Return to Top

The Pioneers of Modern Relief Pitching

By Rob Neyer, December 20, 2013
Despite losing most of his right index finger in a farming accident at the age of seven, Mordecai “Three Finger” Brown went on to become one of the best National League pitchers of the early 20th century.

With the inevitable (but still regrettable) retirement of Mariano Rivera, we can look forward to his Hall of Fame induction in six years. Trevor Hoffman, who’s got more career saves than anyone but Rivera, might be elected even before Rivera. If both make it, that will be seven Hall of Famers inducted largely on the strength of their relief work (with the qualifier added mostly because of Dennis Eckersley, who was a fine starting pitcher for some years).

How much has baseball changed, though? While there are starting pitchers in the Hall of Fame who spent their entire careers in the 19th Century, the oldest relief pitcher in the Hall of Fame didn’t retire until 1972. Granted, knuckleballer Hoyt Wilhelm was nearly 50 when he finally hung up the spikes and put the oversized catcher’s mitt in the attic. But Wilhelm didn’t debut in the majors until 1952, more than 75 years after the National League sprouted into existence.

But the history of relief pitching, the history of great and important relief pitchers hardly begins with Hoyt Wilhelm and his baffling butterfly ball.

During a unique 20-year pitching career as primarily a relief specialist, Hoyt Wilhelm played for nine major league teams. In 1985, Wilhelm became the first relief pitcher inducted into the Baseball Hall of Fame.

Early in the 20th Century, the best relief pitchers were . . . starting pitchers.

From 1908 through ’11, Mordecai “Three Finger” Brown ranked among the National League’s best pitchers, going 102-43 in that span. But in addition to his 123 starts, Brown also made seventy relief appearances, and led the National League in saves in all four of those seasons.*

*Note: Saves were not officially defined until 1969, at which point they were retroactively figured for all the pitchers who had come before. Without comprehensive play-by-play data, a save was awarded for any relief outing that concluded a victory. Still, considering how close most of the games were in the Dead Ball Era, we might assume that most of Brown’s saves would qualify under today’s tighter definition.

In later years, other top starting pitchers – perhaps most notably, Brown’s fellow Hall of Famers Lefty Grove and Dizzy Dean – would mix in relief work with their brilliant work as starters. But when it comes to the creation of true relief specialists, two managers rise above all others: John McGraw and especially Clark Griffith.

Late in the 1904 season, McGraw, managing the New York Giants, purchased a right-handed pitcher named Claude Elliott from the Cincinnati Reds. Elliott pitched just a few innings for the Giants in ’04, but in 1905 he got into 10 games . . . and only two of those were starts. Of his eight relief appearances, six would later be counted as saves. In the parlance of the day, Elliott was perhaps the first dedicated (and so-called) “rescue pitcher.” However, he didn’t actually pitch well, and was never seen in the majors again after 1905.

New York Giants manager John McGraw became one of the first managers to use a “rescue pitcher” after purchasing Claude Elliott from the Reds in 1904. Elliott relieved eight times for McGraw’s 1905 champions but did not appear in the World Series.

McGraw tried again in 1906, but this time with Cecil Ferguson, a 19-year-old rookie. He started one game – and pitched a shutout! – but relieved in his other 21 appearances, and pitched effectively enough. Not effectively enough for McGraw, who soon shuffled Ferguson off to Boston.

Finally, in 1908 McGraw found the relief specialist he’d been looking for: James Otis Crandall. Within two or three seasons, Crandall became one of history’s more unique players: a good pitcher who worked mostly out of the bullpen and ranked annually as one of the National League’s better pinch-hitters. Beginning in 1909, his second season, the righty with the rubber arm led the league in games finished in five straight seasons. In 1910, Damon Runyon wrote, “Crandall is the Giants’ ambulance corps. He is first aid to the injured. He is the physician of the pitching emergency” – Hence, Crandall’s famous nickname “Doc” –  “without an equal as an extinguisher of batting rallies and run riots.”

James Otis “Doc” Crandall is generally considered the first true relief specialist of the Deadball Era. 

In the early 1920s, McGraw once again led the way, employing two relief specialists. One, curve-balling Rosy Ryan, started 46 games in those years, but also made 78 relief appearances and twice led the National League in relief victories. But the other, Claude Jonnard, started just four games in those three years, while relieving in 108 more. Granted, he “saved” only 15 games in those years, but the Giants were National League champions every year, and you might guess that McGraw’s tactics were quickly adopted by his managerial colleagues.

They were not. While most teams in the early ‘20s did have relief specialists, they were generally used when a game was out of hand, just to give the starting pitcher a break. Most managers remained loath to remove a starting pitcher who was actually winning, regardless of how well he was actually pitching.

But not McGraw, and not Washington Senators co-owner Clark Griffith. By the 1920s, Griffith had been an outstanding major league pitcher, manager of three different teams, and controlling partner of the Senators’ ownership group. After the 1920 season, he gave up the managerial reins to focus on the front office.

Back in 1904 and ’05, Griffith was the only pitcher in the majors who made more than half his appearances as a reliever, and he thrived in that role, posting a 2.26 ERA in the span. Nearly 20 years later, Griffith made a concerted effort to build the best specialized bullpen in the majors.


As a pitcher and player-manager, Clark Griffith was increasingly used a reliever for the New York Highlanders in the early 1900s. Griffith and his cross-town rival, John McGraw, are considered revolutionaries in the use of relief pitchers. 

First, just before the 1923 season Griffith traded for Allan “Rubber Arm” Russell, one of the 17 spitball pitchers allowed to continue throwing their wet ones after 1920 thanks to a grandfather clause. How good was Russell? “In 1923 Russell made 52 appearances for the Senators, 47 of them in relief, which was easily a new major league record,” historian Craig Wright notes. “He led in relief wins (9) and saves (9), and his 144.1 relief innings were a new record.”

But as Wright says, Griffith’s real find came during that same season, when Fred “Firpo” Marberry arrived. Fitting the later definition of an ideal relief specialist, Marberry had a great fastball but not much else. Griffith didn’t mind, and neither did young player-manager Bucky Harris; in 1924, as Wright points out, “Russell and Marberry finished first and second in the majors in relief appearances, and thus was born the first modern bullpen.” That fall, the World Series featured Griffith’s Senators and McGraw’s Giants, and the Senators’ four victories all included key relief stints: one by Russell, and three by Marberry.

The Senators won another American League pennant in 1925, and again Russell and Marberry played important roles. Russell was gone after the season, but Griffith replaced him ex-Yankees screwball pitcher Garland Braxton, who set a new (and long-standing) record with 146 relief innings in 1926. Marberry would ultimately start a fair number of games, but he remained one of the league’s top relief pitchers for some years. 

Fred “Firpo” Marberry is credited with being one of the first prominent relievers. Of his 551 games from 1923 to 1935, 364 were in relief.

After the 1932 season, Griffith traded Marberry to the Tigers. But he also acquired a sinkerballer named Jack Russell from the Indians. Russell led the league in saves in both 1933 and ’34, and in the latter year he became the first relief specialist to make an All-Star team.

Throughout the 1920s and early ‘30s, nearly all of the best relief specialists were Griffith’s men, as Allan Russell, Marberry, Braxton, and Jack Russell did most of the heavy lifting. But in the late ‘30s, a few other teams finally began to follow suit. In the American League, Johnny “Grandma” Murphy did yeoman work for Joe McCarthy’s dynastic Yankees, while submarine-screwball artist Clint Brown—once a pretty good starter with Cleveland—became a pure reliever with the White Sox and twice saved 18 games.

The National League’s managers weren’t as enthusiastic about Griffith’s innovation, and so it wasn’t until after World War II that an N.L. pitcher saved more than 15 games in a season. In 1950, the Phillies’ Jim Konstanty saved 22 games, went 16-7 with a 2.66 ERA, and was named Most Valuable Player. By then, American League All-Star Teams had included not just Jack Russell, but also the Yankees’ Johnny Murphy and Joe Page, Cleveland’s Russ Christopher, and Detroit’s Al Benton.

Of course, it would be many years yet before the “closer”—the one-inning relief specialist like Dennis Eckersley and Mariano Rivera—would be created, let alone the lefty specialist often asked to retire just one batter. But in the 1920s and ‘30s, Clark Griffith hit upon the novel idea that some pitchers were better-equipped for relieving than starting, and that some of those pitchers might actually have a better chance of holding a lead than whoever had started the game.

Yes, today that seems obvious. But a century ago, it must have seemed crazy.

Garland Braxton, Johnny Murphy and Clint Brown were some of the best relief pitchers during the first half of the 20th century.


If anyone has any additional information or questions about our artifacts and columns,
please do not hesitate to contact us at or