Joe DiMaggio's 56-Game Hitting Streak: Why It Still Stands
Joe DiMaggio hit safely in 56 straight games in 1941. This article explains the streak, the arithmetic of why it is so unlikely, how Pete Rose's 44 compares, and what has changed in the game since.
From May 15 to July 16, 1941, Joe DiMaggio of the New York Yankees got at least one hit in every game he played, 56 in a row. No one has come within 12 games of it. The closest approach since is Pete Rose's 44 in 1978, and the record before DiMaggio, Willie Keeler's 44 in 1897, had stood for 44 years. The streak is the rare baseball record that ordinary fans can state from memory, and one of the few that mathematicians, physicists and statisticians have argued over for decades.
The reason it endures is simple to say and hard to appreciate: a hitting streak is a coin that must come up heads 56 times, and the coin is not fair. Even an excellent hitter fails to get a hit in about one game in six. This article looks at the streak itself, then at the arithmetic that makes it so improbable, then at the streaks that came closest and the changes in the game that make a repeat harder, not easier.
What happened in the summer of 1941
DiMaggio was already a star. He had joined the Yankees in 1936 and won the American League Most Valuable Player award in 1939. In 1941 he played 139 games, batted .357 with 193 hits, 30 home runs and 125 runs batted in, and won the Most Valuable Player award for the second time. His 125 runs batted in led the American League. His strikeout total for the season was 13 in 541 at-bats, against 76 walks, which is the line of a hitter who almost never misses.
The streak began on May 15 with a single against the White Sox. Early in the streak it drew little notice. By late June it was the story of the season: DiMaggio passed George Sisler's mark of 41 games and went on to tie and then pass Keeler's 44.
It ended on July 17 in Cleveland, where he went hitless against the Indians. Third baseman Ken Keltner made two strong backhand plays on hard-hit balls, and the reliever Jim Bagby Jr. retired DiMaggio in his last time up. The next day, he started another streak, which reached 16 games. The Yankees finished the season 101-53, won the pennant by 17 games over the Red Sox, and beat the Brooklyn Dodgers four games to one in the World Series.
For the streak itself, the commonly cited line is 91 hits in 223 at-bats, a .408 average, with 15 home runs and 55 runs batted in. Subtract that from his season and the rest of 1941 comes out to 102 hits in 318 at-bats, a .321 average. In other words, DiMaggio hit about 87 points higher during the streak than outside it, which is a clue to the mathematics. The streak was not a straight line of his talent. It was a hot streak on top of a great year, and it needed both.
The rules that decide what counts
A hitting streak counts games, not at-bats. A player extends it by getting at least one hit in a game, and he ends it by playing in a game and failing to get one. Under the scoring rules in the official rulebook, a game in which a player comes to the plate but has no official at-bat, because every plate appearance ended in a walk, a hit by pitch, catcher's interference or a sacrifice bunt, does not end the streak. A sacrifice fly, by contrast, does end it. A game in which the player does not appear does not end it either.
Those details matter because they decide edge cases, and because the streak depends on official scoring decisions. A ball that is ruled a hit instead of an error on a close play keeps a streak alive, and a ruling the other way ends it. Long streaks often depend on at least one close scoring call. The record is a count of official decisions as much as a count of swings.
The historical record does not carry game-by-game logs, which is why this article relies on the widely published figures for the streak itself and on the historical season lines for the rest. The season data is solid. The daily detail comes from other sources.
The arithmetic: why 56 is so hard
Take a hitter who bats .357 and gets four at-bats in a game. The chance of going hitless in a game is 0.643 to the fourth power, which is about 0.171. So the chance of getting at least one hit is about 0.829. The chance of doing that 56 games in a row is 0.829 raised to the 56th power, which is about 0.00003, roughly one in 36,000.
Now use the streak-time average of .408 instead. The chance of a hitless four-at-bat game drops to 0.592 to the fourth power, about 0.123, so the chance of a hit in a game rises to 0.877. Raised to the 56th power, that is about 0.00065, roughly one in 1,500. The difference between those two answers, 1 in 36,000 and 1 in 1,500, is the whole story of why the debate is lively. If you believe DiMaggio's true ability was .357, the streak was extremely lucky. If you believe he was a .408 hitter for those two months, it was not so strange.
| Hitter's average | Chance of a hit in a game (four at-bats) | Chance of 56 straight |
|---|---|---|
| .300 | 76% | about 1 in 4.7 million |
| .357 | 83% | about 1 in 36,000 |
| .408 | 88% | about 1 in 1,500 |
The table is a simplified model, and the simplifications deserve a mention. It assumes each game is independent, that the hitter always has four at-bats and that the probability of a hit is the same in every at-bat. None of that is quite true. DiMaggio sometimes had three at-bats and sometimes five, pitchers differ, parks differ and a hitter's probability changes with the opponent. The model overstates the odds in some games and understates them in others. It is useful for the order of magnitude, which is that for a .300 hitter, a 56-game streak is essentially out of reach, and for an exceptional hitter it is a freak event.
The other piece of the arithmetic is the number of chances. A single hitter has a very small chance of a 56-game streak in a given season. But thousands of players have played in more than a century of seasons, and any one of them might have produced a long streak. The right question is not what the odds were for DiMaggio, but what the odds were that someone would do it. That calculation is harder, and different researchers have reached different results, depending on how they model each player's true ability and how they treat his performance as a variable that moves from week to week.
What the experts argued
The best-known popular treatment came from the biologist and essayist Stephen Jay Gould, who wrote in 1988 that DiMaggio's streak was the one record in baseball that stood well outside what random variation would produce. His argument was that most records sit near the edge of what chance would produce, while 56 was far beyond the other long streaks in the data. Later work used computer simulation: build a model of each player's season, run baseball's history many times, and see how often a streak of 56 or more appears. Those studies found that the streak is rare but not impossible, and the answers depend heavily on how much a hitter's ability is allowed to change from stretch to stretch. This article does not give a figure from those studies, because the answer depends on the assumptions.
What is not in dispute is that the streak is an outlier. The next longest streaks are 44, twice, followed by 41 for Sisler in 1922, 40 for Ty Cobb in 1911, 39 for Paul Molitor in 1987 and 38 for Jimmy Rollins across the 2005 and 2006 seasons. The gap between first and second place is a wide one. For comparison, Dom DiMaggio, Joe's brother, hit in 34 straight games in 1949.
Pete Rose and the closest approach
Pete Rose of the Cincinnati Reds hit in 44 straight games in 1978, starting on June 14. He tied the National League record set by Keeler and passed Sisler's 41, and the pursuit of DiMaggio received national attention through July. It ended on August 1 in Atlanta against Gene Garber, who struck him out in the ninth inning in his final plate appearance. Rose finished 12 games short of the record.
There are two ways to think about that near miss. The first is that 44 is a long streak by any standard, and Rose, a high-contact hitter, was a plausible candidate. The second is that having fallen 12 short, he shows how far a good run can fall from the record. Streaks do not taper like a bell curve: they grow rarer faster than most people expect. Each additional game requires the hitter to survive another day at roughly 85 percent, and the cost compounds.
Why a modern hitter has it harder
The game changed in ways that cut against long streaks, and the 1941 season sheet shows one of them. In 1941 the major leagues averaged 3.55 strikeouts per team game. By 1995 the figure was 6.30. The ball is put in play far less often. Every strikeout is a plate appearance with no chance of a hit, so a hitter in a high-strikeout game has fewer chances to get one.
Pitching changed as well. In 1941 starters finished far more of their games than they do today, so a hitter would often see the same pitcher three or four times, which tends to help the batter. Today a starter leaves earlier and a series of relievers follow, usually chosen for a matchup, and each has been selected to be difficult for a batter of a particular hand. A hitter who has had a hit by the seventh inning is facing someone fresh and specialized on his fourth and fifth look.
Streaks of 20 games or more still turn up regularly and draw headlines. But in the whole history of the sport only five streaks have reached 40 games, and a streak of 56 needs a run that no one has yet matched.
DiMaggio's season and the Ted Williams question
The streak was one reason DiMaggio won the 1941 Most Valuable Player award, over Ted Williams of the Boston Red Sox, who hit .406 with 37 home runs, the most in the league. Williams's .406 remains the last time a major leaguer batted .400 for a full season, and many people then and since have argued that the voters picked the wrong man. Voters often reward the team that wins: the Yankees won 101 games and the pennant, while Boston finished 17 games back.
Seen together, the two feats make 1941 an unusual year. One player reached a number that nobody has matched, 56 games, and another reached a number that nobody has matched since, .400. The decision between them is not one that a database settles. It is a question of what the award is for.
The streak in a career
DiMaggio's career is easy to summarize: 1,736 games in 13 seasons, 2,214 hits, 361 home runs, 1,537 runs batted in, a .325 average and a .977 OPS. He missed 1943 through 1945 for wartime service, three years that would probably have added to every total. He struck out just 369 times in his career against 790 walks, and he hit 361 home runs, a rare balance of power and contact. He was elected to the Hall of Fame in 1955 and played in 10 World Series, with the Yankees winning nine of them.
The streak is the thing about him that grew in memory. Without it he would still be one of the best center fielders of his era; with it, he is a name that non-fans know. That is partly a result of the record's simplicity. A home run total requires context. A number of consecutive games requires none.
Common questions
How many games did the streak last? Fifty-six, from May 15 to July 16, 1941. It ended on July 17 in Cleveland.
Who has come closest? Pete Rose, with 44 games in 1978. Willie Keeler also hit in 44 games in 1897, but his streak came before DiMaggio's.
Does a walk end a hitting streak? No. A game in which a player has no official at-bat because of walks, hit by pitch, catcher's interference or sacrifice bunts does not end it, though it does not extend it either. A sacrifice fly does end it.
Did DiMaggio hit .400 that year? No. He hit .357 for the season. Ted Williams hit .406.
Will anyone break it? Possibly, in the sense that nothing makes it impossible. The arithmetic says that it would take an exceptional hitter and an exceptional run of luck at once, and the changes in strikeout rates and pitching usage make that harder than it was in 1941.
Where to go next
The streak has relatives among baseball's other records that look unreachable. Cal Ripken's streak of 2,632 consecutive games is the durability version, the last batters to hit .400 explains why Ted Williams's 1941 average has never been matched, and the 300-win club shows another number that the modern game has made unlikely. For the full list of those who have won the hardest batting prize, see every Triple Crown winner. The site's leaders pages let you sort single seasons by hits and average, and the history topic page collects more reading.