The Lowest ERAs in Baseball History: 1.12 and the Pitcher-Friendly Years

By 4Baseball Staff · November 5, 2026 · 10 min read

Bob Gibson's 1.12 in 1968 is the lowest earned run average of the live-ball era, but Dutch Leonard in 1914 and Pedro Martinez in 2000 did more against their leagues. Here is how to compare them.

Bob Gibson's 1.12 earned run average in 1968 is the lowest figure recorded by any pitcher with 150 or more innings since 1920 in the historical record. It is not the lowest ever recorded, though, and by one fair measure it is not even the most dominant season relative to its league. Which of those statements matters depends on what you want an earned run average to tell you.

This article lays out the lowest seasons in the database, then does the comparison that raw numbers hide: how far each pitcher stood above the league he pitched in. Three seasons anchor it. Gibson's 1968 is the famous one. Dutch Leonard's 1914 with the Boston Red Sox is often cited as the lowest of the modern era. Pedro Martinez's 2000 is the one that happened in the highest-scoring environment of the three, and therefore, by one measure, is the equal of any.

What the number measures

Earned run average is earned runs allowed, multiplied by nine, divided by innings pitched. A pitcher who allows 38 earned runs in 304 2/3 innings, as Gibson did, has an ERA of 38 times 9 divided by 304.67, which is 1.12. Innings are counted in thirds, so 304.1 means 304 and one third. The statistic exists to isolate what the pitcher is responsible for by removing runs that scored because of fielding errors and certain other mistakes, and it is normally quoted only for pitchers with enough innings to be meaningful. The usual cutoff for a league title is one inning per team game; this article uses a flat 150 innings so that the lists are consistent.

The number has two built-in limits. It depends on the scoring environment, because 1.12 in a league that scores 3.4 runs per team game is a different achievement than 1.12 would be in a league that scores five. And it depends on the official scorer's decisions about what counts as an error. Both limits recur through this article.

A note on the data. The historical record through 2025 includes Negro Leagues seasons from 1920 to 1948 as major-league data, and it also carries a few earlier Negro-leagues rows. Short, partial schedules in those leagues produce ERAs on small samples, so I have flagged them where they appear rather than ranking them in the main lists.

The lowest seasons on record

Counting every pitcher with at least 150 innings since 1893, when the mound was moved back to its present distance, these are the lowest earned run averages in the major leagues in the database.

One earlier row would top this list. Dick Redding's 1917 appears in the historical record at 0.82 over 153 2/3 innings, but it comes from a short, partial schedule and I have left it out of the ranking.

Nine of the ten came before 1920, in the deadball era. The exception is Gibson. Since 1920 the next-lowest figures are well above his.

The Negro Leagues seasons of Slim Jones in 1934 (1.60 over 203 innings) and Bullet Rogan in 1925 (1.74 over 155 1/3) also fit in the range, with the same caveat about sample size and schedule that applies to the whole of those leagues' statistics. Gibson's 1.12 sits 41 hundredths of a run below the next live-ball season on the list, a gap with no parallel anywhere else in the table.

1968: the year the sport changed a rule

To understand Gibson you have to understand 1968, which the historical season sheet describes in plain figures. The 20 major-league teams scored 3.42 runs per team game, the American League batted .230 as a whole and the National League .243, and the American League batting champion, Carl Yastrzemski, won the title at .301, a mark usually cited as the lowest winning average in major-league history. Those are numbers from a game tilted toward the pitcher.

Within that setting Gibson's year stands out. He started 34 games, completed 28 of them, threw 13 shutouts, allowed 198 hits and 62 walks in 304 2/3 innings, struck out 268 and gave up only 11 home runs. With the National League's earned run average at 2.99, he allowed about 37 percent as many earned runs per nine innings as the average team. He won the National League Most Valuable Player award and the Cy Young Award, and the Cardinals won the pennant before losing the World Series to Detroit in seven games.

Gibson did not win the most games. Detroit's Denny McLain won 31 for the Tigers, the last 30-win season in the database, and he took the American League Most Valuable Player and Cy Young awards. Luis Tiant had a 1.60 earned run average in the American League, and Sam McDowell, Wilbur Wood and the others fell below 2.00 as well. The whole season favoured pitchers, which is why the sport responded with rule changes.

After 1968 the rules were changed. The mound was lowered from 15 inches to 10, and the strike zone was reduced, so that it ran from the armpits to the top of the knees rather than from the shoulders to the knees. Expansion brought four new teams in 1969, and runs per team game rose from about 3.4 in each league in 1968 to about 4.1 the next year. The extreme end of the scale did not return, and nobody in the live-ball era has come within four-tenths of a run of Gibson's number.

1914: Dutch Leonard and the deadball floor

Dutch Leonard pitched for Boston in 1914, and the historical record lists a line of 19 wins and 5 losses in 36 games, 25 of them starts, with 17 complete games and 7 shutouts. In 224 2/3 innings he allowed 139 hits and 60 walks, struck out 176 and gave up just 3 home runs, and only 24 earned runs. That is an earned run average of 0.96.

A note on the number. Records from the deadball period rest on scorers' reports that were less uniform than modern ones, so a few hundredths in either direction should not be over-read. The story is the same either way: a pitcher whose earned run average was about one run per nine innings, the lowest in the major-league list above.

The American League earned run average in 1914 was 2.73, so Leonard allowed about 35 percent of what the average pitcher allowed. The 1914 environment is also worth describing. The major leagues scored 3.86 runs per team game that year, hit only 0.19 home runs per team game (about one every five games) and struck out 3.97 batters per team game. With the ball rarely leaving the park, offence depended on singles, walks and baserunning. Here is a number for the era that the historical record shows clearly: in 1914 the American League allowed 3.69 runs per nine innings but only 2.73 earned, which means more than a quarter of all runs scored were unearned. Fielders made many errors, and so earned run averages were lower than a modern reader expects relative to the runs that actually crossed the plate.

2000: Pedro Martinez in the run-scoring peak

The third anchor is the season that looks least like the others. In 2000, the major leagues scored 5.14 runs per team game and hit 1.17 home runs per team game, the figures of a hitter's era. The American League earned run average, computed from the team totals, was 4.91. Pedro Martinez of the Red Sox pitched 217 innings, started 29 games, completed 7 and threw 4 shutouts, going 18-6. He allowed 128 hits and 32 walks, which works out to a walks-plus-hits-per-inning figure of about 0.74, gave up 17 home runs, struck out 284 and allowed 42 earned runs. His earned run average was 1.74.

The size of the gap is the point. The next-best earned run average in the American League among qualifiers was Roger Clemens' 3.70, nearly two full runs per nine innings higher. Pedro won the American League Cy Young Award. In 2000 the National League's Cy Young Award went to Randy Johnson, who struck out 347 and had a 2.64 earned run average. The runner-up in his own league was not close, and the National League's earned run average leader, Kevin Brown, finished at 2.58, more than eight-tenths of a run higher.

Pedro's 1.74 is the highest raw number of the three, against 1.12 and 0.96. Measured by distance from the league, it is among the best ever.

Comparing across eras: the league-relative table

To compare pitchers across different run environments, divide the league earned run average by the pitcher's. A ratio of 2.0 means the pitcher allowed half as many earned runs per nine innings as the league average. Here are the best major-league seasons by that measure, again with 150 innings required. The league figure is computed from team totals in the database and includes the pitcher himself, so the true gap against everyone else is slightly wider.

By this measure Leonard and Pedro are close to a tie at the top, Maddux's strike-shortened 1994 is third, and Gibson, with the lowest ERA of the live-ball era, is fourth. A few Negro Leagues seasons in the database, such as Bullet Rogan's 1925, post even higher ratios, but on short schedules and small samples. The ranking does not make Gibson less great. It says that his 1.12 was produced in a league where average pitchers were already at 2.99, so the room for dominance was smaller than in 2000, when the league ran at 4.91.

Maddux's 1994 season needs one more note: it came in the strike-shortened year, 202 innings in 16-6, so the sample is smaller than a normal full season. The figure is real, but a full-length season gives a pitcher more chances for a bad stretch.

Why some eras produce low earned run averages

Four forces push a league's earned run average down, and the lowest seasons show them in different mixes.

The ball and the home run. In 1914, only about one team game in five produced a home run. Without the long ball, a pitcher could allow many hits and still prevent big innings. By 2000 the figure was 1.17 per team game.

The mound and the strike zone. The 1968 conditions came from a mound that stood 15 inches high and a strike zone that ran from the shoulders to the knees, which together gave pitchers a steep downhill angle and a large target. The 1969 changes reversed both.

Scorers and gloves. A deadball infield was rough, gloves were small, and errors were frequent. That inflates runs allowed but not earned runs, which is how 1914 produced a league earned run average of 2.73 on a runs-per-nine figure of 3.69. By 1968 the unearned share in the American League was down to about 13 percent, and in 2000 about 8 percent.

Workload. Leonard threw 224 2/3 innings, Gibson 304 2/3 and Johnson 346 in 1913. Starters of earlier eras worked far more innings than starters today, which makes a low rate harder to sustain. For a reader interested in how workloads have changed, the disappearing complete game tells the story.

What a low ERA does not tell you

An earned run average leaves things out. It does not say how many unearned runs the pitcher allowed, which in the deadball years could be a lot. It does not adjust for the park, and the historical record has no park factors. It rewards a pitcher for strikeouts only indirectly, as fewer balls in play means fewer chances for hits and errors. And it is a rate, not a total: Gibson's 304 2/3 innings and Leonard's 224 2/3 are different loads.

A word about the strikeout trend. Gibson struck out 268 in 1968, when the major leagues averaged 5.89 strikeouts per team game, and Pedro struck out 284 in 2000, when they averaged 6.45. Leonard's 176 came in a league where the average was 3.97. The progression is covered in the evolution of the strikeout, and it matters here because strikeouts take the fielders and the scorer out of the picture.

Common questions

Is Bob Gibson's 1.12 the lowest ERA ever? It is the lowest since 1920 in the historical record for pitchers with at least 150 innings. Leonard's 1914 (0.96 in the historical record) and Brown's 1906 (1.04) were lower, both in the deadball era.

Who has the best ERA relative to his league? By the ratio used in this article, Dutch Leonard in 1914 narrowly leads Pedro Martinez in 2000, with Greg Maddux in 1994 third among major-league seasons.

Why was 1968 called the Year of the Pitcher? Runs fell to 3.42 per team game, the American League hit .230 as a group and Yastrzemski won the batting title at .301. The rules were changed afterward.

Did anyone else have a sub-2.00 ERA in 2000? Among pitchers with 150 innings, only Pedro, at 1.74. The next-best marks were 2.58 in the National League and 3.70 in the American League.

Does a low ERA mean the pitcher was the best that year? Not by itself. Denny McLain won 31 games in 1968 with a higher earned run average than Gibson's, and voters gave both of them the awards for their leagues. Wins, strikeouts, innings and team context all matter.

Where to go next

The batting side of the story is told in the all-time batting average leaders, and the base-running side in the stolen base leaders and the running game. If you want to see how the best pitching performances end up in the scorebook, start with how to read a baseball box score and no-hitters and perfect games. To work out an earned run average yourself, use the stat calculator, and to see the current leaders, the leaders page.