The New Way To Spell Relief

SI's comprehensive rating system proves there's more to judging relief pitchers than just counting their wins and saves
April 12, 1982

After the 1981 season ended, SPORTS ILLUSTRATED asked the Elias Sports Bureau of New York City to research a statistical system that would fairly—and definitively—evaluate relief pitchers. It was our opinion that relievers have never been analyzed properly and that the systems used to rate relievers have substantial flaws. For instance, the Rolaids Relief Man and The Sporting News awards are determined only by a reliever's won-lost record and saves. The antacid promotion—baseball's official standard of reliever supremacy—uses a system that awards two points for each relief win or save and subtracts one point for each relief loss. The Sporting News honor goes to the pitcher with the highest total of relief wins and saves. Both awards use earned run average as a tiebreaker. (Baseball's Basic Agreement offers a third method, but it judges relievers over a two-year period, includes their starting appearances and puts special emphasis on number of appearances and innings pitched.)

The Rolaids and The Sporting News awards are necessarily limited to relievers who get the most opportunities for wins and saves, namely the short relievers. In 1981 Milwaukee's Rollie Fingers and St. Louis' Bruce Sutter—two of baseball's archetypal short men—received the American and National League awards from both Rolaids and The Sporting News; Fingers also won the American League's Cy Young and MVP awards. Nine of the top 10 relievers worked exclusively or primarily in short relief.

The rating system created by SI goes far beyond anything hitherto concocted. Our ratings incorporate eight categories and evaluate long, middle and spot re' lievers as well as short men. And now, with no apologies to Fingers and Sutter, our winners are...Goose Gossage of the Yankees and lefthander Joe Sambito of the Houston Astros, who were Nos. 2 and 7, respectively, in the other ratings. In fact, six of our top 10 finishers (see charts on pages 83 and 86) didn't even appear among the top 10 on the other two lists: long relievers Sammy Stewart of Baltimore and Dave LaRoche of the Yankees, middle reliever Ron Davis of the Yankees, short-middle reliever Jeff Reardon of the Mets and Expos, and short men Tug McGraw of the Phillies and Kevin Hickey of the White Sox. Their subtle contributions went unnoticed in a mere accounting of wins and saves. At the same time, four pitchers who finished high in the awards competitions—Mark Clear of the Red Sox, Steve Howe of the Dodgers, Tippy Martinez of the Orioles and Steve Comer of the Rangers—are barely to be found in our rankings.

The flaw in the Rolaids and The Sporting News systems is their criteria. Wins? A reliever is supposed to hold a game in check, not win it. In fact, relievers often get wins after bad performances. Inheriting a 4-3 lead over the Mets last June 9, the Reds' Tom Hume yielded the tying run on three singles and a walk. But because the Reds rallied to score four runs in the ninth, Hume got the win. Fact is, many premier relievers have relief records that border on .500; Fingers (101-90), Sutter (35-35), Sambito (31-29), Mike Marshall (92-98). Saves? They're awarded only to pitchers who finish games. A successful reliever need not be the Gossage, Fingers or Sutter who gets the last out. He may pitch a couple of innings of useful middle relief, as the Yankees' Davis does better than anyone. He may throw three or more innings of long relief to keep a game close, the specialty of Baltimore's Stewart, who was second in the American League last year with a 2.33 ERA. Or he may be the classy old lefty—Grant Jackson, now with Kansas City, comes to mind—who strolls in at a critical juncture to retire just one or two men. To the awards people, these guys are real nowhere men. (If Rolaids "spells R-E-L-I-E-F," why can't Elmer's Glue-All "spell H-O-L-D?" Elmer could name his award after Horatius, the fellow who held the bridge.)

Back to reality. ERA serves as a good method for judging a pitcher's effectiveness against the batters he faces, but it tells you nothing about his ability to keep the previous pitcher's runners from scoring. Early last season the Royals led the Orioles 2-0 when KC starter Dennis Leonard loaded the bases to start the eighth. Dan Quisenberry relieved and gave up a two-run double and a sacrifice fly, leading to a 3-2 Oriole victory. Because all three runs were charged to Leonard, Quisenberry's disastrous outing wasn't reflected in his stats for the game: 2 IP, 0 ER.

Clearly, the times cry out for new criteria, even if California's Manager Gene Mauch does rail, "We've got too damn many stats in baseball now." Mauch feels that any new stat will inevitably become "just another negotiating tool for some agent," but the fact is that many teams and some individual players have always kept private files. So a public accounting is now in order.

Using the services of Elias, the official statistician of the National League, the NBA and the NFL, and box scores from the 1981 split season, SI has come to the rescue and developed a unique method of measuring relief. Four of our eight statistics—holds, first-batter effectiveness, run prevention and percentage of a pitcher's appearances in which he gets a win, save or hold—have never before been tabulated for all major league teams.

The hold category is an ideal measure for a pitcher like Davis. "Any reliever who helps his team but doesn't win or save should get a hold," he says. "He should get one if he comes in and retires the only batter he faces. He should get one if he preserves a lead or prevents a losing game from getting more one-sided. More than one pitcher on a team can get a hold and you can get them whether or not your team wins." SI's hold category took all of these circumstances into consideration with the proviso that the reliever completed at least one inning.

Davis' leadership in this category isn't surprising. Pitching almost as effectively in the seventh and eighth innings as Gossage did in the ninth, Davis has helped create a new concept in baseball: the six-inning ball game. Opponents had better be leading the Yankees after six; if not, Davis-Gossage will surely shut them down the rest of the way. An instructive outing was June 6 in New York. With the Yankees leading Chicago 2-0 after six, Davis relieved starter Doug Bird. Davis pitched two innings of one-hit ball, Gossage cleaned up in the ninth and the Yankees won 2-0. Strictly routine. Bird got the win, Gossage got the save—and Davis got nothing. We say he deserved a hold. Having created the six-inning ball game, Davis and Gossage now give the Yankees the luxury of harboring six-inning starters.

Middle relievers are notoriously unappreciated. "Middle relief is just treated as nothing," says Sparky Lyle of Philadelphia, who won the Cy Young Award as a Yankee short man in 1977. "You have no chance to make any money." Middle relievers get few saves, and the rare years in which they win big often come when their teams have successful short men. In 1977 Pittsburgh's Kent Tekulve had a 10-1 record, but his teammate Gossage had 26 saves. In 1980 Davis was 9-3 but Gossage was again the bullpen favorite: 33 saves.

The hold also benefits long relievers, who may be the most frustrated pitchers of all. They occupy the lowest rung on the relief ladder and get virtually no credit for one of the grimmest jobs in baseball. "A long reliever just has to work his way up to short reliever," says Bob Sykes, who was traded from the Cardinals to the Yankees over the winter and was sent down to Columbus last week. Long men need at least three pitches; they often warm up without entering the game; and they must be durable enough to last six or seven innings.

Players back the hold concept overwhelmingly. Speaking eloquently for all relievers, Minnesota short man Doug Corbett says, "The hold brings grandeur to the long or middle reliever who otherwise ends up in oblivion."

So does first-batter effectiveness. A few relievers don't like this category, claiming they're often asked to pitch around the first batter. However, such managers as Mauch and the Cardinals' Whitey Herzog disagree. "Getting the first guy out is the only thing that matters," says Mauch. "So many times the game's right there," says Herzog.

Especially when men are in scoring position, which brings us to run prevention. This category awards relievers points for men who are left in scoring position and subtracts points for runners who are allowed to score. With two out in the ninth, runners on first and third and a one-run lead over the Cubs last May 4, Sambito had just one job upon entering the game: Retire Leon Durham. Sambito struck him out. "A reliever's job is to keep runs from scoring," says the Yankees' recently acquired Shane Rawley. Hume agrees, saying, "It bothers me more when I give up a run that's charged to the pitcher I replaced than to give up one that's charged to myself." Managers and pitching coaches point out that some situations (man on first, two out) are less challenging than others (men on second and third, none out). Thus, SI restricts the run-prevention category to men in scoring position; it judges only effectiveness under stress.

Among our other categories, players consider ERA a fair standard for long relievers who start many of their own innings, saves a good measure for short relievers and win/save/hold percentage a means of equalizing pitchers on winning and losing teams. The best equalizers of all may be our final two categories—runners on base per nine innings and innings pitched in relief. The runners-on-base category applies equally to all relief pitchers. Relief innings, surprisingly, are as fair a gauge for short relievers as for long. A long reliever may pitch more innings in a single outing, but he can't pitch every day and may be diverted by spot starts. A good—and healthy—short man will rack up innings galore.

To be eligible for our charts, a pitcher had to make at least 25 relief appearances in 1981. He got 10 points for each first-place finish in a category down to one point for each 10th.

Some of the findings are anything but surprising. As the most celebrated and salaried of their kind, short relievers dominate the overall standings, and Gossage and Fingers are the best. You expected Bobby Sprowl?

The most impressive new stat on Gossage is his win/save/hold percentage: He got one of the three in 87.5% of the games in which he appeared. Why wasn't Gossage—baseball's best relief pitcher—the American League's MVP and Cy Young winner instead of Fingers? Because redoubtable Rollie appeared in 47 games to 32 for Gossage; Goose, remember, missed 40 days of the already shortened season with various back, shoulder and groin ailments.

As expected, Davis dominated the hold category and finished fifth overall in the league. But how did a third Yankee, Dave LaRoche, who starts this season as a player-coach in Columbus, inch into the top 10? In a typical outing LaRoche—and his LaLob delivery—pitched five scoreless innings against the Royals on May 20 and left with the score 4-4 in the ninth. He was forgotten by the time the Yanks won 5-4 in 11. "Anybody who follows a team knows who's doing his job," says LaRoche, who figures to be recalled later in the season.

Even devoted Astro fans may be surprised at Sambito's first-place finish in the National League. Sambito was understandably overlooked on a staff featuring the highest-salaried starting rotation in baseball and two other good relievers, Dave Smith and Frank LaCorte. But Sambito's grip is solid. He was the only pitcher in either league to place among the top 10 in seven of the eight categories. No National Leaguer was more effective at retiring the first batter with men in scoring position. After Aug. 27 Sambito relieved with 12 men on base, and none scored. No wonder Houston won the NL West's second-half title.

Jeff Reardon's second-place finish is equally notable. When the Mets traded him to the Expos for Ellis Valentine on May 29, Reardon not only had to change teams but also styles. A sometimes-middle reliever with New York, he switched to short relief with Montreal. As the Expos held off St. Louis to win the NL East's second half, Reardon allowed just one run in his last nine appearances.

The third-place finisher is a name that few people outside San Diego know at all: the Padres' Gary Lucas, a top 10 finisher in six different categories. Two other high-ranking pitchers on low-ranking teams are Atlanta's Rick Camp, whose ERA has been under 2.00 the last two seasons, and San Francisco's Greg Minton, who hasn't allowed a home run in his last 255 innings.

Sutter's fifth-place finish may seem surprising, but not to those who watched him pitch from the stretch. On nine of 11 occasions when Sutter relieved with fewer than two outs and men in scoring position, the opposition scored at least one run. Sutter was much better with two outs, but his overall score of 10 is unimpressive, to say the least. Other notable absentees from run-prevention leadership are Gossage, Fingers, Quisenberry, Tekulve and Basic Agreement leader Corbett. Hume, the National League leader in this category, got out of scoring threats 80% of the time, both with two outs and fewer. (The major league average was approximately 75% with two outs and 50% with fewer than two.) Sambito survived all eight two-out appearances and six of seven low-out showdowns. American League leader Tom Burgmeier of Boston escaped 15 of 16 two-out jams. Talk about pitching under pressure: Detroit's Kevin Saucier, who came in with men on base 29 of 38 times, was the best at retiring the first batter in the midst of scoring threats.

Run prevention is an important category because it is the statistic of ultimate accountability, giving little-used but oh-so-effective specialists their due. Just check out the American League leaders: two Red Sox other than Clear, two players from the purportedly weak A's bullpen, erstwhile softballer Hickey and the almost unknown Don Cooper of Minnesota.

Because the pitchers in the National League can be lifted for pinch hitters, they are more likely than American Leaguers to start innings. Sutter led the NL in saves because there were no men on base in 31 of his 48 appearances. As for the American Leaguers, Gossage and Fingers were somewhat pampered; Gossage reported with men in scoring position eight times, Fingers 16. The statistic that really separates them is first-batter effectiveness, in which Gossage was among the leaders and Fingers wasn't. But when it comes to cranking it up for a single batter, both could have taken lessons from golden oldies McGraw (37) and St. Louis' Jim Kaat (43).

Eleven of the 20 hold leaders played for teams with losing records. Of the holdsters on winning teams, the Yankees' Davis and LaRoche and the Brewers' Jamie Easterly used the category to escape from the shadows as stoppers.

Surprisingly, not only long and middle relievers picked up holds; short men from weak teams or from bullpens where the work load was shared also got holds. Note the totals for Tekulve (19), Sambito (18) and Corbett (11). The long man's long man, Baltimore's Stewart, wasn't among the hold leaders; nevertheless, he led all relievers in innings pitched, and all long and middle men in ERA, and was seventh among American League pitchers in first-batter effectiveness.

A final conclusion: The best long men inevitably play for strong clubs. If they pitched for teams with weak rotations, they'd be converted to starters.

Unfortunately, statistics in only two of our eight categories—ERA and saves—will be regularly published during the 1982 season. We'll have to evaluate daily performances from the box scores. "Watch the hits-to-innings and strikeouts-to-walks ratios," says statistician Steve Hirdt of the Elias Bureau. "They're good indices."

Meanwhile, several managers suggested to SPORTS ILLUSTRATED that we should tabulate "chokes." Other observers mentioned "scares," to be defined as "when the sight of a reliever warming up scares the opposition into stopping a rally." Or how about the reverse situation, says Quisenberry. "Call it a Carole King. It's when you're loose and ready to enter a game, but you look up and the starter has just given up a two-run homer to lose, and 'It's too late, Baby, it's too late.' "

Count on Quisenberry, the Royal Mad Hatter, to see the big picture: "How about setting up something like in diving? Have a degree of difficulty for save situations—only no judges from Eastern Bloc countries. Maybe you could have half saves, like half sacks in the NFL. But I think the real need is to come up with something different at the end of games. Coming off the mound with a fist or arms raised is old hat. After one of my games last year, I walked straight to the dugout and didn't shake hands with anyone."


1. Joe Sambito, Hou


2 (1.83)

7 (10)

2 (8.72)

5 (18)

9 (.673)

6 (.208)

2 (+18)

2. Jeff Reardon, NY-Mont


5 (2.19)

10 (70)

1 (8.61)

1 (20)

3 (.721)

6 (+8)

3. Gary Lucas, SD


4 (2.00)

5 (13)

1 (90)

10 (10.20)

2 (19)

7 (.684)

4. Rick Camp, Atl


1 (1.78)

4 (17)

5 (76)

6 (9.24)

4 (.708)

5. Bruce Sutter, St. L


1 (25)

3 (82)

3 (8.89)

4 (.708)

6. Dave Smith, Hou


6 (75)

4 (9.00)

5 (18)

2 (.738)

4 (.190)

7. Tug McGraw, Phil


7 (10)

7 (9.41)

8 (.676)

2 (.156)

8 (+6)

8. Tom Hume, Cin


5 (13)

6 (.686)

1 (+19)

9. Neil Allen, NY


3 (18)

1 (.744)

9. Greg Minton, SF


2 (21)

2 (84)

11. Al Holland, SF


4 (79)

5 (+11)

11. John Littlefield, SD


8 (17)

1 (.143)

13. Sparky Lyle, Phil


6 (75)

2 (19)

14. Fred Breining, SF


9 (73)

5 (18)

8 (.227)

14. Kent Tekulve, Pitt


9 (2.49)

2 (19)

16. Woodie Fryman, Mont


3 (1.88)

16. Gene Garber, Atl


5 (9.15)

9 (.229)

16. Grant Jackson, Pitt-Mont


3 (+14)

16. Gary Lavelle, SF


3 (.161)

16. Lee Smith, Chi


10 (.231)

4 (+13)

21. Pete Falcone, NY


22. Jim Kaat, St. L


23. Dick Tidrow, Chi.


24. Dave Stewart, LA


25. Rawly Eastwick, Chi


25. Dave LaCorte, Hou


27. Ron Reed, Phil


27. John Urrea, SD


29. Joe Price, Cin


30. Bobby Castillo, LA


31. Dan Boone, SD


31. Doug Capilla, Chi


31. Enrique Romo, Pitt


34. Steve Howe, LA


35. Doug Bair, Cin-St. L


35. Larry Bradford, Atl


35. Mark Littell, St. L


35. Paul Moskau, Cin


35. Mike Proly, Phil


35. Ray Searage, NY


35. Elias Sosa, Mont


1—Awarded to a reliever who, in a minimum one-inning appearance, prevents a lead from decreasing, a deficit from increasing or a tie from becoming a deficit.
2—The percentage of a reliever's overall appearances in which he earns a win, save or hold.
3—If the first batter is walked intentionally, the second batter is the criterion.
4—A reliever who enters a game with fewer than two outs earns +2 for each runner in scoring position who doesn't score and -2 for each who scores. If there are two outs, the point totals are +1 and -3.