The Vanishing Complete Game
One pitcher in 1980 finished as many games by himself as all thirty teams managed in all of 2024.
The Assignment
For a hundred years a starting pitcher was expected to finish what he began. The complete game was not a feat; it was the job. Then, across a single lifetime, it disappeared almost entirely. The assignment is to measure that disappearance honestly, decade by decade, and to find the one comparison that makes the scale of it impossible to wave away.
The Findings
In 1904, pitchers completed the games they started almost nine times out of ten. Of 2,496 games started that season, 2,186 ended with the same pitcher on the mound who began them, a rate of 87.6 percent. Finishing was ordinary. Leaving early was the exception that needed explaining.
The slide never reversed. 1922 was the last season in which more than half of all starts were completed, at 50.2 percent. By the 1950s the decade rate was 33.5 percent; by the 1980s, 15.6; by the 1990s, 7.4. In 2024, of 4,858 games started across the major leagues, 28 were completed. That is six tenths of one percent.
More baseball is played now, not less. There were nearly twice as many starts in 2024 as in 1904. The games did not vanish. The finishing did.
In 1980, Rick Langford of the Oakland Athletics threw 28 complete games, 22 of them in a row. In 2024, all thirty major league teams combined threw 28 complete games. One arm, one summer, matched the entire sport forty-four years later.
None of this is an accident, and none of it is laziness. Pitch counts, the five man rotation, the specialized bullpen, the velocity that shortens careers, and a front office math that treats the third time through a lineup as a danger to be managed, all of it points the same direction. The complete game did not fail. It was engineered out of the job. Whether the game is better for it is an argument worth having. That it is a different game than the one the record books were built on is not an argument at all. It is a measurement.
Methodology
What was counted. A complete game is a game in which one pitcher pitched the entire game for his team. Season figures are the combined totals of both major leagues. Era figures labeled as decades (1950s, 1980s, 1990s) are the complete-game rate across that full decade; single years (1904, 1922, 2024) are that year alone.
What was excluded. Minor league, college, and postseason games are not included. Shortened seasons are noted where relevant but not used as headline comparisons. The 1980 figure is a single pitcher's season total, placed beside a league-wide season total only to show scale, not to equate the two as like measures.
Sources. Complete-game totals, rates, and the 1904 and 2024 start counts are from the figures compiled in source 1, which draws on the season records in sources 2 and 3. Rick Langford's 1980 total is from source 3.
Sources
- "Complete game," Wikipedia, complete-game rates and season totals including 1904 (2,186 of 2,496, 87.6 percent), 1922 (50.2 percent), 1950s (33.5 percent), 1980s (15.6 percent), 1990s (7.4 percent), 2023 (35), and 2024 (28 of 4,858, 0.6 percent). en.wikipedia.org/wiki/Complete_game
- Baseball Reference, Major League season pitching totals. baseball-reference.com
- Baseball Almanac, Complete Games Year-by-Year Leaders, including Rick Langford, Oakland Athletics, 28 complete games in 1980. baseball-almanac.com/pitching/picomg4.shtml
Parker Intel arrives weekly. One tough assignment, settled with maps, charts, and cited sources.
Subscribe to the Features Desk, $5 a week