clock menu more-arrow no yes mobile

Filed under:

The Relief Pitcher Championship Belt

From Hoyt Wilhelm to Mariano Rivera and everyone in between, who was the best MLB reliever each year since the adoption of the save statistic?

AJ Dungo

Consider four relief pitchers, whom MLB Network just rated the four best in the majors right now. And yet, they were all ineffective, undesired, or both just a couple of seasons ago.

  • Liam Hendriks, now on the White Sox, was designated for assignment in June 2018, going unclaimed on waivers.
  • The Padres’ Drew Pomeranz was a failing starter midway through the 2019 season, with a 6.10 ERA and sub-replacement WAR, when he moved to the bullpen.
  • The Brewers’ Devin Williams missed the entire 2017 season due to Tommy John surgery, then allowed 26 runs (22 earned) in 34 innings at the High-A level the following year.
  • The Rays’ Nick Anderson was a 28-year-old relief prospect so anonymous, entering the 2019 season, that Baseball Prospectus didn’t even mention him in its annual guide, which included commentary on more than 2,000 players. That offseason, Anderson was traded for minor leaguer Brian Schales in a 40-man roster reshuffle; the top comments on the Twins’ subreddit thread about the deal were “I’m going to be honest I have no idea who that is” and “Just a triple A guy, not really that noteworthy.”

The point is: Most great relief pitchers are comets in a world without telescopes, burning bright and brief and flashing atop leaderboards without warning. Ace relief pitching is tricky and unpredictable, except for a select few legends—and it has never been more important, as teams devote ever-increasing innings totals to their bullpens, especially in the postseason.

But which reliever is the best and most reliable? And who, out of the thousands of relievers who have populated major league rosters, grabs that title from past seasons? There’s no better way to answer these questions than with a good old championship belt exercise, starting in 1969, the first official year for Jerome Holtzman’s save statistic. (Apologies to 1950 NL MVP Jim Konstanty.)

But first, some ground rules:

  • This exercise aims to answer the question: Which relief pitcher would the plurality of fans have picked as the best in the majors at the end of each season? Or, put another way: In a Space Jam scenario in which the humans led the aliens by one run entering the bottom of the ninth inning, who would humankind vote to take the mound?
  • Starting pitchers are ineligible. From 2014 Madison Bumgarner to 2001 Randy Johnson to 1999 Pedro Martínez, the sport’s best starting pitchers have proved they’re the best relief options in key situations, too. This much is obvious. But Bill Barnwell already awarded the starter championship belt at Grantland, so we’re sticking to pitchers who predominantly work in relief here.
  • Advanced stats must be considered gradually. Fans in 1977 didn’t know Bruce Sutter’s fWAR, so for the most part, we can’t just rely on our WARs and WPAs. Yes, that means raw save counts matter, even if they’re largely a function of opportunity rather than skill. But we’ll also figure our hypothetical voting panel is full of savvy, forward-thinking fans, so saves and ERA aren’t the only factors that matter—especially in more recent years, when advanced stats are more the norm.
  • Track record matters—to a point. In most cases, one dynamite season isn’t enough to rocket a reliever with a lesser track record past a still-strong stalwart; José Valverde went 49-for-49 in saves in 2011, but it’s hard to figure anybody would have trusted him more than Mariano Rivera in a must-win game. However, if there’s any position with which to be a tad more aggressive, it’s relievers, so this rule is flexible.
  • The playoffs count, too—for better or for worse. Typical baseball award voting doesn’t factor in postseason performance. But if the fate of humanity rested on the arm of one man, wouldn’t you feel a little concerned if he had just blown two games in the World Series, and maybe choose a backup option instead?

With those rules in mind, here is the history of the Relief Pitcher Championship Belt.

1969: Hoyt Wilhelm, California and Atlanta

14 saves, 2.19 ERA, 2.46 FIP

The belt begins in a fitting spot, with the first great closer and, eventually, the first reliever to make the Hall of Fame. Counting saves awarded retroactively, Wilhelm held the career saves record until 1980; he also epitomized the circuitous journey most take to relief pitching by not breaking into the majors until the age of 29.

If the belt were awarded before 1969, Wilhelm might have it for the entire decade: From 1961 through 1969, from the ages of 38 to 46, Wilhelm recorded an ERA of 1.99 over nearly 1,000 innings. And in 1969, despite, again, being 46 years old, Wilhelm was still a valued reliever in his last great season. In early September, Atlanta, two games back in the playoff race, traded for the knuckleballer; he proceeded to allow one run in 12 1/3 innings down the stretch as the team jumped to a division title.

1970-1971: Ken Sanders, Milwaukee

13 saves, 1.75 ERA, 2.34 FIP; 31 saves, 1.91 ERA, 2.97 FIP

The Athletics didn’t have much use for Sanders, who toiled at Triple-A for all but 10 2/3 innings from 1967 to 1969. But after an offseason trade to the Brewers, entering their first season in Milwaukee, Sanders forced his way to the majors by the end of May and reeled off two marvelous seasons in relief.

Sanders also typified the fireman mentality of this early era of relief pitching, entering games at any point from the first through 12th innings and pitching any amount up to 7 2/3 innings in a single game (!) out of the bullpen. Counting his minor league innings, he threw a combined 262 2/3 innings across these two seasons. This was how relievers operated in the 1970s (and most of the 1980s): Every remaining selection in the decade also exceeded 100 innings in their belt-holding seasons.

1972: Rollie Fingers, Oakland

21 saves, 2.51 ERA, 2.19 FIP

Fingers was not the best reliever in the 1972 regular season—that honor probably goes to Tug McGraw, Sparky Lyle, or Mike Marshall—but an incredible playoff run gives him the belt. Fingers pitched in six of seven World Series games, including a two-inning save in Game 7, and his 1972 postseason is the most valuable for any reliever in MLB history, per championship win probability added (cWPA), which calculates that Fingers alone was worth 62 percent of a title that October.

Also: Imagine if Dan Marino’s famous fake spike play had come in the Super Bowl. That’s essentially what Fingers pulled in Game 3 of the World Series, when he struck out Johnny Bench with a fake intentional walk.

1973: John Hiller, Detroit

38 saves, 1.44 ERA, 2.25 FIP

Hiller suffered a heart attack in January 1971. He lost nearly 70 pounds and didn’t pitch all season; in 1972, he worked as a minor league coach until July, when he returned to the mound without any rehab appearances.

A year later, over 125 1/3 innings, he produced one of the greatest stat lines in MLB history, setting a single-season saves record that wouldn’t be broken for another decade. The only pitcher in the live-ball era to throw as many innings with such a low ERA is Bob Gibson in 1968. Ace relievers can come from anywhere.

1974: Mike Marshall, Dodgers

21 saves, 9 holds, 15-12 record, 2.42 ERA, 2.59 FIP

Marshall didn’t start a single game in 1974. He still threw 208 1/3 innings. For context, Stephen Strasburg led the National League in innings in 2019—with 209, less than a single frame ahead of Marshall’s relieving total 45 years before.

Marshall was traded to the Dodgers by Montreal before the season, and threw in a record 106 games in 1974; according to a Sports Illustrated feature later in the decade, “He considers himself most effective when pitching four or five times a week.” And boy was Marshall effective, with a devastating screwball and rubber band arm. In the World Series, although the A’s and Fingers were victorious, Marshall allowed just one run over nine innings, pitching all five games.

To cap his all-time season, Marshall became the first reliever to win a Cy Young—thus paving the path for more frequent reliever awards in the next two decades.

1975-1976: Rollie Fingers, Oakland

24 saves, 2.98 ERA, 3.03 FIP; 20 saves, 2.47 ERA, 2.29 FIP

Here’s a dilemma: Goose Gossage’s 1975 season, like Hiller’s and Marshall’s before, represents the pinnacle of the fireman performance that defined this era of relief pitching. Baseball-Reference WAR counts it as the best season for a reliever in MLB history (8.2 wins above replacement).

But in that season’s award voting, Gossage (tied for sixth in Cy Young, 17th in MVP) finished well behind Fingers (third and fourth, respectively), suggesting that the latter still held greater popular sway. Perhaps that’s because, through 1975, Gossage had never appeared in a playoff game, while Fingers had a 1.97 ERA in 50 1/3 playoff innings from 1972 to 1975, including three championships for the dynastic A’s. Fingers was so consistently clutch over that stretch that he’s now third all-time in career postseason cWPA, behind only Mariano Rivera and Madison Bumgarner.

From a 2021 vantage point, Gossage would win the belt this season—but when we’re trying to identify the best reliever at the time, it’s hard to discount that awards comparison.

In 1976, Gossage returned to the starting rotation for a year while Fingers—who was essentially the same effective pitcher, with nearly the same annual statistics, for the half-decade from 1972 through 1976—stayed in the pen, giving the impeccably mustachioed Athletic an easy win in his final season before embarking to San Diego in free agency.

1977: Sparky Lyle, Yankees

26 saves, 13-5 record, 2.17 ERA, 3.18 FIP

By FanGraphs’ version of WAR, Bruce Sutter’s 1977 season—with a 1.34 ERA and 1.61 FIP in 107 1/3 innings—is the best for a reliever in MLB history. But Lyle, a two-time saves champion and the originator of closer entrance music, was more celebrated after winning the AL Cy Young in a fractured vote that saw him edge past Jim Palmer and Nolan Ryan for the honor. Then, in the best-of-five ALCS, Lyle threw 5 1/3 scoreless innings in Game 4 with the Yankees facing elimination, before securing the final four outs in Game 5 the next day. He recorded the win in both games—and then nabbed another win in Game 1 of the World Series just two days later, with 3 2/3 more scoreless innings, as the Yankees moved toward a title.

1978: Goose Gossage, Yankees

27 saves, 2.01 ERA, 3.00 FIP

Gossage signed with the Yankees in free agency, took Lyle’s closer job, and stole his belt, too. A fading Lyle—who had gone from “Cy Young to sayonara,” in the words of third baseman Graig Nettles, and wrote a spicy memoir about the season—didn’t pitch at all in the World Series, while Gossage tossed six scoreless frames en route to another New York championship. A trade sent Lyle to Texas after the season.

1979: Bruce Sutter, Cubs

37 saves, 2.22 ERA, 1.89 FIP

Sutter was wrongly denied in 1977; not so two years later, when he won the NL Cy Young with another resplendent season. This is the third-best relief season ever by fWAR, giving Sutter two of the top three. That’s the lofty outcome, apparently, when a reliever registers a sub-2.00 FIP in more than 100 innings.

1980-1981: Rollie Fingers, San Diego and Milwaukee

23 saves, 2.80 ERA, 2.72 FIP; 28 saves, 1.04 ERA, 2.07 FIP

For almost the entire span from the mid-’70s through the mid-’80s, the Hall of Fame trio of Fingers, Gossage, and Sutter were in contention for the belt every season, and that battle was never closer than in 1980. Gossage had the best regular season of the trio, but he allowed a pennant-losing homer to George Brett that October. Fingers, meanwhile, had bounced back after an aberrantly walk-heavy 1979; after the 1980 season, a Sports Illustrated piece observed that Fingers “may be the best reliever in history,” but that “if Fingers isn’t the best relief pitcher in history, Sutter probably is.”

The Phillies’ Tug McGraw, with a 1.46 regular-season ERA and World Series title, was also in the mix. Yet his ERA had risen to 5.16 the previous year, so it would have been hard to trust him fully.

By a hair—a perfectly coiffed, handlebar hair—Fingers regains the belt because of a crucial tiebreaker: In 1980, he set a new career saves record. It’s hard to deny a record holder still performing at a high level.

Awarding Fingers, then a Brewer, the belt following the strike-shortened 1981 season should brook no argument—not after he led in saves, WPA, and both WAR calculations en route to winning the Cy Young and MVP.

1982-1983: Dan Quisenberry, Kansas City

35 saves, 2.57 ERA, 3.45 FIP; 45 saves, 1.94 ERA, 2.86 FIP

Perhaps with the exception of the knuckleballer Wilhelm, Quisenberry is the most unlike a modern closer of any pitcher on this list. Quiz was undrafted and scarcely a prospect at all; his oft-used sinker touched only the low 80s, and even for his era, he posted pitifully low strikeout totals. At his closing peak, he barely eclipsed a 10 percent K rate. But he combined pinpoint command with tremendous deception from his submarine delivery, and during his peak, he contended for awards every season, as well as broke Hiller’s decade-old single-season saves record in 1983.

How superlative was Quiz’s control? He threw four wild pitches in his career—fewer than the likes of Craig Kimbrel and Aroldis Chapman tally in a typical season. His catchers, Joe Posnanski wrote for The Athletic, never allowed a single passed ball.

“I’d never heard of him,” Royals great George Brett once said about Quiz’s arrival in the majors. “He looked funny, he threw funny, he was funny, and I wanted to know why we didn’t go out and trade for somebody. Now I know.”

1984: Guillermo “Willie” Hernández, Detroit

32 saves, 1.92 ERA, 2.58 FIP

This exercise doesn’t have many hard-and-fast rules, but here’s one: If a relief pitcher wins MVP, he automatically wins the belt, too. Hernández was a fine if unspectacular reliever when he landed in Detroit just shy of the 1984 season (in his second trade in a year). He responded by twirling 140 1/3 innings and leading all relievers in both types of WAR and WPA for a team that won 104 regular-season games and cruised to a title. Hernández finished five of the Tigers’ seven playoff wins.

1985: Dan Quisenberry, Kansas City

37 saves, 2.37 ERA, 3.05 FIP

If not for Hernández, Quiz would’ve kept the belt four years in a row. Alas, he’ll have to settle for three belt seasons, plus a 1985 World Series trophy and a number of near-awards. From 1982 to 1985, Quiz led the AL in saves every year and finished second twice and third twice in AL Cy Young voting, with the most total votes—by far—for any player in that span.

1986: Mark Eichhorn, Toronto

10 saves, 7 holds, 14-6 record, 1.72 ERA, 2.31 FIP

From one submariner to another: Eichhorn’s rookie season recalled the best years from Hiller and Marshall a decade before, as he threw 157 innings with half of his appearances going more than two frames. He would have stolen Roger Clemens’s ERA title had he thrown five more innings to qualify.

But this season saw another crucial development in the long arc of relief pitching: Dennis Eckersley completed his last season as a starter for the Cubs and went to Oakland—and manager Tony La Russa—in a trade before the next season. A new kind of closer was on the way.

1987: Steve Bedrosian, Philadelphia

40 saves, 2.83 ERA, 3.79 FIP

In Eichhorn’s follow-up campaign, his strikeouts dropped as his walks and home runs grew. He lost the belt. And by this point, the first generation of firemen was mostly extinguished: Fingers was retired while Gossage, Sutter, and Quisenberry were past their primes, and none of the trio made another All-Star team after 1985. Enter Bedrosian, whose questionable peripheral stats and presence on a sub-.500 team didn’t stop him from eking out an NL Cy Young victory in the closest vote in that award’s history.

1988-1992: Dennis Eckersley, Oakland

44 saves per year, 1.90 ERA, 2.06 FIP

From a 2021 vantage point, knowing how the relief pitching profession has evolved, Eckersley’s 1988 season might be the most important of any on this list. For the first dozen seasons of his career, Eck was a good—sometimes very good—starter, but after a challenging 1986 season both personally and professionally, and a trade to Oakland and La Russa, he transitioned to the pen.

In a fireman-esque role, Eckersley actually beat Bedrosian in the various rate stats in 1987—but it was the following season, when Eckersley became a true closer with a laser focus on the ninth and his innings total dropped from 115 2/3 to 72 2/3, that his legacy, and the entire position, really changed.

From 1988 to 1992, only 35 percent of Eckersley’s saves lasted more than one inning, and none were more than two innings. For comparison, from 1982 to 1985, 70 percent of Quisenberry’s saves had gone longer than an inning, and 33 percent had gone longer than two innings.

And with Eckersley’s unparalleled success for a team that reached three World Series in a row, and four ALCS rounds in five years, suddenly the whole league was moving to a one-inning-closer model. For a simple illustration of this shift, here’s a count of players with at least 20 saves and 100 relief innings in the same season, grouped by decade:

  • 1970s: 47
  • 1980s: 39
  • 1990s: 3
  • 2000s: 0
  • 2010s: 0

Compared to his peers, for belt purposes, Eckersley was untouchable in this stretch. In 1989, San Diego’s Mark Davis won the NL Cy Young award; Eckersley beat him in every important rate stat, then tallied four playoff saves en route to the title. In 1990, Bobby Thigpen set a record with 57 saves for the White Sox; Eckersley, though, completed a trifecta by leading all qualified relievers in ERA (an eye-popping 0.61), FIP, and WPA.

This five-year run culminated with an MVP for Eckersley—to this day, the last for a reliever. No reliever has finished even top three in the voting since.

1993: Duane Ward, Toronto

And then, all of a sudden, Eckersley was never the same world-beating force again; in 1993, his ERA rose above 4, and he never made another All-Star team. Thus, a power vacancy emerged and, with Mariano Rivera still working as a starter in Single-A ball, a gap in belt ownership.

Ward, long an understudy to Tom Henke in the Blue Jays’ bullpen, grabbed the leading role in 1993 with his teammate’s free-agent move to Texas. He shone in the spotlight, with a save total tied for the AL’s best and a fifth-place Cy Young finish, and followed with four playoff saves—plus the win when Joe Carter touched ’em all—as Toronto won its second consecutive title.

Ward was 29 years old in 1993, but threw just 2 2/3 more MLB innings as injuries derailed the rest of his career.

1994: Sonia Sotomayor, United States District Court for the Southern District of New York

This was the strike year, and befitting that complication, no reliever stood apart from the pack. The saves leader when the season stopped was Lee Smith, who ranked 33rd among qualified relievers in ERA, 43rd in FIP, and 50th in WPA. The ERA leader was Mike Jackson, who barely pitched after May because of an elbow injury. Jackson’s teammate, Rod Beck, was a perfect 28-for-28 in save attempts, albeit with a negative fWAR.

Sotomayor, on the other hand, was a federal judge—and future Supreme Court justice—who completed the most clutch “save” of the following winter when she closed down the owners’ plan to use replacement players, thus saving the 1995 MLB season. She’s the first lefty (get it?!) to hold the belt since Hernández a decade earlier. And sure, she might not have thrown a pitch to warrant this award, but she had a higher fWAR than Rod Beck.

1995: José Mesa, Cleveland

46 saves, 1.13 ERA, 2.70 FIP

Mesa finished second in Cy Young voting and fourth in MVP voting, and he led all qualified relievers in both ERA and saves while finishing games for a juggernaut in Cleveland. He pitched well in the playoffs, too, including a scoreless three-inning outing in a must-win World Series game. Still two years away from blowing a save in Game 7 of the World Series, he’s an obvious choice for the belt-holder this year.

1996-2013: Mariano Rivera, Yankees

Just kidding! An 18-year run for Rivera might be accurate—he was the first unanimous Hall of Famer for a reason—but it wouldn’t make for a very entertaining read.

So we’re going to be properly critical of Rivera when necessary and consider other relievers who excelled for stretches of his storied career—particularly early on, when he hadn’t yet established his legacy as the greatest closer ever. That said, the 1996 winner is still …

1996: Mariano Rivera, Yankees

5 saves, 26 holds, 2.09 ERA, 1.88 FIP

It’s strange to say this about the most decorated closer in history, but Rivera’s 1996 season—before he was the Yankees’ closer; before he had discovered the cutter that would propel him toward Cooperstown unanimity—might have been the best of his career.

Like with most pitchers in this exercise, Rivera moved to the bullpen as a last resort. As a 25-year-old rookie in 1995, he’d recorded a 5.94 ERA in 10 starts. And despite a strong showing in the pen that postseason—including 3 1/3 scoreless innings in his first playoff game, an extra-innings win—he still wasn’t viewed as a special player yet. Infamously, he was nearly traded to Seattle before the 1996 season for Felix Fermín; George Steinbrenner wanted a shortstop because he didn’t think that a 21-year-old named Derek Jeter could get the job done.

That dynasty-killing disaster averted, Rivera immediately demonstrated just how special he really was. As a setup man, Rivera occupied a sort of pseudo-fireman role, throwing an MLB-high 107 2/3 innings in relief in just 61 games, a near match for Sutter’s totals two decades earlier. And he didn’t just add length, but also recorded what would be the best strikeout rate of his career while allowing just one home run despite a leaguewide spike in homers.

By bWAR, Rivera’s ’96 season is tied for the most valuable season for any reliever since Eckersley redefined the closer role; by fWAR, it’s tied for second, behind just one season still to come.

And that was just in the regular season. In the playoffs, as the Yankees claimed their first title since 1978, Rivera allowed one run in 14 1/3 innings, going more than a single frame in all but one appearance.

Maybe all those stats aren’t necessary; just one quote can suffice to show why Rivera is the right belt-holder for 1996 despite not closing games. That April, in the middle of a run in which Rivera threw 14 hitless innings in a row, Twins manager Tom Kelly declared, “He needs to pitch in a higher league, if there is one. Ban him from baseball. He should be illegal.”

1997: Randy Myers, Baltimore

45 saves, 1.51 ERA, 2.77 FIP

Rivera was excellent in his first season as a closer—but in the ALDS, with a chance to close out Cleveland, he allowed a game-tying home run to Sandy Alomar. The Yankees lost in five.

So instead the belt goes to Myers, who would be out of the league a year later but first turned the clock back to his Nasty Boys days for one last hurrah. Since the official introduction of the save, Myers in 1997 is one of just five qualified relievers to lead the majors in both saves and ERA in the same season. He was the last to do so until 2019.

And while Myers, like Rivera, blew a playoff game to Cleveland that year, it’s hard to blame the lefty for his bizarre loss.

1998: Trevor Hoffman, San Diego

53 saves, 1.48 ERA, 2.04 FIP

Like Wilhelm before him, Hoffman, the Hall of Famer and one-time career saves leader, epitomizes how so many great closers fall into the role as a backup plan. In 1991, Hoffman moved from shortstop—his position in college and the early minors—to the mound. In 1992, he went to the Marlins in the expansion draft. In 1993, he pivoted to the Padres in a trade.

Finally in a stable home in San Diego, Hoffman started recording saves and didn’t stop for about a decade and a half. He was at his best early in his career: By bWAR, Hoffman’s four best seasons were 1994, 1996, 1997, and—most of all—1998. Hoffman actually received the most first-place NL Cy Young votes that season, but he was left off more ballots and finished second overall to Atlanta starter Tom Glavine. (Glavine became the first player to win the award without the most first-place votes.)

Remarkably, although Hoffman had 12 more seasons and 413 saves to go after 1998, he not only never won the belt again—he never even seriously challenged for it.

1999: Billy Wagner, Houston

39 saves, 1.57 ERA, 1.65 FIP

Rivera led the majors in saves in 1999 and might have reclaimed the belt this season, especially after a playoff run that ended with a World Series MVP. But in 1999, Wagner led all relievers in ERA, FIP, and WPA, and he set a new strikeout record, breaking the previous record he’d set the year before. Through the three-season stretch from 1997 through 1999, Wagner whiffed 40.4 percent of opposing hitters, and in 1999, he cut his walks and home runs, too.

The remaining competition is so fierce that, like Hoffman, Wagner wore the belt only this once, but he should be a Hall of Famer, too. Among all retired pitchers, he’s the second-most effective on a per-inning basis, behind only Rivera.

2000-2001: Mariano Rivera, Yankees

36 saves, 2.85 ERA, 3.28 FIP; 50 saves, 2.34 ERA, 2.28 FIP

During the Yankees’ three-peat from 1998 to 2000, Rivera converted all 18 of his postseason save opportunities; at one point, he didn’t allow a single run through a record 23 consecutive playoff appearances. He’s an easy choice to hold the belt after 2000.

But what about 2001? Is it wrong to pick Rivera for the season that he faced the very scenario outlined in the intro—bottom of the ninth, up by a run, in the most important game—and failed? Perhaps. But even counting Game 7 of the 2001 World Series, Rivera’s career playoff ERA after this season was 0.80 across 79 innings. Yes, with a zero to start. Part of the reason the Arizona loss was so surprising was because Rivera was the most trustworthy reliever in the sport, and one fluky inning involving an error and bloop single didn’t change that perception.

2002: John Smoltz, Atlanta

55 saves, 3.25 ERA, 2.39 FIP

With a 1.97 ERA and 1.80 FIP, Eric Gagné was the best reliever in baseball in 2002—but would he have claimed the belt with Rivera sidelined by three separate stints on the injured list? The bespectacled Dodger entered the season as a failed starter, with a below-average ERA in his career, while his chief competition was a former Cy Young winner in Smoltz who, in his first season as a full-time closer, set the NL single-season saves record.

That season’s Cy Young voting provided a natural tiebreaker: Smoltz finished third in the NL voting, one spot ahead of Gagné. Therefore, the belt is his.

(The most intriguing, off-the-wall possibility is young Francisco Rodríguez, who set the majors aflame in the 2002 postseason. But with just 5 2/3 regular-season innings in his career at that point, he doesn’t quite push ahead of a future Hall of Famer performing at a high level.)

2003-2004: Eric Gagné, Dodgers

55 saves, 1.20 ERA, 0.86 FIP; 45 saves, 2.19 ERA, 2.05 FIP

But not for long. Even though Smoltz raised his game in 2003—with a 1.12 ERA and 1.54 FIP—Gagné put forth the best single relief season on a per-inning basis in MLB history, as well as the most valuable, by fWAR, since the pre-Eckersley period. He won the NL Cy Young with 28 of 32 first-place votes; broke Wagner’s strikeout record, with a 44.8 percent K rate; and set a record with a perfect 55-for-55 save mark that may never be broken.

The 2004 season offers the most crowded field of the 21st century. In addition to Gagné, who ran his save streak to 84 in a row before blowing one, candidates included:

  • Rivera, who led in saves and bWAR and placed third in AL Cy Young voting, but played a role in the Yankees’ ALCS collapse
  • First-time closer Joe Nathan, who posted a 1.62 ERA
  • First-time closer Brad Lidge, who led in strikeout rate, fWAR, and WPA
  • K-Rod, who led in FIP and tied Lidge for the fWAR lead
  • Hoffman and Smoltz, future Hall of Famers with typically steady if not peak seasons of their own

Ultimately, given Gagné’s stupendous three-year run, nobody from that eminently qualified group could overtake him. He saved 84 games in a row! (Except for a blown save that didn’t count, because it came in the 2003 All-Star Game.) It’s hard to get more trustworthy with a ninth-inning lead than that.

2005-2006: Mariano Rivera, Yankees

43 saves, 1.38 ERA, 2.15 FIP; 34 saves, 1.80 ERA, 2.84 FIP

Injuries sidelined Gagné for most of the 2005 season, and Rivera was ascendant once again, with the lowest ERA of his career and a second-place Cy Young finish. The competition was tighter in 2006, with strong cases from the likes of Rodríguez, Nathan, B.J. Ryan, Takashi Saito, and rookie Jonathan Papelbon, who toted a 0.92 ERA—but by this point, Rivera’s track record meant only an anomalous stumble (or a truly magical season from a competitor) could get in his way.

2007: Jonathan Papelbon, Boston

37 saves, 1.85 ERA, 2.45 FIP

Speaking of anomalous stumbles: This was the worst season of Rivera’s relief career, as his ERA rose all the way to—brace yourself—3.15, a mere 44 percent better than average. (Every other season from 2003 through 2011, Rivera’s ERA began with a “1.”)

Across the AL East, Papelbon built on his stellar rookie campaign and led all qualified relievers in strikeout rate. In the regular season alone, he still wasn’t able to distance himself from Nathan, Saito, and J.J. Putz, who led in ERA and WPA—but Papelbon added 10 2/3 scoreless playoff innings en route to a title. Even as a youngster, he didn’t choke in the clutch. (He saved his choking, of course, for Bryce Harper in the dugout eight years later.)

2008: Brad Lidge, Philadelphia

41 saves, 1.95 ERA, 2.41 FIP

In 2008, Rivera bounced back with aplomb, to the tune of a 1.40 ERA and MLB-best WAR total. Nathan was phenomenal once again, with a 1.33 ERA. Rodríguez set a record with 62 saves (albeit in 69 attempts; as Sam Miller once wrote for ESPN, “the history of the save record is really the history of the save opportunity record”).

But nobody could overtake Lidge, who, between the regular season and playoffs, was a perfect 48-for-48 in save tries and won a title. If humanity were on the line, it would be hard to pick against someone who hadn’t goofed once in the previous year.

The next season, Lidge would post a 7.21 ERA and the worst single-season WPA for a relief pitcher in MLB history. Relievers are weird.

2009-2011: Mariano Rivera, Yankees

44 saves, 1.76 ERA, 2.89 FIP; 33 saves, 1.80 ERA, 2.81 FIP; 44 saves, 1.91 ERA, 2.19 FIP

If not for Lidge’s perfect season, Rivera would have reclaimed the award for a fourth distinct stretch (re-re-reclaimed?) in 2008; as it stands, he’s the easy choice for 2009 as both the best reliever in the regular season and the winning World Series closer for the final time.

And while he faced more pressing competition in both 2010 (from the Giants’ Brian Wilson) and 2011 (Valverde and an electric rookie named Craig Kimbrel), Rivera was firmly in “greatest closer ever” territory at this point. Maintaining both his minuscule regular-season ERAs and his playoff superiority meant nobody could surpass his status.

About that playoff superiority: From 2006 through 2011, across 24 postseason appearances, Rivera allowed one total run—and that came in a two-inning appearance in which he still converted the save. Although Rivera never approached his triple-digit innings total from 1996 again, at least in the playoffs, he more closely mimicked the firemen of yore. He didn’t enter games quite as early as Hiller or Marshall might have, but he was called upon to secure more outs than any of the other top relievers of the post-1988 era.

Best Relief Pitchers Since 1988 in the Playoffs

Player Games Outs/Game
Player Games Outs/Game
Mariano Rivera 96 4.4
Andrew Miller 29 4.0
Wade Davis 29 3.6
Aroldis Chapman 35 3.5
Brad Lidge 39 3.5
Kenley Jansen 49 3.4
Dennis Eckersley 27 3.4
Craig Kimbrel 20 3.3
Zack Britton 21 3.3
Only pitchers with at least 20 postseason games included; one playoff start apiece for Davis and Eckersley have been removed from their totals.

Rivera is the career leader in the following statistical categories in the postseason: games (96), ERA (0.70), saves (42, more than double anyone else), and cWPA (1.83 championships added). And in the 2011 regular season, for good measure, he passed Hoffman for the career regular-season save record.

2012: Craig Kimbrel, Atlanta

42 saves, 1.01 ERA, 0.78 FIP

With Rivera seriously injured for the first time in his career—he tore his ACL shagging flies during batting practice—the belt was up for grabs, and Kimbrel seized it by spinning the best FIP for a qualified reliever in MLB history. After being named NL Rookie of the Year the previous season, Kimbrel became the first pitcher ever to strike out at least half of opposing hitters (50.2 percent). With apologies to Fernando Rodney and his then-record-setting 0.60 ERA in Tampa, there’s no real argument here.

2013: Koji Uehara, Boston

21 saves, 13 holds, 1.09 ERA, 1.61 FIP

Although Kimbrel didn’t quite match his 2012 level in 2013 (1.21 ERA, 1.93 FIP, 38 percent strikeouts), he remained ludicrously effective and a fine choice to retain the belt. But Uehara, in his first year in Boston, was almost as dynamic in the regular season and even better in the playoffs, where he posted a 0.66 ERA and 16 strikeouts to zero walks. He even picked off a runner to end a World Series game, helping Boston to another title.

2014-2015: Wade Davis, Kansas City

3 saves, 33 holds, 9-2 record, 1.00 ERA, 1.19 FIP; 17 saves, 18 holds, 8-1 record, 0.94 ERA, 2.29 FIP

There are just nine seasons in the live-ball era in which a pitcher recorded an ERA of 1 or lower in at least 60 innings. Nobody has more than one such season—except Davis, who had two in a row. And he earned those numbers, with incredible mastery of the Three True Outcomes, most notably by not allowing a home run over 125 consecutive innings at one point.

For most of this span, Davis was a sort of Rivera ’96 candidate: the majors’ most dominant reliever, who just happened to be a setup man in Ned Yost’s rigid bullpen hierarchy. But he grabbed the closer job from Greg Holland—another strong belt candidate around this time—after the latter tore his UCL in September 2015. And all the while, he helped redefine a bullpen’s potential in the postseason, leading the Royals to consecutive World Series and one title.

Davis’s escape from a first-and-third, no-outs jam in the ninth inning, after a lengthy rain delay, to send Kansas City to the 2015 World Series ranks among the most exciting postseason moments in recent memory. During this two-year stretch in the playoffs, Davis allowed two runs (one earned) in 25 innings, with three wins, three holds, and four saves.

2016: Zack Britton, Baltimore

47 saves, 0.54 ERA, 1.94 FIP

In most seasons, 2016 Andrew Miller would have grabbed the belt due to his postseason heroics during Cleveland’s run to the World Series. But this wasn’t most seasons, because Britton set an ERA record, didn’t blow a single save, and, on this very website, inspired arguments to win both Cy Young and MVP.

The sinkerballing southpaw induced an 80 percent groundball rate, the highest mark on record, just ahead of Britton’s own high from the year before. He was incredible. It’s too bad nobody informed Buck Showalter before the wild-card game.

2017: Andrew Miller, Cleveland

2 saves, 27 holds, 1.44 ERA, 1.99 FIP

Kenley Jansen was by far the best relief pitcher in the 2017 regular season, leading the pack in ERA, FIP, and WPA. To start the season, he collected 51 strikeouts before issuing his first walk. But then the World Series rolled around, and Jansen blew a save in Game 2, allowed a home run in Game 4, and lost to a walk-off in Game 5. Unlike Rivera after 2001, Jansen enjoyed no such benefit of the doubt: He allowed as many home runs in this World Series as Rivera allowed in his entire 141-inning playoff career.

Who’s the best backup option? After a couple down years, Kimbrel had a strong case with a bounce-back season in Boston (1.43 ERA, 1.42 FIP), but Miller had been more consistently dominant for years at this point. He didn’t have the save total of an ace reliever, but from 2014 to 2017, the 6-foot-7 lefty recorded a 1.72 ERA in the regular season and a 1.10 ERA in the playoffs. More importantly, he inspired a sort of mini-reliever revolution, as other teams sought a Miller of their own. Remember the magical month of Chris Devenski? The superlative seasons for Chad Green and Archie Bradley? You couldn’t click on a baseball website in 2017 without stumbling upon a “next Andrew Miller … ” story. In part due to his own performance, then, and in part because he became the very namesake of the modern fireman role, Miller is a fine choice for the belt in 2017, his last elite season before declining.

2018: Blake Treinen, Oakland

38 saves, 0.78 ERA, 1.82 FIP

Might as well flip a coin between Treinen and 57-save man Edwin Díaz. The former has the ever-so-slight edge for two reasons. First, he placed one spot ahead of Díaz in the AL Cy Young vote; second, he was a much better fit for the “next Andrew Miller” role that was all the rave.

Of Díaz’s 57 saves, 54 were exactly one inning; the other three were four outs apiece. Treinen, conversely, threw more than one inning in 19 games this season, during which he collected a 0.52 ERA with 38 strikeouts to six walks. This is the only season that Treinen has ever made an All-Star team or received votes for an end-of-season award—but he made the most of his brief peak.

2019: Kirby Yates, San Diego

41 saves, 1.19 ERA, 1.30 FIP

Yates led all relievers in both ERA and saves in 2019—the first to pull the double since Myers in 1997—and the advanced stats liked him, too. Only Josh Hader—who allowed a bushel of homers and blew a save in the wild-card round—had a better strikeout-minus-walk rate than Yates, and nobody had a better FIP.

In a 17-month span from November 2015 through April 2017, before he blossomed into a belt-holder, Yates was traded twice for cash and waived twice more. He was extraordinary for a couple seasons; then, in 2020, he had a 12.46 ERA in six games before undergoing elbow surgery. One last time: Relievers are absolutely, unconditionally, maddeningly unpredictable.

2020: Liam Hendriks, Oakland

14 saves, 1.78 ERA, 1.14 FIP

Had the 2020 season featured a full schedule and Devin Williams twirled a 0.33 ERA, 0.86 FIP, and 53 percent strikeout rate over 60-plus innings rather than the 27 he actually threw, he would have grabbed the belt. But with the pandemic-shortened schedule, Hendriks is the best choice because he was just as effective in 2019 (1.80 ERA, 1.87 FIP) as 2020.

Hendriks signed with the White Sox in free agency this winter, becoming the first belt-holder since Myers in ’97 to change teams the following offseason. Playing for a new team, in a less pitcher-friendly home park, his hold on the belt may be tenuous. Let’s see what happens in 2021.

What about relievers who never held the belt? There are some surprising omissions. Eckersley’s half-decade run swallowed up the best seasons from a bunch of great closers, including:

  • Hall of Famer Lee Smith, the career saves record-holder until Hoffman
  • John Franco, the career leader in saves by a lefty
  • Nasty Boy closer Rob Dibble, the single-season strikeout rate leader until Wagner
  • Tom Henke, who had a 2.29 ERA and 143 saves from 1988-92—the same period during which Eckersley had a 1.90 ERA and 220 saves

Poor Joe Nathan finished his career with an 89.1 percent save conversion rate—the same mark as Rivera, and one of the five best in history (minimum 100 saves). But Nathan was unfortunate to overlap with Rivera and peak Gagné, and to spend his prime with the woebegone Twins, never giving him a playoff run to push to the top of the list.

Despite his overall dominance, Aroldis Chapman also never held the belt. Season-by-season, it’s difficult to find any one when Chapman was the best reliever—even 2014, when he posted a record 52.5 percent strikeout rate, Davis beat him in both WARs and WPA, making Davis a strong candidate even before factoring in his playoff dominance. As The Ringer’s Michael Baumann noted when consulted about this list of pitchers, Chapman is “probably the second- or third-best reliever in like six seasons.”

Jansen and Chapman are the two greatest snubs of the past decade, and both suffered from repeated playoff failures in clutch moments: Jansen by blowing multiple games in both the 2017 and 2018 World Series; Chapman by allowing the Rajai Davis home run in the 2016 World Series and series-losing homers to José Altuve in 2019 and Mike Brosseau in 2020.

Surveying the half-century-plus time period this exercise covered, Rivera is unsurprisingly the most-decorated belt-holder, with eight wins, followed by Eckersley and Fingers with five apiece and Quisenberry with three. Modern fans might not recognize just how great Quiz—the only non–Hall of Famer there—was at his peak. Rivera, Fingers, and Quiz are also the only pitchers to lose the belt and reclaim it.

In the present day, it’s difficult to imagine any recent belt-holder reclaiming it, given how steeply the likes of Kimbrel, Miller, and Treinen have fallen since their peaks. With relievers, of course, it’s impossible to forecast the future; maybe Hendriks holds the belt for another couple years to come, or Williams takes over with a full dazzling season, or Treinen, now a World Series winner with the Dodgers, recaptures his glory days.

The only confident prediction is that nobody will catch Rivera—in overall legacy, in Hall of Fame unanimity, in playoff success, or, most of all, in Relief Pitcher Championship Belt honors.

Thanks to Kenny Jackelen of Baseball-Reference for research assistance.

Off the Pike

Alex Cora on the MLB Offseason and the 2023 Red Sox

The Full Go With Jason Goff

Embarrassing the City of Chicago

Off the Pike

Should the C’s Stand Pat at the Trade Deadline? Plus, Chad Finn on the New Red Sox NESN Booth.

View all stories in MLB