In the last two weeks I have been bombarded by readers asking how to divide baseball history into eras. I answered the first of these questions honestly: I said I had no idea. I have a better answer for you now, and we’ll get to that in due time.
I first became aware of the concept of dividing baseball history into eras with the publication of the Neft/Cohen Baseball Encyclopedia. The Sports Encyclopedia: Baseball (1974) was a marvelous book in its day. It largely ignored the 19^{th} century—a wise choice—and segmented "modern" baseball history into four eras—the Dead Ball Era (19011919), the Lively Ball Era (19201945), the PostWar Era (19461960), and the Expansion Era (19611973). Then it presented, as I recall, a twopage statistical spread about each league each season, with a writeup about the year and some interludes summarizing player’s careers. It was a seminal work.
Those era breakdowns worked until about 1980, but after about 1980 they didn’t really work anymore. The world decided that 19^{th} century baseball was major league baseball, ignoring all of the obvious indicators to the contrary, and the "postexpansion era" would now stretch to more than half a century if we were still in that. The commissioner’s office decided to promote the idea that the arrival of Jackie Robinson was the biggest event in the history of the game. At some point we didn’t know where to draw the lines any more.
OK, how do we figure this out?
The same way I figure anything else out: I create a mathematical image of the problem. There are many, many events which can be used to draw dividing lines across baseball history, some of them large and some of them small. These events include:
1) Rule Changes. The Designated Hitter Rule makes the game after 1973 different from the game before 1973.
2) Structural Changes. The split into divisions in 1969 makes baseball post1969 different from baseball pre1969. Wild Cards and Interleague play are meaningful dividing lines in history.
3) Changes in the conditions of the game. Night baseball is a very important dividing line. Integration (Jackie Robinson) is an important dividing line. The banning of steroids is an important dividing line, and the arrival of steroids is a dividing event, although it is hard to find the start of it.
4) Changes in how the game is played. The 5man pitching rotation (replacing the 4man rotation) is a dividing line. The development of the modern closer is a dividing line.
5) Radical changes in statistical standards are dividing lines. The 19631968 era, when the pitchers dominated, is sort of a selfcontained era, although this also can be marked off by rules changes. If I say "the stolen base era" you know what I mean by that—the Rickey Henderson/Vince Coleman/Tim Raines era.
6) Franchise moves are dividing lines. The move of the Dodgers and Giants to the coast (1958) is an important dividing line.
7) New ballparks are dividing lines, to an extent. The arrival of the Sterile Ashtray Stadiums (about 1970), the arrival of artificial turf, of domed stadiums. . .these things are dividing lines. The building of Camden Yards was a significant dividing line, because Camden Yards showed what modern ballparks could be with intelligence and imagination.
8) External events are dividing lines. World War II is a dividing line. 911 may be a dividing line (although I forgot to include it in my study, which I will report to you in a few minutes here.)
9) Players’ careers are dividing lines. We speak of the 1920s as the Babe Ruth era, of the 1940s as the Joe DiMaggio era in New York, or the Ted Williams era in Boston, or the Stan Musial era in St. Louis. We speak of the 1950s as the Mickey Mantle/Willie Mays era.
10) Commissioners are dividing lines.
And there are probably others, which I may get to here or may not. Altogether I found 366 dividing lines or dividing events in baseball history—or more than that, depending on how you state things. I put the arrival of Ted Williams (1939) and the departure of Ted Williams (1960) on the same line, so I count them as one event, but that’s actually two separate events.
Some of those events are more significant than others, obviously. I valued events at one through 10 points each, with the biggest and most obvious dividing lines counting as 10point events. Most of the 366+ events were onepoint events, and I had, altogether, 737 points worth of dividing lines separating 1871 from 2012.
Let’s use the 1930s to illustrate the process. The AllStar Games begin in 1933. That’s a dividing line; the "AllStar Era" begins in 1933. I counted that as a 2point dividing line:
Every year after the AllStar games begin is different from every year before they began, so those 2 points carry forward to all future years, including 2012; those are 2 of the 737 points that separate 1871 from 2012.
The modern MVP Awards begin in 1931. That’s a dividing line; I counted that as a 1point event, making every year after 1931 "in" the MVP era, and every year before 1931 out of the era:
The Hall of Fame opened in 1939; I counted that, again, as a twopoint event:
The first night baseball game was played in 1935. Night baseball is a big deal in baseball history, a much more real and meaningful separator than the awards, but on the other hand, they didn’t suddenly start playing a whole lot of night baseball in 1935. They started playing a whole lot of night baseball during World War II. I counted "night baseball" as a 7point event, but put two of those points in 1935, when the first night game was played in the majors, and the other five in 1943. I’ll add a totals line here:
1940 is separated from 1930, so far, by a total of 7 points. Bob Feller and Joe DiMaggio both reached the major leagues in 1936, Ted Williams in 1939:
I started with a list of the 25 biggest superstars in baseball history, found a few other guys that I couldn’t keep off the list, and wound up with I think 31 superstars whose arrival is a onepoint dividing line in baseball history, and whose departure is another onepoint dividing line. Everybody was one point, but I treated Babe Ruth different from anybody else, in that I also gave one point to Ruth’s arrival in New York (1920) and one point to his departure from New York (1935), as well as a point for his departure from the game:
Eddie Collins and Pete Alexander also retired after the 1930 season and Rogers Hornsby played his last game in 1937 and Lou Gehrig in 1939, so there were also points for that, but. . .you understand the process, right? If you add in those four markers 1930 is at 6 and 1940 is a 22, a 16point separation between those two years. We add one point for the turning of the calendar in 1940 (three points for the turning of the calendar in 1900 and 2000). John McGraw’s retirement is a point.
There was only one ballpark added to the major leagues in 1930, Municipal Stadium in Cleveland (and it was only used for weekend games), but a new ballpark is a onepoint event.
I marked off the decline in complete games in this way: that on the line for "declining complete game percentage" I started at zero in 1871 and added 1 for each 5% decline in the percentage of games that were completed. In 1931 48% of major league games were completed by the starting pitcher, so the score on that line is "10", meaning that we have crossed ten thresholds dividing baseball from the time when the starter almost always finished his work (starter completes less than 95% of games—1, starter completes less than 90% of games—2, starter completes less than 85% of games—3, etc.) In 1934 the percentage of complete games dropped to 43%, so that’s an 11^{th} marker in that column. There is a huge change in how the game is played; we have to construct a system that takes note of changes in how the game is played. This is one of the ways in which 2012 is very different from 1975.
One of the biggest changes in baseball in the 1930s was the organization of the minor leagues into farm systems. That happened between 1920 and 1940, mostly. It’s a huge dividing line in baseball history—a 10point divider, one of the big ones—but it did not happen all at once, so I counted that as 1 point in 1920, 1 point in 1922, 1 point in 1924, 1 point in 1926, etc., until the 10 points were all in in 1938.
The offensive norms changed. In 1931 (as best I can place it) the leagueleading norm for hits dropped below 240 hits; that’s a point. In 1931 the leagueleading norm for triples dropped under 20. In 1933 the leagueleading norm for batting average dropped under .380. In 1934 the leagueleading norm for batter’s strikeouts went over 100. In 1937 (as best I can place it) the leagueleading norm for RBI dropped under 160. In 1939 the leagueleading norm for runs scored dropped under 140, and the leagueleading norm for hits dropped under 220. (It hasn’t changed since 1939, by the way. It’s been in the same "bracket"—200 hits to 220—since 1939, so there has been no point in that category since 1939.) In 1939 the leagueleading norm for doubles dropped under 50. All of these things are separation points. (The leagueleading norm for doubles dropped under 40 in 1970, but went back over 40 in 1973. The norm for a leagueleading doubles total went over 50 in 1997, went back below 50 in 2009.)
If you add up all of the points separating 1930 from 1940, you have a total of 33 separation points. This is a very, very low total for a decade, one of the lowest ever. These are the point totals by decade since 1871:
1871

0







1880

47


1871

to

1880

47

points

1890

123


1880

to

1890

76

points









1900

186


1890

to

1900

63

points

1910

253


1900

to

1910

67

points

1920

311


1910

to

1920

58

points

1930

357


1920

to

1930

46

points

1940

390


1930

to

1940

33

points









1950

432


1940

to

1950

42

points

1960

483


1950

to

1960

51

points

1970

569


1960

to

1970

86

points

1980

614


1970

to

1980

45

points

1990

646


1980

to

1990

32

points









2000

696


1990

to

2000

50

points

2010

729


2000

to

2010

33

points

The 1930s miss by only one point, in my system, of being the least "active" decade in baseball history, in terms of changes in the game. No separating line can reasonably be drawn through the 1930s, because there’s just not much happening here—no rules changes, franchise shifts, strategic revolutions. There’s the invention of the three major awards, the comings and goings of a good number of superstars, the beginning of night baseball, some changes in the statistical norms of the game, the gradual development of the farm systems, and one new ballpark; that’s it.
The major dividing lines in baseball history, as I scored them, are:
The reformation of the National Association into the National League, 1876 (10 points),
The gradual introduction of fielders gloves in the 1870s and 1880s (6 points),
The switch of pitchers to pitching overhand from underhand, 18821883 (10 points), (The changes in the ball and strike limits in this era are also extremely important, but since these rules were changed several times, there are several smaller point values for them, rather than one large immediate point total.)
The collapse of the American Association, 1892 (6 points),
The change to the modern pitching distance, 1893 (10 points),
The contraction of the National League from 12 teams to 8, 1900 (6 points),
The formation of the American League, 1901 (8 points),
The beginning of the World Series, 1903 (8 points),
The emergence of modern stadiums, dated 1908 (7 points),
The banning of the corrupt players, 1920 (5 points),
The banning of the spitball, 1920 (6 points),
The arrival of the Commissioner system, 1921 (7 points),
The Federal League ruling, 1921 (4 points),
The beginning of baseball on radio, 1921 (5 points),
The widespread use of night baseball, 1943 (5 points),
The breaking of the color line, 1947 (7 points),
The switch from travel by train to travel by airplane, 19521960 (5 points),
The required use of batting helmets, 1956 (6 points, which seems now like about 3 points too many),
The move of the Giants and Dodgers to the coast, 1958 (6 points),
The first expansion and the switch from a 154 to a 162game schedule, 19611962 (a total of 17 points for various things—2 points for each new team, 3 points for the expansion itself, 2 points for the move of the Twins to Minnesota, 2 points for the extra eight games on the schedule, 1 point for the new stadium in Washington.)
The redefinition of the strike zone, 1963 (5 points),
The institution of the amateur draft, 1965 (4 points),
The second expansion in 1969 (8 points, 2 for each new team),
The split into divisions in 1969 (5 points),
The lowering of the mounds/restoration of the strike zone in 1969 (5 points),
The DH Rule, 1973 (5 points),
The Seitz decision granting free agency, 1977 (7 points),
The switch from 4man to 5man starting rotations, 19761986 (6 points, gradual),
The development of modern closers, 19781984 (6 points, gradual),
The banning of steroids, 2005 (5 points).
The franchise shifts of the years 19531972 are also huge dividing lines, if taken aggregately, as are the aggregate changes in statistical standards which we account for one point at a time. The other expansions are all 4 points at a time (2 points per team), although with one of them there is an extra point or two for the Brewers moving to the National League. I gave 5 points every time the major league ERA moved from over 4.00 to under 4.00 (on a firm basis), or from over 3.00 to under 3.00. The major league ERA dropped under 4.00 in 2011, but we don’t know yet if that’s a firm change. If it is, that’s a 5point marker separating 2010 from 2011.
You are of course free to debate any of these choices, values or omissions as you see fit, and if you know me at all, you know I’m not going to debate them with you. I’ll also post my spreadsheet in which I did this accounting so that you can download it and study it at more length if you want to do that.
OK, now we have a time line which measures changes in the game. Our next challenge is to use that time line to divide baseball history into eras.
Well, let me jump ahead to the conclusion. As a result of doing this work, I now have a clear notion of where the dividing lines across baseball history should be drawn. The reason we do this, of course, is to enable us to think more clearly about the issue. By breaking the large, vague question (where are the dividing lines between baseball eras?) into a long series of small, specific questions which have definitively correct answers or answers requiring that we make a very small guess, we are able to create a much more focused picture of baseball’s eras. These are the lines that I will advocate:
Era 1 (The Pioneer Era), 18711892
Era 2 (The Spitball Era), 18931919
Era 3 (The Landis Era), 19201946
Era 4 (The Baby Boomers Era), 19471968
Era 5 (The Artifical Turf Era), 19691992
Era 6 (The Camden Yards Era), 19932012
You can call them something else. . .you can call them the Cap Anson Era, the Ty Cobb/Honus Wagner Era, the Babe Ruth Era, the Willie Mays/Mickey Mantle Era, the Mike Schmidt Era, the Bud Selig Era. You can call them the Brickyard era, the Sharp Spikes Era, the Golden Era, the Jackie Robinson Era, the DH/Free Agent era, and the Steroid Era. You could call the last two eras the Labor Strife Era and the Labor Peace Era. I’ll sell you the naming rights for $50.
Anyway, how did I arrive at these dividing lines, given what I have told you so far?
I tried about 40 different ways to break baseball history down into eras, and two of those ways worked better than the other 38; I’ll skip the failures.
One of the two things that worked was, I planted a post every 25 years beginning in 1880—1880, 1905, 1930, 1955, 1980, 2005. By definition, there are going to be six eras in baseball history, and each will be about 25 years in length, more or less. Then I measured the changes in the game from the posts.
For example, I planted a post in 1930—assuming that the 1930 season would be in the center of an era—and I planted a post in 1955, assuming that the 1955 season would be in the center of an era. The question then becomes, where do we draw the dividing line between those two?
We start with simple math. The "Accumulated Change Score" in 1930 is 357; in 1955 it is 461. In 1940 the ACS score is 390, which is 33 points from 1930 but 71 points from 1955. Obviously, that goes in the era with 1930, rather than the era with 1955.
1941 is 41 points from 1930, but still 64 points from 1955.
1942 is 46 points from 1930, but 58 points from 1955.
1943 is 51 points from 1930, but 53 points from 1955.
1944 is the same, 51 and 53.
1945 is 53 points from 1930, but 51 points from 1955—so we COULD draw the dividing line between 1944 and 1945.
But that doesn’t make sense, because there is only a twopoint separation between 1944 (408) and 1945 (410). We can’t draw a dividing line where there is no space.
1945 and 1946 is better; there is a fivepoint separation there based on the end of the war (3 points) and some other things. But still, 5 points is not a very large or satisfying separation.
There is an 11point separation between 1946 and 1947 based on the breaking of the color line, the inauguration of the Rookie of the Year award, and some changes in statistical standards. I chose, then to break between 1946 and 1947.
What we are doing, essentially, is asking ourselves "which is a more rational dividing line—the end of the war, or the end of segregation in baseball?" By working through this process we have narrowed our options down to those two, and it seems to me that the end of segregation in baseball is a better place to draw a line. That’s what this is. . .it’s a process of forcing ourselves to focus on the most relevant question.
The end of this era is the same thing; the 1968 season is actually a couple of points closer to 1980 than it is to 1955, but there is a 3point separation between 1967 and 1968, and a 24point separation between 1968 and 1969. Obviously, we draw the line where there is a 24point separation, rather than a 3point separation, so we end that era at 1968, and begin the new one in 1969.
The 1965 season has a surprisingly strong argument to be used as a dividing line. I did not anticipate this, since 1965 is in the middle of the 19631968 pitchers’ epoch, and I would not be very willing to divide the eras in the center of a clearly identifiable period like that. But 1965 is
a) the beginning of the artificial turf era, and
b) the beginning of the amateur draft era,
two fairly significant markers. Combining that with the election of a new Commissioner that year (Spike Eckert) and some other little stuff, 1965 makes a surprisingly good case to be considered the start of an era. But we rejected its appeal.
OK, so that system worked—the six pillars approach.
The second approach that worked was this. Suppose that we say that a new era begins whenever three standards are met:
1) At least ten years have passed since the commencement of the previous era,
2) The accumulated changes in the game since the commencement of the previous era total up to at least 100 points, and
3) There is a separation between two consecutive seasons of at least 10 points.
That method, it turns out, draws exactly the same lines as the six pillars approach. . ..fudging just a little bit on one dividing line, but essentially it draws the same lines as the six pillars approach. The three markers of sufficient change approach has two other advantages: one, that it does not start with an a priori decision that there should be six eras each lasting about 25 years, thus cleans up one of the arbitrary elements of the process, and two, that it tells us where we are at the present time.
Using the six pillars approach, we can assume that we will transition into a new era sometime between now and 2030, but when? We really don’t know where we are with respect to the era that we are in at the present time.
The three markers of sufficient change approach tells us that we are nearing the end of the Bud Selig era, nearing the end of the Camden Yards era, but that we are probably at least five years away from the next line. We are more than ten years into the era—mark one—and the accumulated changes in the game since 1993 total up to 74 points as I have scored them. I could have scored some of those differently; it is harder to get perspective on more recent events. I gave two points for the start of interleague play; maybe it should have been more. I gave no points for the AllStar game determining home field advantage in the World Series; maybe there should have been points for that. I gave 5 points for the banning of steroids in 2005; maybe it should have been 8. I gave no points for the major league ERA dropping under 4.00 in 2011, but I will if it stays under 4.00. Historical perspective requires time and distance.
But there will be changes next year. We’re realigning the divisions next year (2 points, probably), moving the Astros to the American League (1 point), plus we’re probably selecting a new Commissioner within the next year or so (2 points). We have a new set of rules this year about the signing of amateurs taken in the draft; maybe I should have given a point for those. There are going to be very important new rules in the next few years reducing the salary discrepancies between rich teams and poor teams. It’s hard to place a value on those rules until they are in place and we see what they do.
One more point, and then I’ll shut up. Each era can be divided into epochs (or each epoch could also be divided into eras. I studied dictionaries, trying to figure out whether an "era" or an "epoch" was a larger event. It’s not really clear; the definitions are fungible. I decided to use eras as larger periods and epochs as smaller.)
Anyway, each era can generally be broken down into about five epochs, some clear and some fuzzy. The war years (19421945) are obviously an epoch within the Landis era. The pitchingdominated part of the 1960s (19631968) are obviously an epoch within the Baby Boomer era.
In the second era in baseball history (18931919) there is a very important little "bubble" after the 1910 introduction of the corkcentered ball. We think of that era as the dead ball era. From 1906 through 1909 the league ERAs were around 2.50 every year, then the new "lively" ball, the corkcentered ball, was introduced, and for two or three years there were pretty big hitting numbers. Cobb and Joe Jackson hit over .400; runs scored were way up.
Then the scuff ball—invented in 1910—took over and spread around the league, and by 1915 the league ERAs were back to around 2.70. But there is a little epoch there, within the era, which is very different from the rest of the era, but it was so long ago, 100 years ago, that only the most fanatical baseball fans even know that this offensive boomlet ever occurred.
The issue of epochs within eras is an interesting one, but this article is too long to go into that now. Thanks for reading. I think we will be in a new era in baseball history by 2017.
Bill James
Click here to download Bill's spreadsheet.