felix Heredia, who this season is in the top 30 for the first time.

Middle men
HIGHEST PAID MIDDLE RELIEVERS
(2003 figures)
Player salary
S. Hitchcock, Cards $6 million
Steve Karsay, Yanks $5 million
Steve Sparks, A's $4.5 million
Antonio Alfonseca, Cubs $4 million
Jeff Nelson, Yanks $3.98 million
Paul Shuey, Dodgers $3.91 million
John Franco, Mets $3.83 million
David Weathers, Mets $3.73 million
Curt Leskanic, Royals $3.7 million
Arthur Rhodes, M's $3.5 million
Brian Meadows, Pirates $3.5 million
Turk Wendell, Phillies $3.35 million
Paul Quantrill, Dodgers $3.33 million
Gabe White, Yanks $3.27 million
Jim Mecir, A's $3.26 million
Felix Rodriguez, Giants $3.25 million
Jay Powell, Rangers $3.22 million
J. Zimmerman, Rangers $3.2 million
Matt Anderson, Tigers $3.2 million
Buddy Groom, O's $3 million

Go through these two lists on this page, and realize that Steve Karsay, Mendoza and Mike Stanton are making more than $12 million between them next season, while Brendan Donnelly, Francisco Rodriguez, Ben Weber and Scot Shields -- who between them make $1,317,000, nearly $2 million less than Matt Anderson -- set up Troy Percival in what is, for the second year in a row, the best bullpen in the American League.

But while the Angels are in the top five in the game (by Baseball Prospectus' numbers) in successive years, they are only the fourth team to have repeated as a top-five bullpen, and they are one of nine teams to have made the top five more than once in the last six seasons. Conversely, go back to 2002: the Braves, Angels, Cardinals, Twins and Giants were the top five. This season, the Braves are 18th, the Cardinals have fallen all the way to the third worst. Follow it out to a top eight from last season, and the Pirates and Orioles, 7-8, have sunk to the bottom eight, while the Yankees' aggregate numbers (the Dodgers are No. 1 at 89.6, the A's fifth at 36.1) have fallen from 40.7 in 2001 to 31.6 in 2002 to -4.5 this season.

As we have seen with every world champion since 1996 -- when Mariano Rivera was setting up John Wetteland -- through the Angels, a strong bullpen is an integral component in winning, and a major reason that so many are skeptical about the chances of the Red Sox, Braves, Cardinals or Yankees in going all the way. Going back six years, only two years -- including the 2003 Angels -- have had top-five bullpens and had a losing record. From 1998 through 2002, the aggregate winning percentage of top-five bullpen teams was .572, while the winning percentage of the bottom five teams was .428. All five teams with top-five bullpens in 2001 and 2002 made the playoffs; this year's top five are Los Angeles, Anaheim, Houston, Seattle and Oakland (Red Sox fans take note). However, if you're looking at the Seattle model, realize that Kazuhiro Sasaki, Jeff Nelson and Arthur Rhodes have had off years or been traded, while it's been Shigetoshi Hasegawa, Rafael Soriano and Julio Mateo who have been the mainstays.

Baseball Prospectus Top 30
(As of Sept. 12)
Player rating salary
S. Hasegawa 28.4 $1.8M
B. Wagner 27.3 $8M
Eric Gagne 27.0 $550K
R. Cormier 26.9 $2.98M
B. Donnelly 26.7 $325K
G. Mota 25.3 $675K
D. Riske 22.7 $314K
J. Smoltz 22.4 $10.67M
K. Foulke 22.3 $6M
S. Shields 20.3 $305K
D. Marte 19.5 $330K
P. Quantrill 18.4 $3.3M
J. Valverde 17.0 $300K
L. Hawkins 16.8 $3M
R. Soriano 16.8 $320K
Fr. Rodriguez 16.4 $312K
B. Fuentes 15.9 $300K
J. Mateo 15.4 $300K
C. Bradford 15.0 $331K
T. Gordon 15.0 $1.3M
O. Dotel 14.8 $1.6M
B. Weber 14.6 $375K
C. Reitsma 14.4 $350K
J. Speier 14.2 $850K
F. Heredia 13.9 $600K
B. Lidge 13.1 $300K
M. Rivera 12.9 $10.5M
M. Mantei 12.6 $7.07M
K. Farnsworth 11.9 $600K
L. Ayala 11.8 $300K

In fact, the 35-year-old Hasegawa is the No. 1 reliever, statistically, in the majors. He'd never cracked the top 30 in seven previous seasons in the major leagues.

So while you recall that Donnelly's road to Anaheim included a stint as pitching coach in Nashua, N.H., and an independent league game in which he knocked down an opposing woman batter, or that Ben Weber's was through three independent leagues and Taiwan, consider some of these facts off the Baseball Prospectus ratings:

  • Only six of the top 30 in 2001 are in the top 30 in 2003;

  • Only 10 of last year's top 30 are in this year's top 30;

  • Only three pitchers -- Foulke, Billy Wagner, Octavio Dotel -- have made the top 30 three consecutive seasons, and only Foulke has made it four straight years;

  • Kyle Farnsworth made the top 30 in 2001, the bottom 10 in 2002 and the top 30 in 2003;

  • Only five of the 2000 top 30 made the 2001 list, but Koch, Nelson, Armando Benitez, Curtis Leskanic, Jim Mecir, Gabe White and Bobby Howry parlayed those seasons into long-term contracts that because of injuries or decreased performance became white elephants.

    The are obvious problems with making long-term judgments on relievers off one or two seasons (yes, Mary, Jeff Tam was in the top 20 in 2000 and 2001). The first is burnout, when managers fall in love with the hot pitcher(s) and grind their arms into sawdust, thus leaving them unable of regaining arm speed, touch or command the following spring. The second is that the job requirement is one or two pitches, and if either the arm slot or touch goes the following season, so do the numbers (or, Mary, yes, Chris Hammond was No. 3 and Tony Fiore No. 8 in 2002).

    The third problem is that relievers offer such a small sample. Hammond only had to get 228 selected outs last season. Derek Lowe, in his breakout year as a starter, got 660 in all innings and in all situations at all slots in opponents' batting orders. We knew a lot more about Lowe than we did Hammond.

    Epstein was never convinced that the Bill Jamesian theory of six good arms doth a bullpen make, although repeatedly he decried the pitchers that he, Epstein, had assembled during the winter other than Timlin. Whether he be Foulke, Bob Wickman (who is re-claiming his Indians job at the risk of Danys Baez being non-tendered) or Worrell, the ninth inning does require someone who can take that inning's heat and accept its responsibilities. But the ideal model is to find what the Angels, Dodgers and 1996-2000 Yankees had -- power, strikeout pitchers who can walk into the most critical situations in the sixth, seventh and eighth innings and deliver the closer the lead with three, four or five outs remaining.

    Of course, it's great to set your plan to find a young Mariano Rivera or another Francisco Rodriguez or Guillermo Mota. It's another thing to find them, much less develop them properly with the right mix of minor-league innings before becoming a reliever.

    It may be more difficult to go to the freeagentrelievers.com and fill your shopping cart with established relievers. That's how Jay Powell, Todd Van Poppel, Hammond, Matt Anderson, Danny Patterson and Karsay got those three- or four-year deals.

    For instance, for all Donnelly has done the past two seasons for a two-year total of less than $600,000, ask yourself this: if you were Bill Stoneman, would you give Donnelly a three-year, $10 million contract? Emotionally, yes. Logically, no.

    So if you're Walt Jocketty or Theo Epstein or Brian Cashman and everyday you're looking at bullpen roulette, logic has nothing to do with it. You've got a chance to run into Moonlight Graham.