My last article in this series, Your Typical Starting Pitcher, got long-winded, so I’ll get right to the point in this one. This chart gives the record of a typical relief pitcher in each decade since 1920:
Decade
|
Count
|
G
|
IP
|
|
W
|
L
|
WPct
|
SO
|
BB
|
|
GF
|
SV
|
|
ERA
|
1920s
|
4
|
57
|
142.0
|
|
10
|
7
|
.579
|
|
65
|
55
|
|
37
|
15
|
|
3.11
|
1930s
|
8
|
53
|
114.1
|
|
9
|
6
|
.612
|
|
43
|
41
|
|
37
|
11
|
|
3.80
|
1940s
|
23
|
58
|
114.1
|
|
8
|
6
|
.575
|
|
49
|
48
|
|
37
|
11
|
|
3.23
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1950s
|
69
|
58
|
102.2
|
|
7
|
5
|
.580
|
|
56
|
38
|
|
35
|
11
|
|
3.31
|
1960s
|
256
|
60
|
97.2
|
|
6
|
5
|
.526
|
|
67
|
36
|
|
35
|
12
|
|
3.06
|
1970s
|
353
|
61
|
97.0
|
|
7
|
6
|
.543
|
|
63
|
37
|
|
37
|
13
|
|
3.17
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1980s
|
509
|
61
|
91.0
|
|
6
|
5
|
.515
|
|
65
|
35
|
|
35
|
13
|
|
3.19
|
1990s
|
835
|
63
|
72.0
|
|
4
|
4
|
.509
|
|
59
|
29
|
|
28
|
10
|
|
3.60
|
2000s
|
986
|
65
|
67.1
|
|
4
|
4
|
.513
|
|
57
|
26
|
|
25
|
9
|
|
3.73
|
Notes:
1) My standard of a “regular” relief pitcher was 50 game appearances with no more than 5 starts. There is no one listed before the 1920s because there are no such pitchers before 1920.
2) All four of the qualifying relievers of the 1920s pitched for the Washington Senators.
3) The number of relievers increased exponentially each decade until the 1960s, and since the 1960s has settled in at a more modest 40 to 60% per decade growth rate.
4) Relief appearances on average have been getting shorter since the role was first defined. In the 1940s a typical reliever pitched 2 innings per outing; in the 1960s, 1.6; in the 1980s, 1.5; today, just over 1.0. We can project that, by the year 2025, a typical reliever may be pitching about 70 games and 58 innings per season.
5) Early relievers tended to have high winning percentages because they normally entered the game after the starter was knocked out—thus, when their team was behind. They were in position to be credited with a win more often than they were in position to be charged with a loss.
6) The strikeout/walk ratio of a typical reliever has gone from essentially even to a little better than 2-to-1.
7) The strikeouts per nine innings of a typical reliever have gone from a little over four to almost eight.
8) In the 1920s almost two-thirds of relief appearances ran until the game was over. Today that figure is a little over one-third, and still headed downward sharply.
9) The number of Saves per season for the average reliever has declined as the number of relievers has grown much more rapidly than the number of Saves. (The total number of saves is growing hardly at all anymore, as complete games are almost extinct.)
10) In the 1940s the average age of a major league reliever was 31.69 years, whereas the average age of a starter was 29.47. 31.69, for relievers, many of whom were old guys, minor league veterans. This dropped to 31.14 in the 50s, 30.01 in the 60s, and 28.74 in the 70s. Since then it has gone up, to 28.93 in the 80s, 29.63 in the 90s, and 30.48 in our own decade.
Relievers in the 70s used to pitch 75 games, 130 innings a year. As usage rules have evolved to limit the innings of relievers there have been more “career” relievers, which probably drives the average age up.
11) The ERA of relievers has always been lower than the ERA of starters in the same era, and lower than the league ERA, in part (but only in part) because the “earned run” rules favor a reliever. The relative ERA of relievers has been between .83 and .87 in every decade since the 1940s.
12) In the 1970s a typical reliever issued 7.96 intentional walks per season. This is now down to a less unreasonable 3.73.
13) The percentage of regular relievers who are in the Hall of Fame, by decade:
Team
|
Hall
|
1920s
|
0%
|
1930s
|
0%
|
1940s
|
0%
|
1950s
|
9%
|
1960s
|
3%
|
1970s
|
4%
|
1980s
|
2%
|
1990s
|
1%
|
2000s
|
0%
|
The 9% in the 1950s is all Wilhelm. There are only 69 relief seasons there, so six seasons of Wilhelm is 9%.
14) I figured the relative strikeout rate of regular relievers in each decade. Remember, the first two groups are so small that the data is meaningless. Since then, the strikeout rate has been high—and trending higher.
Team
|
Rel K
|
1920s
|
1.50
|
1930s
|
.97
|
1940s
|
1.06
|
1950s
|
1.07
|
1960s
|
1.06
|
1970s
|
1.12
|
1980s
|
1.19
|
1990s
|
1.19
|
2000s
|
1.16
|
This really is one of the largest engines of change in baseball—the realization that many pitchers who have sub-marginal stuff in extended outings can throw hard enough, in short outings, to be effective. A lot of guys who strike out five men per nine innings as starters strike out seven to eight as relievers—and there are lots and lots of pitchers who are just not quite good enough.
A reliever has an ERA advantage over a starter based on the fact that the run is charged to whoever puts the runner on base, which means that if a starter and a reliever share the inning, the run is normally charged to the starter. But this ERA advantage only amounts to about .20 per nine innings. Even if you adjust for that, the ERA of relievers has always been lower than the ERA of starters.
With that in mind, perhaps it is not surprising that the number of relievers used keeps growing. If you get a 4.20 ERA from your starters and a 3.20 ERA from your relievers, who do you want more of and who do you want less of? History is logical, in a sense. As long as we continue to see meaningfully lower ERAs from relief pitchers than from starters, perhaps we should expect to see more and more relievers.