Boston's Bullpen Gamble Spurning high-priced closers, the RED SOX are touting a new strategy and testing the odds with a mix of lowcost relievers to get them out of jams

An object lesson in the use of the modern bullpen: Last Aug. 21
the Boston Red Sox, 3 1/2 games out in the American League
wild-card race and gasping for their playoff lives, were tied
3-3 with the Texas Rangers entering the top of the seventh at
Fenway Park. Red Sox manager Grady Little hooked starting
lefthander Casey Fossum and inserted righthander Bobby Howry, a
failed closer with the Chicago White Sox before being obtained
for two Class A pitchers three weeks earlier. Howry faced pinch
hitter Todd Hollandsworth, followed by the first five batters in
the lineup--six Rangers hitting a combined .289 and averaging 21
home runs apiece--and retired them all. After Boston took the lead
in the bottom of the eighth, Little, as push-button strategy
dictates, sent in closer Ugueth Urbina, who sailed past Carl
Everett, Herb Perry and Mike Lamb (a combined .260, averaging 11
homers each) in the ninth for his 29th save.

This is late-inning division of labor. "Closers come in, fresh
inning and bases empty," says Howry, "but the guy relieving the
starter in the seventh, with men on or facing the middle of the
order, that's the guy in the tough situation. Yet nobody knows
who the middle guys are. That's what drives pitchers to want to
close. Everybody knows who the closers are, and once you get that
title of closer, you triple your salary."

In bullpen society the pampered finishers such as Urbina--who
collected 28 of his 40 saves last season by starting the ninth
with a two-or three-run lead and earned $6.7 million doing
it--are the leisure class, amassing gaudy save totals and fat
contracts while proletarian setup men do the heavy lifting.

Though the formulaic use of Howry and Urbina worked against the
Rangers, Boston's pen was dreadful overall, going 15-22 with a
4.25 ERA and 17 blown saves last season; it was largely
responsible for the club's 13-23 record in one-run games.
Believing that personnel as well as philosophy were responsible,
Theo Epstein, after having been the Red Sox' assistant general
manager last season and as the new G.M. this year, has made
changes all around. He jettisoned the pricey Urbina, replacing
him with a corps of low-cost middlemen--lefthander Alan Embree
and righthanders Howry, Ramiro Mendoza and Mike Timlin--and
decreed that Boston would no longer waste a relief ace in
accumulating low-pressure, ninth-inning saves. Instead, top
relievers would work the most critical outs, with late-inning tie
games atop the priority list.

"The fact that we don't have a high-profile, proven, late-inning
relief pitcher is probably a financial decision," Epstein says.
"Using our best relievers to get the most critical outs, that's
purely philosophical. We could have had closer X, but we would
still be committed to using him in the seventh inning on a given
day if it was appropriate. Instead of religiously waiting till
the very end, we're picking our spots."

That is not a new idea. Not until the 1980s did managers begin
their slavish adherence to the conditions of the save statistic.
Before that, 100-plus-inning seasons and double-digit win totals,
as well as healthy save numbers, were typical of durable
relievers like Rollie Fingers and Mike Marshall. But what started
as an effort to save wear and tear on relief aces soon devolved
into the robotic deployment of the closer as a ninth-inning
specialist, and the comfort of handing the ball to a known
quantity at the end of each game became a powerful crutch. "When
you go to your closer and he doesn't get the save," says New York
Yankees manager Joe Torre, "there's no second-guessing, because
that's what you're supposed to do."

Little will be subject to lots of second-guessing. Add to his
plate the day-to-day tasks of playing matchups and massaging the
emotions of his apt-to-be-addled relievers, and the Red Sox have
chosen a psychologically taxing course. "Psychology," says Bill
James, "is never a reason to behave illogically."

James is the intellectual forefather of much of today's
statistically driven analysis. The Bill James Historical Baseball
Abstract, published in 1985, with an updated edition issued in
2001, a year before Boston hired him as a senior adviser, is an
iconoclastic reconsideration of baseball's conventional wisdom,
grounded in empirical methods. In a section of the latest edition
entitled "Valuing Relievers," James introduces a computer model
that assesses the impact of an ace reliever in several late-game
situations; at no time is he more valuable, James concludes, than
when pitching in a tie game in the eighth or ninth, when allowing
a run is most costly. "The traditional construction of bullpens
in recent years has a terrible blind spot with respect to a tie
game," he says. "It's focused on protecting the lead, when the
most important situation a bullpen faces is a tie game. To treat
that situation as an afterthought doesn't make sense."

In James's simulation, a relief ace's pitching one inning of a
tie game in the eighth or ninth improves his club's winning
percentage from .500 to .574; with a one-run lead in the ninth
the team improves from .810 to .867. Yet with a two-or three-run
cushion in the ninth--a situation that accounted for 32 of
Urbina's 61 appearances last year--the impact is minimal. "If you
use your relief ace to save a three-run lead in the ninth
inning," James writes, "you'll win that game 99% of the time. If
you don't use your ace in that situation, you'll win 98% of the
time. So is that a smart thing to do, to use a precious asset in
that situation?"

James only reinforced the conclusion Red Sox executives had
reached on their own last June, when Little was briefed on the
plan to forgo a standard save specialist in '03. The manager
responded enthusiastically when Boston soon after acquired Embree
from the San Diego Padres for two minor league pitchers and then
got Howry.

In Embree, 33, who re-upped with Boston for two years at $5.5
million in November, Epstein has the hard-throwing strikeout
pitcher he coveted. Armed with a mid-90s fastball and a tumbling
slider, Embree had a 4-to-1 strikeout-to-walk ratio and averaged
11.8 whiffs per nine innings last season. (Pedro Martinez's 10.8
led all starting pitchers.) Though he has just six career saves,
Embree has a closer's intensity, a demeanor he has sharpened
during off-season workouts over the past two winters with Randy
Myers, who was a closer for most of his 14-year career and three
times was a league leader in saves. "When I enter a game, I feel
it's on the line, no matter what inning it is," Embree says.
"Anywhere from the seventh on, I want to throw hard for one
inning and try to stop the game at that point. I would love to
close games here. I think I would be stupid not to say I want
that opportunity."

In the off-season Epstein, who believes the market generally
overvalues relief pitchers, spent chastely on Mendoza (two years,
$6.25 million), Timlin (one year, $1.85 million) and Chad Fox
(one year, $500,000). The highest expectations among that group
are for Mendoza, 30, a workhorse who often languished as a mop-up
pitcher with the Yankees. He's a straight sinkerballer who
pitched two or more innings in 22 of his 62 appearances last
season. Timlin, 37, who is with his sixth team in seven years,
has performed in every late-inning bullpen role. (The 32-year-old
Fox, who missed most of 2002 nursing a sprained elbow ligament
and rotator cuff pain in his right arm, has a great upside if
healthy.)

Little has lots of options. He just has to be wary of how each
pitcher handles being used in a variety of circumstances. "It's
difficult to run a bullpen like that," says Baltimore Orioles
manager Mike Hargrove, who had a closer-by-committee system with
the Cleveland Indians in 1993 and '94. "I found that people
perform better when their roles are defined."

Says Little, "In a perfect world, set is the way you'd like it,
but with the personnel we have, most of them have never been in
that position for an extended period." Little also believes the
team's lack of an endgame intimidator will be counterbalanced by
the unpredictability of his bullpen moves, allowing him to
dictate matchups instead of allowing the opposition to plan for
an expected setup-closer combination.

There is a sign on the clubhouse door at the Red Sox' minor
league facility in Fort Myers, Fla., that reads NO EGOS ALLOWED
ON FIELDS OR IN THE CLUBHOUSE, a maxim that Boston's bullpen gang
of four has adopted. The absence of a ninth-inning gunslinger,
entering a game to blaring heavy metal and leaving batters
quaking, bothers nobody. "Whatever you might lose intangibly, you
gain from a sense of unity and cohesion in the bullpen," Epstein
says. "By forgoing conventional roles and individual stats and
replacing them with the common goal of getting the last out and
getting a win--not a save or hold but a win--I think that builds
a certain esprit de corps in the bullpen."

Relievers of the Red Sox, unite! You have nothing to lose but
your saves.

Read Tom Verducci's Inside Baseball column every Tuesday at
si.com/baseball.

COLOR PHOTO: PHOTOGRAPHS BY AL TIELEMANS PEN PALS With a combined 30 years of relief work, (from left) Embree, Timlin, Mendoza and Howry will operate as a unit.

COLOR PHOTO: PHOTOGRAPHS BY AL TIELEMANS RAMIRO MENDOZA Best pitch: Sinker. Best situation: When double play is needed. Had a 2.04-to-1 ground-ball-to-fly-ball ratio last season, and induced eight double plays.

COLOR PHOTO: PHOTOGRAPHS BY AL TIELEMANS BOBBY HOWRY Best pitch: Low-90s fastball. Best situation: The start of an inning. Batters had an OBP of .364 against him with men on last season, .283 with the bases empty.

COLOR PHOTO: PHOTOGRAPHS BY AL TIELEMANS MIKE TIMLIN Best pitch: Mid-90s sinking fastball. Best situation: With righthanded hitters due up. They hit .236 against him over the last three seasons; lefties fared better, hitting .272.

COLOR PHOTO: PHOTOGRAPHS BY AL TIELEMANS ALAN EMBREE Best pitch: High-90s fastball. Best situation: Runners in scoring position, lefthanded hitters due up. Hard-to-handle fastball complemented by excellent control.

COLOR PHOTO: ORLIN WAGNER/AP BYUNG-HYUN KIM Has a rubber arm, but will have to get pitch count down from 16.1 per inning.

COLOR PHOTO: MATTHEW STOCKMAN/GETTY IMAGES DANNY GRAVES Favored sinker out of bullpen, but will mix in sliders and changeups as starter.

STARTING WITH CLOSERS
Boston isn't the only team experimenting with its staff. Arizona
and Cincinnati hope that last year's finishers can move into
their rotations

Rescued from what he considered bullpen purgatory, Arizona
Diamondbacks righthander Byung-Hyun Kim is a starter this
spring--and thriving. Despite a 2002 season that cemented the
24year-old Korean's reputation as one of the National League's
toughest closers--36 of 42 save chances converted and a 2.04
ERA--Kim has been moved into the rotation, as has long been his
wish. To those who witnessed Kim's lightning-strikes-twice
meltdowns in the 2001 World Series, which cost Arizona wins in
Games 4 and 5, that's an understandable impulse, but those
white-knuckle ninths weren't the only reason. "Part of it was a
cultural thing," says manager Bob Brenly. "If you were in the
bullpen in Korea, it's because you weren't good enough to be a
starter. Even though he was one of the best in the game, he
didn't like being a closer. He wants to start. Because of Matt
Mantei's [improved] health, now's the time."

Mantei, 29, is fully recovered from the Tommy John surgery that
cost him most of the last two seasons. This spring he's been
gunning his high-90s fastball with the same effectiveness that
enabled him to save 41 of 48 career chances for the Diamondbacks,
and no one's happier about his comeback than Kim.

Besides having the mental makeup of a starter, Kim possesses a
rubber arm, a necessity given his penchant for nibbling at
batters with an average of 16.1 pitches per inning. On game days
last season he'd play lengthy games of catch with batboys,
throwing a variety of pitches at near-full strength, then retreat
to Bank One Ballpark's indoor cages and continue throwing,
winging balls off the backstop by himself if he couldn't find a
partner. Still, durability is a concern. "He has to cut down on
his pitch counts," Brenly says. "As a starting pitcher you cannot
afford to have a 25-or 30-pitch inning early in the game and
expect to be in there long enough to get a decision. The role
change will help. As a closer a lot of times he came in with the
game on the line, and there were certain hitters he just refused
to give in to."

The early returns are positive: In his last two starts through
Monday, Kim had thrown eight shutout innings, striking out seven
and using only 95 pitches. In addition to his fastball and the
looping, submarine slider with which he mowed down righthanders
last season (they had a .198 batting average against him), Kim is
working on a changeup to throw at lefties. "If he has done as
well as he has the last few years in a role that he did not
like," Brenly says, "I'm anxious to see what he is going to do in
a role that he wants."

The Cincinnati Reds are conducting a similar experiment, but not
at their closer's request. Righthander Danny Graves, who averaged
30 saves over the last four seasons, is being tried in the third
slot of the rotation because the club believes that hard-throwing
Scott Williamson has more typical closer's stuff. Graves is
optimistic about the switch--he was 1-0 with a 1.89 ERA in four
starts late last year--but he struggled out of the gate. In three
spring starts he allowed 19 hits and nine runs over 9 1/3
innings, with an 8.68 ERA, and was not expected to make
Wednesday's start because of inflammation in his right knee.
Because a starter has a greater margin for error and can afford
to scatter hits, Graves is using his slider and changeup more,
rather than sticking solely with the sinker he predominantly
threw out of the pen. "All of his pitches will get better as he
uses them more," says manager Bob Boone. "It will make him more
efficient as a starter." --D.G.H.

SAVE COUNT
At week's end only three teams had pitching staffs in spring
training that accounted for fewer saves in 2002 than Boston's
current crop of hurlers. Here are the teams with the fewest and
most 2002 saves in their camps.

FEWEST SAVES

PITCHERS 2002
TEAM WITH SAVES SAVES

ROYALS 2 2
DEVIL RAYS 3 4
TIGERS 3 4
RED SOX 6 12
A's 4 15

MOST SAVES

PITCHERS 2002
TEAM WITH SAVES SAVES

BRAVES 3 82
YANKEES 5 81
RANGERS 6 72
DODGERS 5 56
WHITE SOX 4 56

Boston will use its strongest arm in late-inning tie games.
Protecting three-run leads in the ninth will be a lower priority.