Population Versus Expansion

            In 1960, before the first expansion, each major league team represented or was drawn from a population of about 12 million people. 

            This article does not address the relative strength of the American League versus the National, but rather, the relative strength of the major leagues as a whole over time.  After the first expansion, in 1961-1962, it was commonly said that this had "diluted" the quality of play.   Reporters said in 1962—yes, I am old enough to remember this—observers would say that the major leagues now consisted of 16 teams full of major league players mixed with 4 teams worth of minor league players.   Actually, they would continue to say this for many years after that—that expansion had diluted the quality of the product, so major league teams were not as strong as they were years ago, before all them minor league players were added to the majors. 

            To assume that the number of fully qualified major league players does not vary over time is of course silly, and to assume that the number of fully qualified major leaguers varies proportional to the population is fraught with problems.  There are, after all, nations of many millions of people which don’t produce any major league players at all, because people there just don’t play baseball.  You can’t assume that the number of players within a population is even remotely the same in all countries, nor can you safely assume that it is remotely the same in different decades. 

            But the argument that expansion diluted the quality of play in the major leagues must be true to some extent, so we need to try to model the problem and estimate what that effect might be, as best we can. 

            In 1876 the population of the United States was about 45,500,000, more or less.  There were eight teams that we now recognize as major leagues teams, rightly or wrongly, so that is a ratio of about 5.7 million people for each team.   In 2021 the population was about 333,000,000 and there were 30 teams, so that is a ratio of about 11 million people for each team. 

            These numbers have to be modified, however, for at least three things:

1)      Segregation, which limited the number of potential athletes from which each team was drawn,

2)     World War II, which severely limited the number of available athletes for a period of time, and

3)     International Scouting, which has drawn into the major leagues now a very large number of players from Cuba, the Dominican Republic, Venezuela, Canada, Mexico, Japan, Australia, South Korea, Thailand and Zimbabwe.  None from Zimbabwe?   We’ll get there. 

So this is what I did, to modify the raw population numbers to create a more realistic estimate of the relevant population.

1)     Segregation.  For each season prior to 1947, I multiplied the actual population by .70, assuming that 30% of the eligible population of the best athletes was banned from what was then regarded as Major League Baseball.   That’s a conservative assumption.  After 1947, I increased that number (.70) by .01 each year until 1949, by .02 in each year from 1950 through 1959, and by .01 each year after 1960 until the number was 1.00.  This creates a pattern of integration essentially consistent with the research reported here in the article BL3.   The number reaches 1.00 (100%) in 1966.

2)      World War II.  World War II, of course, took away most of the eligible population of baseball players, leaving the leagues to be staffed by whoever was left.   I estimated the percentage that was left as:

90% in 1942

70% in 1943

50% in 1944

30% in 1945

50% in 1946

70% in 1947

90% in 1948 and

100% in 1949

The population of available athletes did not immediately snap back to 100% after World War II because, of course, a great many young men who would otherwise have become major league athletes had been killed or disabled in the war, or had had their development years, ages 18 to 22, completely taken away from them. 

3)      I marked the beginning of international scouting at 1950, which is about when we began to have significant numbers of international players in the majors, although of course there had been some earlier.  

To represent this in the chart, I multiplied the available population by 1.0032, and then multiplied that number each year by 1.0032.  That is saying that the influence of international scouting has been growing at a rate of about one-third of one percent per season, so that by 2021 international scouting had increased the pool of potential players by 26%.   That’s a really conservative number, but again. . .better to make too small an adjustment than too large. 

      Using these assumptions, the effects of the expansions beginning in 1961 were not fully overcome until 2007.   There were repeated expansions—1969, 1977, 1993 and 1998.  Each expansion diluted the talent a little bit more, thus setting back the time when the effects of expansion would be fully overcome.   

In 1876, then, the population of the US was about 45,547,000 (you get different figures from different sources).  However, since black players were not included in the game (with a very few exceptions), I count this as an effective population of 31,883,000.  There were 8 teams in the NL, which makes a ratio of 3,985,000 potential athletes for each team.   Of course, we could reduce this number by eliminating the women folk and the children and aged, but since these adjustments would be essentially the same for each season, they would have little impact on the conclusions. 

 How do we add this to our chart, which is based around .500?  

I divided the "ratio" in the chart above by the same plus 6678; in other words, the effective winning percentage for the National League in 1876, based on the size of the population from which the teams were drawn, would be .374, since (3985/ (3985 + 6678)) = .374.   6678 was the number I used because that number makes the chart center at .500.  Adding that to the chart:

            New leagues forming in the 1880s drove this equivalent winning percentage way down, as low as .171 in 1884, the lowest it has ever been.  In 1884 each team was drawn from an effective population of not much more than a million people. 

      The Player’s Revolt in 1890, leading to the formation of a third league, gave us the second-lowest number ever in 1890, but the consolidation into one 12-team league made the ratio of population to teams much higher, thus presumably making the league much stronger.   Not just "presumably"; there is no doubt that the quality of play in the majors was improving, although the population ratio alone would not say that it was better in 1899 than in 1879:

          The National League shucked off four teams in 1890s, creating the strongest league ever at that time for the 1900 season.  In 1901 the American League formed, dividing the talent between the two leagues, and dropping the presumptive skill level back to the .330s:

          From 1901 to 1960 the ratio of population to teams grew steadily, as the population of the US more than doubled in those years, while the number of teams stayed at 16.   There were two exceptions to that, the first of which was the Federal League, 1914-1915:

          By 1929 the presumptive winning percentage based on population was up to .443:

          And by 1939 it was up to .479, the highest it had ever been except for the 1900 season, when there were only eight teams:

          And then the 1940s were wild.   By 1945 the presumptive quality of baseball (based on the population ratio) had dropped to its lowest point since 1890.   By 1949, aided a little bit by early integration, the presumptive quality had climbed over .500 for the first time ever.  It has never dropped under .500 since then:

          By 1959, with large-scale, normalized integration, the quality of the leagues had reached a dizzying level:

          What we are saying here is that, based simply on the size of the American Population and the introduction of black players and international players into the game, we would conclude that a major league team from 1959 would beat the bejeebers out of a team from 1939. 

          By 1969, however, the major leagues had expanded by 50% in 9 years.  This diluted the product, and set backward significantly the overall quality of play:

          Because of international scouting, the effective population became larger than the American population in 1962.  Remember Camilo Pascual, Roberto Clemente, Orlando Cepeda, Julian Javier, Juan Marichal, Luis Aparicio and the Alou brothers?   Because there was another expansion in the 1970s, not a lot of progress was made toward recovery:

          Much more progress was made in the 1980s, pushing the product back close to the pre-expansion standard:

          But the additions of four more expansion teams in the 1990s left the quality of the league about the same in 1999 as it had been in 1990:

          With no expansions since 1998, the effects of expansion were finally wiped out by 2007, leaving the quality of play stronger than ever:

          And it has continued to grow since 2009:

          Each team now represents an effective population of about 14 million people, including international players.  This is the highest ratio of all time. 

          Just by way of my opinion, I will say that I don’t really believe that it takes as long for the quality of play to recover from expansion as this chart makes it appear.  Major league players do not simply exist; they are created--like lawyers, writers, educators and criminals. They are created out of talent, which is limited by the population size, but they are also created by training, development and opportunity.  Those things are not limited by the population size.   My opinion. . .the effects of the first expansion (1961-1962) were probably mostly gone by 1966, 1967.  Mostly gone, not entirely. 

          But. . .this is just intuitive, maybe, but it seems to me it took baseball longer to recover from the second expansion (1969) than from the first one.  You have two operations in a fairly short period of time, it’s going to take you longer to recover from the second one than it did the first one. 

          And my second opinion:  most people tend to dramatically overestimate the improvement in the quality of play over recent decades.  I know that a lot of people think that a team from 2022 would easily dominate a team from 1982, if a game could be arranged.  I really don’t see convincing evidence for that. 

          But there are a lot of things left that we could or will measure, and all of those things will feed into the estimate we make of the slope of history.  I am trying, as best I can, to take this issue out of the realm of complete speculation, and move it into the realm of educated speculation, careful speculation, organized speculation.  It’s a hard problem; take my work for what you think it is worth.