100 Years Ago: When Managers (Well, John McGraw Anyway) Upended Orthodoxies

Much of the discussion about baseball in today's day and age is about how advanced technologies and analytics increasingly inform managers' roster and dugout decisions, including positioning and strategy in game situations.  A century ago, managers themselves were at the forefront of sophisticated innovations that became part of managers' game-management toolkit that has endured to this day--using relievers to secure victories, making position player substitutions to try to win games, and platooning for advantage at the outset of games.    

100 Years Ago: When Managers Upended Orthodoxies

The Washington Post's longtime baseball writer Thomas Boswell wrote that the Nationals' switch from old-school Davey Johnson to first-time manager Matt Williams is "emblematic of the era."  Specifically: "The 21st century manager generally has a lower profile ... than most famous managers in the previous century, but he remains important because he is an extension of the analytical thinking of the entire organization.  Like good upper-middle managers, they implement the business plan."  http://www.washingtonpost.com/sports/nationals/inside-the-game-most-think-the-manager-can-make-a-big-difference/2014/03/27/7acd416a-b4d6-11e3-8020-b2d790b3c9e1_story.html.  One hundred years ago, major league baseball was also in the midst of an innovative transition in how managers did their jobs. The job of "baseball manager" had become ever more its own discipline, and its professionalism was evolving into greater complexity.

It was increasingly apparent that the most successful teams would be those that were not only the most talented and skilled in execution, but also the most sophisticated in their use of strategy to win games. So managers began to think more strategically about how to win games, which led to a reconsideration or refinement of three established orthodoxies. It should come as no surprise that John McGraw, who burnished his reputation as a baseball strategy genius by always looking for an angle and being willing to try unconventional things at a time when the game was still discovering itself, was at the leading edge of all three.

The first orthodoxy taken on by McGraw was that pitchers were expected to finish the games they started. and certainly victories.  At a time when relief pitchers were rarely in the game at the end of victories--(pitchers completed 88 percent of their starts and fewer than 3 percent of wins were "saved" in 1903, his first full season as Giants manager)--McGraw's genius was to realize that victories don't necessarily have to come from complete games and that sometimes bringing in a fresh arm to complete a game is the best way to secure a win.   Despite a starting rotation including Christy Mathewson and Joe McGinnity that was better than any in the league, with the possible exception of the Cubs, McGraw called upon a relief pitcher to save 102 of the Giants' 663 victories between 1903 and 1909.  That not only was 15 percent of the Giants' total, but accounted for fully one-third of the total 311 saves by National League teams those seven years. 

By the end of the decade, almost certainly because of McGraw's influence, NL managers in particular had seized on the notion of using a reliever to "save" a victory. (The "save" was half a century away from being a recognized pitching statistic, however.)  By 1914--one hundred years ago--the percentage of victories secured by a save had increased to about 13 percent.  While most managers had bought into this concept by now, albeit judiciously (complete games still being perceived as the best way to get the win), none had any one pitcher designated for a relief role; the pitchers getting the saves were established starters, many of them the ace of the staff.  With Doc Crandall from 1909 to 1913, however, McGraw was far ahead of his contemporaries in imagining or anticipating a future of designated relief pitchers--although even he backed off on that after Crandall.

The second orthodoxy McGraw upended was the one where managers rarely replaced anyone in the starting line-up during the game.  The prevailing wisdom at the time he became a manager was that, barring injuries or poor performance, seven of the eight position players in the starting line-up were the same from day-to-day (the understandable exception was inevitably banged-up catchers) and played every inning of every game; the players rounding out major league rosters who sat on the bench were there more for emergencies--to substitute for an injured regular, to give a regular an occasional day of rest, or to take over if the incumbent was ineffective--than for inclusion in the game at critical moments.  Not including pitchers, major league managers made an average of only 23 substitutions for position players in the field in 1903, but McGraw that year made 44.  Quick to see the possibilities in his never-ending quest to gain a key advantage, McGraw was much more inclined to pinch hit and sometimes pinch run for a position player in pivotal moments, which--if this occurred in any but the last inning--then required a defensive replacement in the field.

From 1908 to 1912, when his Giants were one of the powerhouse teams in baseball, McGraw made twice the number of position substitutions in the field (573) than the sixteen-team major league average (270). By 1914--one hundred years ago--substituting for position players during games for tactical advantage was an accepted practice; National League managers, following his lead, averaged 108 position player substitutions that year, but McGraw was still ahead of the curve with 130 of his own.  Indicative of the two leagues being somewhat different in their strategic approach to the game, American League managers lagged behind in position player substitutions, not achieving consistent parity with their NL counterparts until the early 1920s.

Taking position player substitutions to their logical conclusion--platooning in the starting line-up--upended the third established orthodoxy:  that a team should have a set line-up of core regulars, unchanging except for injury or a player proving ineffective at his position.  Although platooning certainly occurred to McGraw, he did not platoon in his starting line-up until a decade after he began making substitutions to gain the "platoon advantage" in key moments of games, and so it is Boston Braves' manager George Stallings who gets the credit for masterminding the concept.  Unlike McGraw, whose strong teams at the time generally had a dependable player to start regularly at every position, Stallings had no such advantage with the woebegone Braves when he became their manager in 1913.  He had no (as in zero) outfielders he felt comfortable starting on a daily basis.  

Stallings was widely regarded as a brilliant strategic manager even at the time, but what made his claim to fame in historical retrospect was how his master manipulation of eight players--four left-handed and four right-handed at any given time during the season--in an outfield rotation involving all three positions, including making substitutions to counter pitching changes, contributed to the 1914 Braves' miracle of rising from the bottom of the heap on July 4th, overtaking McGraw's indisputably better club in early September, and eventually sweeping the historically great Philadelphia Athletics in the World Series. What is unusual is how little attention was paid to this strategy at the time; there were no references to  Stallings' lefty-righty outfield trade offs depending on the opposing starting pitcher in any articles appearing in The Baseball Magazine, the premier publication on the sport at the time, in either 1914 or 1915.  By the end of the decade and through the 1920s, however, platooning was widely practiced by nearly all major league teams.

While George Stallings is the historical midwife of platooning and 1914 is considered the baseline year for that strategic concept, starting line-up data for 1914--the earliest year for which such data is available on the website baseball-reference--shows that both McGraw and Cardinals' manager Miller Huggins also had an outfield platoon that year, although at only one position.  The important point here is that these refinements in game strategies and tactics were more evolutionary than revolutionary; they were institutionalized by the collective wisdom of managers observing and learning from each other and becoming more strategic in their thinking.  Even if basic game strategies and strategies for employing players at key moments in games are now in place, the complexity of the game and its many nuances means there is always new insight and knowledge to be gained. Except today, as Boswell implies in his article, there is social engineering by managers making logical adaptations to not only what they observe on the field of play, but also based on the baseball-use revolution in advanced metrics and technology that can dissect performance. 

  

Hard Stuff, Stinginess, and Control in the Bullpen--A Historical Perspective, Part II

Previously on this blog we observed that ace relievers in the pre-expansion era rarely had the high strikeout per nine innings (K/9) ratios, high strikeout-to-walk (K/BB) ratios, and the very low WHIP (walks-plus-hits per inning pitched)--all far better than the league average for pitchers--that are characteristic of today's best closers.  This Baseball Historical Insight tracks the arc of the modern relief ace from Dick Radatz and Hoyt Wilhelm in the 1960s, through Dennis Eckersley (whose use and excellence popularized the term, "closer"), to the best of them all--Mariano Rivera.

Hard Stuff, Stinginess, and Control in the Bullpen--A Historical Perspective, Part II

The age of enlightenment in relief pitching began in the 1960s.  Even though a strong bullpen was increasingly recognized as an important ingredient to winning success in the postwar period, few teams followed the practice of developing or inserting a dominating pitcher into the role of relief ace.  Jim Konstanty, appearing in nearly half his team's games (74) with a 16-7 record and 22 saves while averaging more than two innings per relief appearance for the Phillies in the only stellar season of his career, had won an MVP award in 1950, but few relievers were considered even by their teams to be as valuable as front line starting pitchers. That changed in the 1960s when managers began relying on their ace relievers as never before, and their bullpen saviors were expected to be as dominating in their short stints as the best starting pitchers in the game.

Boston's Dick Radatz, known as the "Monster" (not to be confused with the Green Monster in his home ballpark) because of his intimidating 6-6, 230-lb. build and his intimidating fastball, was at the beginning of a wave of power pitchers in the bullpen to close out games that has persisted to this day.  From 1962 to 1964, when he averaged 10.6 strikeouts per year while appearing in 207 games and saving 78 of them (along with a 40-21 record), Radatz's ratio of 3.3 strikeouts for each walk was far better than the dedicated relievers who came before and comparable to power starting pitchers with excellent command and control, and his 1.1 WHIP during those three years was appreciably better than the league average.  Averaging two innings per relief appearance and pitching in 42 percent of the Red Sox' games during those years, however, took a significant toll, and Dick Radatz pitched only four more years after 1964.

Meanwhile, Hoyt Wilhelm--while striking out nearly 7 batters per 9 innings, better than the league average--set an unprecedented standard for stinginess in his six years as the White Sox' bullpen ace from 1963 to 1968.  In 358 relief appearances (and 98 saves) for Chicago, Wilhelm allowed fewer than one base runner per inning.  This was all the more remarkable because Wilhelm was a practitioner of the knuckleball, a pitch whose difficulty to control typically inflates a pitcher's WHIP. Despite his relying on a fluttering pitch, Wilhelm had an excellent 3.1 K/BB ratio during his ChiSox years that was better than most of the power relievers of his and the next generation of relief specialists.  For those who recall that he had been an ace in the New York Giants' bullpen from 1952 to 1956, Wilhelm's K/9 ratio (5.7), control (1.4 strikeouts per walk), and WHIP (1.3) were not nearly what he would achieve in the 1960s.

Over next two decades, most of baseball's best relievers had high K/9 ratios, including Goose Gossage, who fanned more than a batter an inning (657 Ks in 651.2 innings pitched) from 1977 to 1983--all but the first of those years with the Yankees; Bruce Sutter, who averaged a strikeout an inning in the five years he was ace of the Cubs' bullpen from 1976 to 1980; and Lee Smith, who averaged more than 10 Ks per 9 innings over six years from 1985 to 1990 (pitching for the Cubs, Red Sox, and Cardinals), and 8.7 per 9 innings over the course of 18 years and 1,016 relief appearances for his career.  Gossage and Sutter "closed" their way into the Hall of Fame, and Smith--now in his fourteenth year of eligibility--has been a strong contender, although his prospects for selection by the Baseball Writers Association of America may be dimming because of the prominent players now becoming eligible.  During the years mentioned for each, Gossage allowed fewer than one runner on base per inning three times in seven years and Sutter twice in five years; Smith did not break the sub-one WHIP barrier.  Sutter was the only one of these three to average better than three strikeouts per walk, including an excellent 5.6 strikeouts-to-walks ratio in 1977--his best year, when he had a 7-3 record and 31 saves for the NL East fourth-place Cubs.  A third Hall of Famer from this era, Rollie Fingers--possibly the most storied reliever until the coming of Mariano Rivera, if for no other reason than his handlebar mustache and pitching for the great (and dysfunctionally colorful) Oakland A's teams in the first half of the 1970s--did not match Gossage, Sutter, or Smith in K/9 ratio or WHIP in their five consecutive best years, but was much better than the league average for starting pitchers in both those categories, as well as in strikeouts per walk.
 
Enter Dennis Eckersley, the first of the modern closers to command the ninth inning, thanks to his manager Tony LaRussa's innovative scheme for the Oakland Athletics.  Eckersley mastered a closer's trifecta in his five best seasons with the A's from 1988 to 1992, during which he notched saves in 220 of his 310 appearances.  In 359.2 innings pitched in those five years, Eck fanned 378 batters for a 9.5 K/9 ratio while walking only 38 batters, yielding a superb control ratio of 9.95 strikeouts for every batter he walked.  Take away his 12 intentional walks on LaRussa's orders, and Eckersley had a K/BB ratio of 14.5 over five years, which trumps by far Koji Uehara's exceptional 11.2 K/BB ratio in 2013.  Eck not only had exceptional control but was about as difficult to hit as it gets, surrendering 6.2 hits per 9 innings, which contributed to his having a WHIP below 0.8 base runners per inning.  No other closer in history--not even Mariano Rivera--had Dennis Eckersley's combination of strikeouts per 9 innings, control, and stinginess over any five-year period.

Most of the great closers since Eckersley have been high strikeout power pitchers with a ratio of strikeouts to walks at least double the league average for all pitchers and a WHIP in the area of only about one base runner per inning.  Billy Wagner, for example, in a 15-year career from 1996 to 2010 had a strikeout ratio of 11.9 Ks per 9 innings, fanned four for every batter he walked, and had a WHIP just under 1 at 0.998 in 903 innings of relief; Trevor Hoffman in 16 years (1993-2008) charging out of the Padres' bullpen to save the day had a 1.04 WHIP, a K/9 ratio of 9.7, and a K/BB ratio of 4.04 in 902 games and 952.1 innings of work; and Joe Nathan in six years as the Twins' closer from 2004 to 2009 had a 0.94 WHIP, averaged 10.4 Ks per 9 innings; and had a 4.3 strikeouts-to-walks ratio.

And finally, appropriately and as always, the last word belongs to Mariano Rivera. The Sandman may not have matched the 1988 to 1992 Eckersley when it came to WHIP, K/9 ratio, and control in any five-year period, but while Eck had only six outstanding seasons as a closer, Rivera's entire 19-year career was outstanding.  In 1,105 games and 1,233.2 innings out of the Yankees' bullpen (including 1996 when he set up for John Wetteland and not including his 10 starts in 1995), Rivera's WHIP was below 1 at 0.97, he struck out 8.3 batters per 9 innings, and had 4.3 strikeouts for every walk he surrendered.  The last time he averaged better than a strikeout an inning was in 2009 (9.8 K/9) at the age of 39, but Rivera remained a control artist and nearly impossible to hit right to the end of his career.  Not including his injury-lost season of 2012 (when he appeared in nine games before ruining his knee shagging fly balls), Rivera allowed fewer than one runner per inning in six of his last eight years closing out Yankee victories.  In his swansong season of 2013, Mariano had a 1.05 WHIP, 7.6 Ks per 9 innings, and struck out six batters for each walk. Although Eckersley's closing brilliance was decisive in his making it to Cooperstown in his first year of eligibility, he had been a closer for less than half of his 24-year big league career, and voters took into account his full body of work--which included a first career as a starting pitcher.  That means Mariano Rivera will be the first pitcher to be elected to the Hall of Fame exclusively as a reliever on his first very ballot--a percentage certainty even greater than the certainty that the game is saved when the Sandman doth enter.

The (AL) Nationals'1924-25 Capital Success

With the Washington Nationals widely expected to repeat as National League Eastern Division winners this season, the following Baseball Historical Insight looks at the keys to success the last time a major league baseball team in Our Nation's Capital won back-to-back titles:  the Washington Nationals (also known as the Senators) winning American League pennants in 1924 and 1925. 

The Nationals' 1924-25 Capital Success

As the 1924 season approached, there was little reason to expect the Washington Nationals to compete for the American League pennant.  The Yankees had won three straight to kickstart their dynasty, including a dominating performance in 1923 when they won the pennant by 16 games, while the Nationals had finished fourth, sixth, and fourth under three different managers since 1921. An original American League franchise, the Nationals with only six winning seasons in their first 23 years had mostly been either deadenders or the definition of mediocre.  But the Nationals had a nucleus of accomplished veterans 30 years or older in all-time pitching great Walter Johnson, southpaw George Mogridge, shortstop Roger Peckinpaugh, and right fielder Sam Rice; an eight-year veteran just under 30 in Joe Judge at first base; and young hustlers in Muddy Ruel behind the plate, second baseman Bucky Harris, third baseman Ossie Bluege, and left fielder Goose Goslin.  With this foundation, the Nationals surprised the baseball world by beating out the Yankees by two games to grab the 1924 title, and winning the World Series beside, and repeating in 1925 (except for the World Series part), taking the pennant by 8-1/2 games over second-place Philadelphia as the Yankees never recovered from not having Babe Ruth--out with a mysterious stomach ailment--for the first month-and-a-half of the season and finished seventh.

Three factors were key to Washington's unexpectedly Capital Success.  The first was Washington owner Clark Griffith's decision to change managers yet again, turning to his second baseman to take charge of the team after at first considering trading him for Eddie Collins so that Collins could be manager.  At 27 with only four major league seasons behind him as a player, Bucky Harris was younger than all but two of the Nationals' position regulars; only Goslin and Bluege--both 23--were younger than he.  Griffith, therefore, was careful to secure support for his decision to make Harris manager from Johnson and Rice, Washington's two biggest stars.  Harris proved a dynamic and an innovative leader, his innovation being introducing the concept of a dedicated relief ace to major league baseball.

That would be Firpo Marberry, the second key factor in the Nationals' success because they would not have won the 1924 pennant and might not have repeated in 1925 without him as their relief ace.  Not until the 1920s was the pitcher with the most saves typically a "genuine" relief pitcher, and they were not elite pitchers groomed for that role.  Being a "relief ace" was definitely not the road to career success, particularly at a time when most teams were still using established starting pitchers to come in from the bullpen to win or save close games that the day's starting pitcher could not himself close out.  Harris's insight was to dedicate a young pitcher with the potential to be a first-rate starter almost exclusively to relief.  (Marberry did start 14 games in 1924, along with 36 relief appearances, but was not used as a starting pitcher at all in 55 games in 1925.)  Marberry, much in the manner of closers today, not only threw hard--very hard--but displayed a fearsome, stalking, intimidating presence on the mound.  Harris made 273 pitching changes in his back-to-back pennant-winning seasons, using Marberry 91 times.  In almost exactly half of those games, Marberry got either the win (15) or the save (30).  He also was the losing pitcher 11 times, meaning Marberry figured directly in the game's outcome 62 percent of the time he was called in from the bullpen.

The decision to use Marberry's talent in the relief role was critical because Harris inherited a pitching staff whose two best pitchers--Walter Johnson and George Mogridge--were both 35 or older, and in 1925 Mogridge was replaced by 35-year old Stan Coveleski.  Having a pitcher as fine as Marberry to rely on in the bullpen allowed Harris to better pace his aging staff aces.  With the reliable Marberry as Washington's relief ace in the hole, the Nationals' old-guy starters benefited by not necessarily having to complete their own victories.  Marberry saved 11 of Johnson's 43 wins in 1924 and 1925, four of Mogridge's 16 wins in 1924, and four of Coveleski's 20 victories in 1925.

Marberry's role helped rejuvenate the twilight of Walter Johnson's career.  The Big Train had not won 20 games since 1919, but had nonetheless completed 70 percent of his starts and had also appeared in 27 games in relief.  With a pitcher of Marberry's excellence in the bullpen, Johnson no longer needed to make relief appearances to save wins for Washington and could be bailed out by Marberry in games he started.  The 36-year old Johnson was 23-7 in 1924, leading the league in wins, winning percentage, ERA, and strikeouts, and was the league's MVP.  In 1925 he went 20-7.  Johnson appeared only once in relief those two years, and completed 36 of his 67 starts--a more reasonable, given his age, 54 percent.

Thirty-five-year old Mogridge, meanwhile, went 16-11 in 1924, completing 13 of his 30 starts, after having completed 57 percentage of his starts in his three previous years in Washington.  And Coveleski, after a relatively mediocre season in 1924 with Cleveland that might have marked the beginning of the end of his great career, was 20-5 with the Nationals in 1925, leading the league in winning percentage and ERA (same as Johnson the previous year), but completing only 15 of his 32 starts.

The final key to Washington's success was decisively winning the season series against their principal rival for the pennant both years and playing well down the stretch to secure the American League flag. In 1924 the Nationals won 13 of their 22 games against the Yankees, including a four-game sweep in June at Yankee Stadium that thrust them into first place in the midst of a 10-game winning streak that began when they were in sixth.  The most pivotal games, however, were in the last days of August when Washington took three of four from the Yankees, again in the enemy's lair at Yankee Stadium, to take a game-and-a-half lead into September.  The Nationals remained in first place the rest of the season, but never by more than two games until clinching the pennant on the next to last day of the season.  Harris guided his team to an 18-7 record in the final month, including winning seven of the ten games they played in September with first place directly at stake--either tied for first, no more than a game ahead, or no more than a game behind--which proved absolutely necessary because the Yankees kept the pressure on Washington down the stretch with an 18-8 record of their own and owned a share of the top soil for three days in mid-September.  The Nationals went on to win a thrilling seven-game World Series against the New York Giants.

And in 1925, Washington was 13-7 against the Philadelphia Athletics, themselves making an improbable run for the pennant after years in the wilderness following owner-manager Connie Mack's break-up of his great 1910-14 team.  Although the Nationals won in the end by a blowout margin, it was a tight pennant race for most of the summer.  From July 15 until August 19, the Nationals were never alone in first place, mostly keeping pace in second, typically about two games behind the Athletics.  On August 27, Washington held a mere half-game lead after losing four in a row, but won nine of their next ten to open up a nine-game lead, while Philadelphia was mired in a 12-game losing streak. This time, the Nationals lost a thrilling seven-game World Series, to Pittsburgh.

Introducing Baseball Historical Insight

More than any other sport, baseball's history informs the present.  Baseball honors its history with both import and reverence, which helps keeps alive its heroes of the past as totems against which the accomplishments of current players, teams, and even managers are invariably measured.  Unlike for statesmen, for whom it is said that to be unaware of history is to be condemned to repeat mistakes of the past, for baseball fans, to be unaware or unmindful of the game's history is to be missing out on what makes the game so special and endearing.  As we are about to be caught up in the drama of the 2013 season and follow the daily highlights, my purpose in Baseball Historical Insight is to remind baseball fans of the legacy implicitly carried by every player, every manager, and every team in the major leagues today.

The following is the first Historical Insight:

JOHNNY MURPHY TIME

As Mariano Rivera embarks on the final season of his Hall of Fame career,it is worth remembering that he follows in a New York Yankees' line of exceptional relief pitchers--typically the best in baseball, or at least the American League, during their time in pinstripes--dating back to Johnny Murphy, who pioneered the role in the mid-1930s.

Despite the success the Washington Senators had winning back-to-back pennants in 1924 and 1925 using hard-throwing right-hander Firpo Marberry nearly exclusively in relief--which was unprecedented for a pitcher with his talent--there was still no epiphany in major league baseball that "relief ace" deserved to be its own pitching discipline, that good pitchers might be developed and used for years specifically in that role.  In 91 relief appearances those two years, in which he finished 70 games, Marberry proved the value of his role by winning 15 games while saving 30 of Washington's 188 victories.  And thus was the relief ace born.

But the concept of a dedicated relief ace was slow to catch hold.  Major league teams were not especially warm to the idea of using talented young pitchers like Marberry in relief roles for very long before making them starters, and starting pitchers were expected to get complete game victories.  Consequently, most teams continued to use multi-role pitchers who both started and relieved to "save" close games, or auditioned pitchers in the relief role before either moving them after a season or two into the starting rotation or determining that they were unlikely to achieve starting-pitcher quality and letting them go.  And rather than use the designated relief pitchers in any given year to "save" victories, managers typically relied on some of their top starting pitchers for such an important task.  Lefty Grove and Dizzy Dean, for example, were both top-flight starting pitchers who, in their peak years, also routinely led their team in saves.  Most pitchers who worked primarily in relief spent much of their time replacing starting pitchers who were knocked out early, or working in games that were lost causes.  Obviously, no self-respecting pitcher with ambition wanted any part of "relief pitcher" as a career path.

It wasn't until the Yankees in 1935 that another team specifically groomed a young pitcher--in their case, Johnny Murphy--to be an ace reliever as a career calling.  The previous year, as a 25-year old rookie, Murphy had started 20 games and relieved in 20 games, and pitched well in both roles, but his manager, Joe McCarthy, did not see Murphy as a front-line starter--at least not on a staff that also included future Hall of Famers Red Ruffing and Lefty Gomez.   Deciding to divide his pitching staff into exclusively-starters and exclusively-relievers, McCarthy persuaded Murphy to give up whatever dreams he may have had of being a starting pitcher to become his ace in the bullpen.  Unlike other managers of his day, McCarthy in all his years of winning success with the Yankees rarely used his top four starters out of the bullpen.  Instead, he routinely carried three (sometimes four) pitchers he used almost exclusively in relief, the best of whom--Johnny Murphy--was both his "fireman," coming in late in the game with runners on base to douse the flames of an opposition rally (and, indeed, one of Murphy's nicknames was "Fireman"), and his "closer" (although the term was not yet used in a Mariano Rivera context), coming in specifically to save games.

McCarthy used Murphy judiciously and efficiently.  From 1935 to his final season with the Yankees in 1946 (not counting the two years--1944 and 1945--he served in the military during World War II), Johnny Murphy appeared in as many as 38 games in relief only once (in 1939) and relieved in 35 games or more only four other times.  But his relief appearances were invariably productive.  Working about two innings per relief appearance, and leading the American League in saves four times, Murphy pitched 262 games in relief for the Yankees from 1936 to 1943, when they won seven pennants in eight years. His record in relief during those seasons was 60 wins and 34 losses with 88 saves, meaning that in 70 percent (182 games) of his total relief appearances, Johnny Murphy figured directly in the outcome of the game.  McCarthy was clearly using Murphy when it mattered the most.

Indicative of McCarthy's genius in handling his pitching staff during the Joe DiMaggio-era Yankee dynasty, the New York Yankees led the league in both complete games and saves in 1937, 1939, and 1942.  This had been done only once previously, by the 1926 St. Louis Cardinals, and would be done only once more in the twentieth century--by the 1988 Los Angeles Dodgers.  The significance of this hardly needs to be said:  at least until the 1990s, when complete games plummeted to less than 10 percent, teams with starting pitching good enough to lead the league in complete games did not need to go to relief pitchers so often that their pitching staff also led the league in saves.  That happened only five times in the twentieth century,including three times in six years by a Yankee team that had excellent starting pitching and a strong bullpen led by Johnny Murphy.  Manager McCarthy was more than astute in how he employed his pitching staff.  He gave his starting pitchers every opportunity to complete a victory, but was intuitive enough to know when it was . . . Johnny Murphy time.