Baseball Toaster was unplugged on February 4, 2009.
When we last left off, we established that teams today retain leads better than any decade since the deadball era. However, the reason for this, apparently, is that teams establish bigger leads today since their success in holding smaller leads (one-, two-, or three-run leads) has changed very little over the last fifty years or so.
However, I didn't offer any direct proof of this assumption. But the table below bears it out. It breaks down the game situations at the end of the sixth inning for all games:
Decade | % Tied | % 1-R Leads | % 2-R Leads | % 3-R Leads | % >3-R Leads |
1900s | 14.91% | 23.35% | 18.41% | 14.44% | 28.89% |
1910s | 14.91% | 24.37% | 18.60% | 14.89% | 27.22% |
1920s | 12.76% | 23.19% | 17.93% | 13.81% | 32.31% |
1930s | 12.97% | 22.04% | 18.74% | 14.09% | 32.16% |
1940s | 14.17% | 23.58% | 18.55% | 14.23% | 29.47% |
1950s | 14.06% | 23.57% | 18.77% | 13.87% | 29.73% |
1960s | 14.81% | 24.84% | 19.21% | 14.38% | 26.77% |
1970s | 13.87% | 24.59% | 19.96% | 14.42% | 27.16% |
1980s | 13.44% | 24.21% | 19.07% | 14.69% | 28.59% |
1990s | 12.85% | 23.14% | 18.87% | 14.98% | 30.16% |
2000s | 12.40% | 22.24% | 18.76% | 13.85% | 32.76% |
Overall | 13.60% | 23.62% | 18.96% | 14.38% | 29.45% |
Note that tie ballgames one-run leads are now at an all-time, and two- and three-run leads aren't that far behind. At the same time, leads of more than three runs are at an all-time high.
Therefore, teams have more leads than ever to hold, but they are bigger leads, and hence are easier to keep.
However, one has to keep in mind that teams are using more pitchers per game than ever. And as the table below demonstrates, innings pitched per start have been dropping steadily, with fits and starts, since the dawn of recorded baseball time (note, based on pure starter stats):
Deacde | IP Per GS | Change |
1870s | 8.83 | |
1880s | 8.54 | -3% |
1890s | 7.91 | -7% |
1900s | 8.08 | 2% |
1910s | 7.00 | -13% |
1920s | 7.28 | 4% |
1930s | 7.47 | 3% |
1940s | 7.34 | -2% |
1950s | 6.88 | -6% |
1960s | 6.96 | 1% |
1970s | 6.97 | 0% |
1980s | 6.58 | -6% |
1990s | 6.32 | -4% |
2000s | 6.13 | -3% |
So maybe I am a bit hard on modern bullpens. Yes, they haven't helped teams today retain close leads any better than their counterparts did in the past. But more and more the leads are being held by members of the bullpen rather than starters. Being on a par with starters is actually quite an accomplishment especially when the guys holding the lead in, say, the sixth and seventh might be the fourth or fifth guy in the bullpen.
What happens as the game gets later and, one would assume, the team with the lead is more likely to use its closer? Let's look at the results for leads after seven and eight innings to see whether teams are improving on retaining leads as the game progresses.
Decade | % Leads Retained after 7 | % 1-R Lead Retained | % 2-R Lead Retained | % 3-R Lead Retained | % >3-R Lead Retained |
1900s | 86.12% | 68.37% | 80.84% | 91.50% | 98.09% |
1910s | 84.61% | 65.88% | 80.46% | 91.30% | 97.88% |
1920s | 83.99% | 62.06% | 79.56% | 87.58% | 97.00% |
1930s | 83.34% | 59.49% | 78.07% | 88.71% | 96.96% |
1940s | 83.65% | 61.47% | 80.10% | 88.64% | 97.79% |
1950s | 83.91% | 61.81% | 78.35% | 91.62% | 97.75% |
1960s | 83.96% | 63.60% | 81.21% | 91.58% | 97.92% |
1970s | 83.99% | 63.69% | 81.53% | 90.37% | 97.41% |
1980s | 84.42% | 62.38% | 81.95% | 91.29% | 97.93% |
1990s | 84.67% | 62.58% | 81.00% | 90.33% | 98.00% |
2000s | 85.49% | 62.26% | 82.19% | 91.44% | 97.87% |
Total | 84.30% | 62.68% | 80.75% | 90.51% | 97.71% |
Again, teams today retain leads after seven better than they have in a hundred years, and again their ability to hold one-run leads hasn't improved in the 2000s. It's actually very slightly below average even though it hasn't changed much in the last two and one half decades. Also, leads over three runs have been retained at the same rate since the Forties.
However, retaining two-run leads after seven is at an all-time high, and retaining three-run leads, though a much tighter group, has improved a full percentage point over the Nineties and is the highest it's been since the Fifties and Sixties. So there is some actual improvement in holding leads that starts to become clear after seven innings.
Let's see what happens after eight innings:
Decade | % Leads Retained in the 9th | % 1-R Leads Retained | % 2-R Leads Retained | % 3-R Leads Retained | % >3-R Leads Retained |
1900s | 92.83% | 80.04% | 90.24% | 96.59% | 99.36% |
1910s | 91.90% | 79.12% | 89.33% | 96.36% | 98.98% |
1920s | 91.58% | 76.65% | 87.44% | 95.28% | 99.08% |
1930s | 91.57% | 75.70% | 88.65% | 94.76% | 98.99% |
1940s | 91.32% | 76.28% | 88.52% | 95.46% | 99.18% |
1950s | 91.83% | 77.06% | 89.60% | 96.05% | 99.30% |
1960s | 91.68% | 77.88% | 90.75% | 95.58% | 99.55% |
1970s | 91.72% | 78.94% | 89.40% | 96.06% | 99.09% |
1980s | 92.12% | 77.90% | 90.82% | 96.32% | 99.25% |
1990s | 92.37% | 77.84% | 91.00% | 96.04% | 99.24% |
2000s | 92.89% | 78.14% | 90.72% | 96.71% | 99.43% |
Total | 91.99% | 77.74% | 89.96% | 95.95% | 99.24% |
Well, there's definite improvement across the boards here, and that's thanks to the closer.
Now, which teams have been the best at retaining leads? Let's look at leads after six first:
Year | Team | % Leads Kept after 6 |
1954 | CLE | 95.00% |
1946 | NYA | 94.20% |
1942 | BRO | 93.83% |
1997 | CIN | 93.22% |
1912 | BOS | 92.11% |
1984 | DET | 91.67% |
2003 | ANA | 91.67% |
1967 | NYA | 91.38% |
1994 | FLO | 90.91% |
1938 | DET | 90.77% |
And here are the worst:
Year | Team | % Leads Kept after 6 |
1962 | NYN | 48.78% |
1928 | PHI | 53.49% |
1945 | PHA | 55.32% |
1979 | OAK | 56.52% |
1988 | ATL | 56.86% |
1979 | SDN | 56.92% |
1952 | PIT | 57.14% |
1931 | CIN | 58.18% |
1950 | PIT | 58.33% |
1957 | WS1 | 58.33% |
Now, here are the best on the ninth:
Year | Team | % Leads Kept in 9th |
1910 | SLA | 100.00% |
2003 | LAN | 100.00% |
1993 | SFN | 100.00% |
1984 | DET | 98.88% |
2004 | BOS | 98.82% |
1948 | NYA | 98.81% |
2002 | ANA | 98.73% |
1996 | NYA | 98.73% |
1912 | NYG | 98.70% |
1933 | NY1 | 98.70% |
1938 | NY1 | 98.67% |
2004 | MIN | 98.61% |
1963 | BAL | 98.61% |
1966 | CLE | 98.55% |
1988 | SFN | 98.53% |
And now the worst:
Year | Team | % Leads Kept in 9th |
1963 | NYN | 76.47% |
1978 | SEA | 77.55% |
1979 | SDN | 77.61% |
1968 | PHI | 77.94% |
1978 | NYN | 78.46% |
1958 | CHN | 79.37% |
1999 | CHN | 79.69% |
1926 | BSN | 80.00% |
1949 | CIN | 80.00% |
1936 | CIN | 80.30% |
Next are the best at retaining a closer lead (three runs or less) in the ninth:
Year | Team | % <4 R Leads Kept in 9th |
1985 | CIN | 100.00% |
2002 | ANA | 100.00% |
1910 | SLA | 100.00% |
2003 | LAN | 100.00% |
1993 | SFN | 100.00% |
1963 | CIN | 98.11% |
1966 | CLE | 97.96% |
1963 | BAL | 97.87% |
1965 | SFN | 97.83% |
1958 | BAL | 97.83% |
1933 | NY1 | 97.78% |
1996 | NYA | 97.73% |
1992 | KCA | 97.67% |
1993 | KCA | 97.62% |
1938 | NY1 | 97.62% |
The worst:
Year | Team | % <4 R Leads Kept in 9th |
1994 | OAK | 65.00% |
1963 | NYN | 66.67% |
1941 | PHA | 66.67% |
1932 | CHA | 66.67% |
1978 | SEA | 67.65% |
1929 | CHA | 67.65% |
1930 | PHI | 68.18% |
1949 | CIN | 68.75% |
1997 | OAK | 68.97% |
1993 | CIN | 68.97% |
1999 | CHN | 69.23% |
1979 | SDN | 69.57% |
Finally, here are the best at holding a one-run lead in the ninth. There were too many teams at one hundred percent, so I listed the perfect teams with the most opportunities:
Year | Team | 1-R Leads Held in 9th | % |
2002 | ANA | 18 | 100.00% |
1976 | DET | 17 | 100.00% |
2003 | LAN | 17 | 100.00% |
1981 | OAK | 16 | 100.00% |
1983 | SDN | 16 | 100.00% |
1993 | SFN | 16 | 100.00% |
1913 | CHN | 15 | 100.00% |
1930 | CIN | 15 | 100.00% |
1969 | DET | 15 | 100.00% |
1965 | SFN | 15 | 100.00% |
1970 | CAL | 14 | 100.00% |
2001 | ANA | 14 | 100.00% |
2001 | CHA | 14 | 100.00% |
1954 | CLE | 14 | 100.00% |
1945 | WS1 | 14 | 100.00% |
1983 | PIT | 14 | 100.00% |
1991 | NYN | 14 | 100.00% |
1985 | CIN | 14 | 100.00% |
1980 | KCA | 13 | 100.00% |
1935 | CIN | 13 | 100.00% |
1979 | LAN | 13 | 100.00% |
1910 | SLA | 13 | 100.00% |
1969 | CHN | 12 | 100.00% |
1912 | NY1 | 12 | 100.00% |
1980 | DET | 11 | 100.00% |
1990 | TOR | 11 | 100.00% |
1996 | CHN | 11 | 100.00% |
1967 | SFN | 11 | 100.00% |
1991 | KCA | 11 | 100.00% |
And the worst:
Year | Team | % 1-R Leads Kept in 9th |
1963 | NYN | 33.33% |
1986 | SEA | 35.71% |
1940 | BOS | 35.71% |
1961 | SLN | 37.50% |
1993 | CIN | 40.00% |
1978 | SEA | 41.18% |
1973 | TEX | 41.67% |
1943 | WS1 | 42.86% |
2001 | COL | 42.86% |
2004 | CHN | 42.86% |
Because, I think there's a big difference between the two. If you're using the latter definition, it's a bit problematic... In the 50s and 60s both teams were using starters very late into games, so if indeed the concept of the relief/closer is valid (in the sense of "maintaining leads more successfully") it would then seem likely that even if the team winning late in the game lost the lead at one point, they would be more likely to regain said once again lead because both pitchers would be tiring.
Again, really interesting stuff. Cheers!
The presence of the 1979 A's on the worst list for holding leads after 6 is some valuable context for Billy Martin's infamous handling of his starting staff in 1980.
Good point. I have to see if Billy's approach actually helped them in 1980.
Underbruin,
Sorry, I may have missed it. By leads retained I mean that the team never relinquished the lead after the sutuation in question. If it's atthe end of the sixth, the leading team could not lose the lead in either half of the seventh, eighth, or ninth innings. By "lose the lead" I mean allow the other team to tie the game or take a lead. However, if team A (the visitor) leads team B by a score of 1-0 at the end of the sixth. Then team A scores 3 runs in the top of the seventh, and team B scores 3 runs (score 4-3), etc., until team A wins 10-9. As long as team A lead by at least one run throughout it counts. By the same token, if team B is trailing 1-0 at the end of the sixth and ties the game in the bottom of the seventh only to lose 10-1, it doesn't count as a lead retained since the lost the lead in the seventh (though it could count for seventh-inning leads).
As for starters staying in games longer in the early days of baseball, to me that just represents a different strategy for retaining a lead that teams employed more often at that time. The way it was employed changed and may have affected how well some teams retained leads as relief pitching developed. I want to look at standard deviation in lead retention to see if it can be measured. But that's another study.
Fabulous work!
I would suggest breaking up your data into top/bottom half for each inning.
Tom
Comment status: comments have been closed. Baseball Toaster is now out of business.