The AlmanacA Longitudinal Study · Est. Retrosheet 1911–2025

The Shape
of the Game

One hundred and fifteen years of baseball, reconstructed at the grain of a single event — every plate appearance, every base taken — to watch the sport quietly rebuild itself.

15,800,829

Retrosheet events

1911–2025

seasons, unbroken

base / out states

estimates — all counted

ACT I

The run-expectancy genome

Twenty-four numbers encode a whole era’s scoring structure: the runs a team can expect from every base-and-out situation. Watch the matrix breathe as the run environment swells, collapses, and swells again.

The Dead-Ball Era

1911

Ballparks are cavernous, the ball is soft and often the only one used all game. Runs are manufactured a base at a time.

0.51R/inn

bases empty, 0 out

9.0R/G

runs per game · both teams

2.57R

bases loaded, 0 out

drag or scroll

ACT II

The disappearing ball in play

A strikeout, a walk, a home run — the “three true outcomes” never involve a fielder. In 1912 they settled one plate appearance in five. Today it is one in three, and the defended, batted, run-down baseball is vanishing from the sport.

Three true outcomes and balls put in play as a share of plate appearances, by season
Season	TTO %	Ball in play %
1911	19.2	79.9
1912	19.1	80.1
1913	19.1	80.2
1914	19.0	80.3
1915	18.8	80.4
1916	18.2	81.1
1917	17.1	82.2
1918	16.3	83.1
1919	15.8	83.6
1920	15.7	83.7
1921	15.7	83.7
1922	15.9	83.5
1923	16.0	83.4
1924	16.1	83.4
1925	16.1	83.4
1926	16.1	83.4
1927	16.3	83.2
1928	16.6	83.0
1929	17.1	82.4
1930	17.4	82.2
1931	17.5	82.1
1932	17.2	82.5
1933	17.5	82.1
1934	17.7	81.9
1935	18.3	81.3
1936	18.7	80.9
1937	19.2	80.4
1938	19.5	80.2
1939	19.5	80.2
1940	19.6	80.0
1941	19.5	80.1
1942	19.2	80.4
1943	18.6	81.0
1944	18.4	81.2
1945	19.1	80.6
1946	20.1	79.6
1947	21.0	78.7
1948	21.2	78.4
1949	21.8	77.8
1950	21.8	77.7
1951	22.0	77.5
1952	21.9	77.6
1953	22.1	77.4
1954	22.5	77.0
1955	23.2	76.3
1956	23.6	75.9
1957	23.8	75.6
1958	23.9	75.5
1959	24.3	75.1
1960	24.7	74.8
1961	25.0	74.4
1962	25.3	74.1
1963	25.5	74.0
1964	25.7	73.7
1965	25.7	73.7
1966	25.7	73.7
1967	25.4	74.0
1968	25.8	73.6
1969	26.0	73.4
1970	25.9	73.6
1971	25.4	74.0
1972	24.8	74.7
1973	24.4	75.1
1974	24.0	75.5
1975	23.3	76.2
1976	23.5	76.0
1977	23.3	76.3
1978	23.4	76.1
1979	22.9	76.6
1980	22.8	76.8
1981	22.9	76.7
1982	23.4	76.2
1983	24.0	75.6
1984	24.4	75.1
1985	25.3	74.3
1986	26.2	73.3
1987	26.2	73.3
1988	25.7	73.8
1989	25.2	74.3
1990	25.6	73.9
1991	25.6	73.9
1992	25.7	73.7
1993	26.2	73.1
1994	27.2	72.1
1995	28.0	71.3
1996	28.4	70.8
1997	28.5	70.7
1998	28.6	70.5
1999	28.8	70.4
2000	28.9	70.2
2001	28.7	70.4
2002	28.2	70.8
2003	28.1	70.9
2004	27.8	71.2
2005	27.9	71.1
2006	27.9	71.1
2007	28.4	70.6
2008	28.9	70.2
2009	29.3	69.8
2010	29.4	69.7
2011	29.7	69.4
2012	30.0	69.2
2013	30.4	68.8
2014	30.5	68.7
2015	31.1	68.0
2016	32.2	66.9
2017	33.2	65.8
2018	34.2	64.8
2019	35.1	63.8
2020	35.5	63.3
2021	35.0	63.9
2022	34.4	64.5
2023	33.9	64.9
2024	34.0	64.9
2025	33.8	65.1

Three true outcomes — SO + BB + HR

Ball put in play

ACT III

The wall of outcomes

Every way a plate appearance can end, charted as its own century-long line. Red marks the years an outcome ran above its own historical average; graphite, below. Drag the year to read any season across all eleven at once.

2025rate per 100 plate appearances

Strikeout22.42%

above its 115-yr avg of 13.50%

Walk8.32%

below its 115-yr avg of 8.50%

Home run3.04%

above its 115-yr avg of 1.96%

Single14.22%

below its 115-yr avg of 16.50%

Double4.23%

above its 115-yr avg of 4.03%

Triple0.36%

below its 115-yr avg of 0.74%

Hit by pitch1.08%

above its 115-yr avg of 0.63%

Ball-in-play out45.32%

below its 115-yr avg of 52.30%

Reached on error0.55%

below its 115-yr avg of 1.33%

Three true outcomes33.77%

above its 115-yr avg of 23.96%

Ball in play65.14%

below its 115-yr avg of 75.41%

ACT IV

The stolen base: death & rebirth

No play has been buried and resurrected as often. The dead-ball running game gave way to the home run, sat near-dormant through the 1950s, roared back with Rickey Henderson’s generation — then, in 2023, a rule change brought it back a third time, on purpose.

Stolen bases and caught stealing per game, by season
Season	Stolen bases / game	Caught stealing / game
1911	2.76	1.95
1912	2.73	1.96
1913	2.64	2.01
1914	2.43	1.92
1915	2.28	1.86
1916	2.09	1.68
1917	2.06	1.62
1918	1.97	1.50
1919	1.79	1.43
1920	1.53	1.29
1921	1.29	1.13
1922	1.24	1.05
1923	1.24	1.03
1924	1.22	1.02
1925	1.14	0.97
1926	1.11	0.94
1927	1.08	0.89
1928	1.10	0.86
1929	1.01	0.77
1930	0.96	0.74
1931	0.86	0.70
1932	0.80	0.69
1933	0.76	0.63
1934	0.73	0.60
1935	0.76	0.60
1936	0.78	0.64
1937	0.79	0.62
1938	0.79	0.60
1939	0.77	0.57
1940	0.75	0.57
1941	0.75	0.60
1942	0.75	0.61
1943	0.75	0.59
1944	0.76	0.58
1945	0.75	0.56
1946	0.72	0.57
1947	0.67	0.55
1948	0.63	0.52
1949	0.60	0.49
1950	0.61	0.49
1951	0.62	0.50
1952	0.62	0.50
1953	0.57	0.48
1954	0.55	0.47
1955	0.56	0.45
1956	0.59	0.45
1957	0.60	0.45
1958	0.64	0.46
1959	0.68	0.46
1960	0.72	0.46
1961	0.77	0.46
1962	0.78	0.45
1963	0.77	0.46
1964	0.79	0.47
1965	0.84	0.50
1966	0.88	0.55
1967	0.89	0.57
1968	0.91	0.58
1969	0.95	0.57
1970	0.95	0.56
1971	0.95	0.56
1972	0.98	0.58
1973	1.10	0.64
1974	1.21	0.68
1975	1.39	0.74
1976	1.44	0.78
1977	1.48	0.80
1978	1.43	0.79
1979	1.47	0.77
1980	1.48	0.77
1981	1.50	0.77
1982	1.51	0.77
1983	1.51	0.75
1984	1.50	0.72
1985	1.50	0.72
1986	1.58	0.73
1987	1.62	0.72
1988	1.59	0.69
1989	1.54	0.69
1990	1.52	0.71
1991	1.54	0.74
1992	1.50	0.74
1993	1.47	0.71
1994	1.44	0.67
1995	1.43	0.62
1996	1.45	0.63
1997	1.41	0.63
1998	1.41	0.64
1999	1.32	0.60
2000	1.30	0.58
2001	1.20	0.55
2002	1.15	0.52
2003	1.09	0.48
2004	1.06	0.45
2005	1.09	0.45
2006	1.13	0.44
2007	1.17	0.43
2008	1.19	0.43
2009	1.20	0.45
2010	1.26	0.48
2011	1.30	0.48
2012	1.26	0.47
2013	1.19	0.43
2014	1.09	0.42
2015	1.07	0.42
2016	1.04	0.41
2017	1.03	0.40
2018	1.00	0.37
2019	0.98	0.35
2020	0.94	0.32
2021	0.97	0.31
2022	1.13	0.33
2023	1.32	0.36
2024	1.44	0.38
2025	1.45	0.40

Stolen bases per game

Caught stealing per game

The 2023 experiment

Bigger bases and a limit on pickoff throws took effect in 2023. Stolen bases jumped from 1.02 to 1.44 a game and the success rate climbed past 80% — a clean natural experiment, visible as a step in the line.