Bill James Guide to Baseball Managers, The (47 page)

Do players today jump from team to team more often than they did in previous eras? It is a commonplace of sports journalism to assume that they do. This is usually cited in a pejorative fashion, as in “How can the teams today expect fans to develop any strong loyalty, when they know the players they root for will be gone in a year or two?”

This is my third published study of the subject.

The first study, published in the late 1970s, looked at the issue by taking a fixed group of players from a base season (1910, 1920, 1930, 1940, etc.), and then checking to see how many of those players were still with the same team one year later, two years later, three years later, etc. The conclusion at that time: There was no such change. The percentage of players who were still with the team one year later or two years later or five years later was the same in 1975 as it had been in 1965, or 1955, or 1915.

The second study, published in the mid-1980s, showed the same thing: There was still no change in the rate of player retention, given a fixed set of players. I added a second study of the issue at that time, designed to answer essentially the same question in a different way. Suppose that you focus on all players as they play their 500th major league game, their 1,000th major league game, their 1,500th major league game, etc., and you ask for each player “Is he still with his original major league team?” This study also found no change. A given player playing his 1,500th major league game in 1985 was as likely to be with his first team as he would have been in 1975, or 1955, or 1935.

I used a slightly different method this time. This study focused on five-year intervals—1990–1995, 1980–1985, etc. The exact question that I asked was this: Given all the major league players who were active in both 1990 and 1995, how many of them were still with the same team at the end of 1995 that they were with at the beginning of 1990? And how does this compare with earlier eras, 1980–1985, 1970–1975, etc.?

This study shows that what sportswriters have been saying for at least twenty years is now true: players today
do
change teams more often than they did in the past.

There were 268 players who played in the major leagues in both 1960 and 1965. Eighty-four of those 268, or 31%, played with only one team throughout that era.

The data from 1970 to 1975 were essentially the same, except that the major leagues had expanded. There were 344 players who played in the majors in both 1970 and 1975. Of those 344 players, 103, or 30%, were still with the same team.

From 1980 to 1985, despite the beginning of the free agent era, the percentage actually
increased
slightly, to 32%—128 of 406.

But when 1995 is compared to 1990, there are 444 players who played in both seasons. Only 100 of those 444 players, or 23%, were with the same team in both seasons. Lets put that in chart form:

The data from the earlier eras, 1950–1985, suggests that about 21% of all major league players changed teams each year, and 79% did not (79% to the fifth power is 31%.) The data from 1990 to 1995 suggests that the one-year moving percentage has increased from 21% to 26%—a substantial change.

In the 120-year history of major league baseball, the standard use of relievers has never arrived at a static equilibrium. The rules effectively governing the use of relief pitchers have been in constant motion—sometimes moving fast (like now), sometimes moving a little more slowly, but always moving. The bullpen of 1996 is radically different from the bullpen of 1986—as the bullpen of 1986 was different from that of 1976, 1976 was different from 1966, and 1966 was different from 1956.

As a starting point, we may assume that baseball began without relief pitching. This is not technically true. Even in 1876, the first season from which decent records may be reconstructed, Harry Wright’s Boston Red Caps used 21 relief pitchers in 70 games. The rules originally allowed a pitcher (or other player) to be replaced when he was injured or with the consent of the other team.

So relievers do not start out at true zero, and Dan Quayle is not a complete idiot, but it’s close enough for government work, and we’ll mark it at zero. The rules were changed to allow substitutions without challenge in 1891, and the number of relievers used began to edge upward. For a hundred years since then, it has gone only upward:

“GIR” stands for “games in relief,” “MLG” is “major league games,” and “RePG” stands for “relievers per game.” The number of relievers used per game has gone down only in the 1970s, and that was simply an effect of the designated hitter rule. The use of relievers went up in the National League in the 1970s, but dipped by more than 20% in the American League in 1973, when the DH rule was adopted.

You will note that the number of relievers used per game is now surging upward more rapidly than at any other point in baseball’s history, but the numbers are getting us ahead of the narrative. There were two strands in the early development of the bullpen, the “relief ace” strand and the “starter as closer” strand. The latter strand is best represented by Mordecai Brown and Albert Bender. Brown rivaled Christy Mathewson as the best pitcher in the National League in his time, compiling records of 26–6, 29–9, and 27–9 with ERAs, over a four-year period, never getting
higher
than 1.47. He also led the NL in saves, retrospectively figured, in 1908, 1909, 1910, and 1911. Bender was an American League contemporary, one of the best pitchers in the American League at that time. The two men shared the save record for many years, with 13 apiece, although neither pitcher was aware of this until he had been dead for a number of years.

Doc Crandall was the top reliever of the era, not counting the pitchers like Brown and Bender, who worked 80% of their innings as starters. As early as 1906, John McGraw had used a teenager named George Ferguson to close out some victories (six or seven saves, depending on the source). McGraw had visions of Ferguson becoming a great pitcher, but Ferguson developed bad habits, and the relief job passed to Doc Crandall. Crandall was a part-time infielder, a lifetime .285 hitter who was used by McGraw as a backup second baseman, shortstop, pinch hitter, and reliever. Christy Mathewson wrote about him in
Pitching in a Pinch:

The McNamaras were the Tim McVeighs of their day, a pair of radical union leaders who left a bomb at the offices of the
Los Angeles Times
, killing twenty-one people. Damon Runyon wrote that Crandall was “without equal as an extinguisher of batting rallies” and “the greatest relief pitcher in baseball.” These quotes are borrowed from
Pen Men
.

Crandall pitched 23 games in relief in 1909, a major league record at the time, then upped the record to 24 in 1910, to 26 in 1911, to 27 in 1912, to 32 in 1913. Other pitchers chased just behind him; by 1913, a half-dozen other pitchers were pitching as many games in relief as Crandall had in 1909.

While Crandall pitched 30 games a year out of the bullpen, he also started 10 to 20 games a season, not to mention pinch hitting and playing second base. As a reliever he went 7–1 in 1910, 7–0 in 1911. What does this tell you?
He was almost always used when his team was behind
. Unlike the modern relief ace, who is used only when his team is
ahead
, Crandall normally came in when the starter was ineffective, and thus, when the team was behind. Crandall had almost twice as many relief wins as saves.

Crandall was a good pitcher. He was not as good as his won-lost records suggest. He was 18–4 in 1910, 15–5 in 1911, but in addition to entering many games protected from a loss by the fact that his team was already behind, he started 33 games over those two seasons, most of them against bad teams. McGraw would use Mathewson and Marquard and Tesreau against the other contenders, figuring that Crandall could beat the Braves and the Cardinals.

Crandall was far from a pure reliever; still, he is the first career reliever of any significance. Crandall pitched in relief 168 times in his career, the most of any pitcher pre-1920. Just two games behind him was his longtime teammate, Red Ames. Ames was a better starter than Crandall; he started about twenty-five times a year, pitched in relief about ten times a season, and had a long career, adding up to 166 relief appearances, although he was basically a starting pitcher.

Sad Sam Jones pitched 39 games in relief in 1915, taking the record away from Crandall, and Dave Danforth pitched 41 games in relief in 1917. These were young pitchers who were trying to establish themselves as starters. When they had some success as relievers, they moved into a starting assignment. Bernie Boland pitched 37 relief games for Detroit in 1916, George Cunningham 36 games for the same team the next season. They were the same, young pitchers trying to get established. Thus, we should note, the cliché about relievers being old broken-down starters is distinctly
untrue
about the first generation of relievers. The bullpen started not as a refuge for old starters, but rather as a kind of try-out camp for young, unproven starters.

This article is not about relief pitchers. That story has been told, and very well, by John Thorn in
The Relief Pitcher
(E. R Dutton, New York, 1979), and Bob Cairns in
Pen Men
(St. Martin’s Press, New York, 1992). This article is about strategy, about how relief pitchers have been used by their managers over time.

In early August 1923, the Washington Senators purchased the contract of Fred Marberry, a twenty-four-year-old Texas right-hander. The press called him “Firpo” because he looked like a boxer of that name, but Marberry didn’t like the name and was called Fred by those who knew him. Marberry had a fastball and nothing much else. The Little Rock team, from whom the Senators had purchased him, had been using him mostly in relief, and Senators manager Donie Bush continued in this way, giving Marberry four starts late in 1924 (all of which he won), but using him seven times in relief.

In the history of relief pitchers, Firpo Marberry is the dog that didn’t bark. What is exceedingly curious about Firpo Marberry is that there
isn’t
a Firpo II, a Firpo III, and a Firpo IV In 1924 the Senators gave Bucky Harris, their young second baseman, the opportunity to manage the team. Harris used Marberry, in 1924, more or less as other young relief pitchers had been used—as a spot starter against weak teams and a patron of nearly lost causes.

As the season wore on, however, Harris gained confidence in Marberry and began using him to
save
games. The term “saves” was not used, but Marberry was the first pitcher aggressively used to protect leads, rather than being brought in when the starter was knocked out. Thus Marberry is, in my opinion, the first true reliever in baseball history. He pitched 55 times in 1925, always in relief, and saved 15 games. He saved 22 games in 1926, retrospectively figured. It would be many years before anyone would save more. He was a modern reliever—a hard-throwing young kid who worked strictly in relief, worked often, and was used to nail down victories.

And the Senators won. The Washington Senators, the most famous down-and-outers of their time, won their first American League pennant in 1924, and the World Series, and their second American League pennant in 1925. They had been sub-.500 in 1922 and 1923; they added Marberry, used him in relief, and won two pennants.

It should have sent shock waves through baseball. “Hey, this reliever stuff really works!” people should have said. This is what happens when someone uses a new and unique strategy, and wins two league championships with an also-ran team.

What happened here was, everybody got credit for the Senators’ success except Marberry. Bucky Harris became one of the biggest stars in the game, the subject of a hero-worship biography. He would manage another twenty-five years on the coattails of his 1924–1925 success. Walter Johnson got credit for it. He was named the MVP in 1924. Roger Peckinpaugh got credit for it; he was named the MVP in 1925. (He then made six errors in the 1925 World Series, prompting a sportswriter to comment that he was not only the MVP of the American League, but of the National League as well.) Sam Rice and Goose Goslin and Joe Judge and Stan Coveleski got credit for it. Firpo Marberry? He was just a reliever.

Even Bucky Harris didn’t follow through on Marberry’s success. He didn’t use Marberry in the seventh game of the 1925 World Series, and Marberry began to slide gradually out of the “closer” role. He started five times in 1926, ten times in 1927. Eventually Harris was replaced, and Marberry began to start more and more.

Marberry’s success was reinterpreted as a failure. Muddy Ruel said about Marberry that “he couldn’t keep the pace through a whole ball game. He was invincible for two or three innings, but he couldn’t go the distance.” This quote, which is borrowed from John Thorn’s
The Relief Pitcher
, is patent nonsense. “Marberry pitched relief,” people said, “because he
couldn’t
start. He was an oddity—a good pitcher, but not able to start.”

Firpo Marberry made 187 starts in his career, and won 94 of them. Marberry won a higher percentage of his career starts than many Hall of Famers, if not
most
Hall of Famers, including his contemporaries. He won a better percentage of his career starts than Don Drysdale, Waite Hoyt, Eppa Rixey, Ted Lyons, Steve Carlton, Tom Seaver, Red Faber, Jesse Haines, Catfish Hunter, Ferguson Jenkins, Robin Roberts, or Red Ruffing. This does not prove that he should be a Hall of Famer or that he would have been a Hall of Fame starter, but to suggest that he failed as a starting pitcher is absurd.

Baseball historians, looking backward, have tended to hang a line between Firpo Marberry and twenty-first-century relievers, and to say that this is the seminal experience of the modern reliever. In fact, the line that begins with Firpo Marberry, ends with Firpo Marberry. It ends, actually, in the middle of Marberry’s career.

What happened is that conventional assumptions had, for the moment, proved stronger than the power of a successful innovation. The traditional way of thinking about pitchers was that

a) a starting pitcher should finish the game whenever possible. If he doesn’t finish the game, he hasn’t done his job.

b) the only pitchers on the staff who really count are the starters. The relievers are just substitutes, bench warmers.

These ideas were so deeply entrenched in the minds of baseball men that it was easier to re-interpret Marberry’s accomplishments as a failure, or at best as a function of Marberry’s “peculiar” limitations, than it would have been to deal with the message of Marberry’s success. Marberry himself said that he much preferred to be a starting pitcher. Bucky Harris, who had made Marberry the first modern reliever, managed until 1956, during most of which time he backed steadily away from the use of relief aces. Managers simply were not ready to accept the idea that a relief pitcher could be as valuable as anyone on the staff.