I’ve always been curious about competitive balance. How balanced is major league baseball today? How balanced has it been over time? How much balance is too much, and how little balance is too little? What are the obvious factors impacting competitive balance?

Anyway, I decided to do a little study of this issue. I should say right up front that I have no idea whether somebody has done this already, and if they have, I apologize. I know that Bill James had an “Index of Competitive Balance” in the New Historical Abstract, but I don’t know how he calculated it.

What I did was very, very simple. For each league/year, I subtracted each team’s winning percentage from .500; then, I took the absolute value of that number and multiplied it by 1000. Here’s the 2002 American League:

I then simply calculated the average score, which in this case was 90. That is the league’s Competitive Balance Score, or CBS. Theoretically, a perfectly-balanced league would have a score of zero, and a totally imbalanced league would score at 500. Of course, those extremes would never happen. Every single major league in baseball history has scored between 35 and 192. At 90, the 2002 American League was the most imbalanced league since 1962.

The post-1900 record is 127, set by the 1909 National League:

Three-quarters of that league was either over .600 or under .400; I’d say that’s pretty imbalanced. The most balanced leagues of all time were the 1968 and 1983 editions of the NL, both scoring at 35:

There has been a general trend over time towards greater balance. The all-time CBS average is 75; looking just at the AL and NL since 1900, the average is 70. Breaking it down decade-by-decade, here are the average league CBS scores over time:

 1870s  141

 1880s  113

1890s   95

1900s   92

1910s   77

1920s   75

1930s   83

1940s   76

1950s   74

1960s   65

1970s   59

1980s   54

1990s   55

2000s   61

This seems to suggest a handful of “eras” in baseball history:

 1870s-1880s   118

1890s-1900s    93

1910s-1950s    77

1960s-2000s    59

Actually, you could begin that last stage in about 1957; the late fifties were very balanced.

General trends aside, there’s a good bit of year-to-year variation. A league might be very unbalanced one year and quite balanced the next; overall, the average year-to-year variance since 1900 has been +/- 15. However, there are trends in there too: for a long time,  the average was +/- 17-18, but since the 1970s, the average variance has been around 10. In other words, leagues are becoming both more balanced and more stable in that balance.

One of the most interesting things I found is also one of the most intuitive: expansion years, without exception, see a significant jump in imbalance:

Those are some of the biggest one-year jumps of all time. In fact, since expansion began, five of the eight largest one-year increases in CBS have come in expansion years, including the two largest (1969 NL and 1977 AL). And the rise is almost always just a one-year spike, with a stabilization the year after expansion.

Overall, expansion has encouraged competitive balance. So has integration, the amateur draft, and the influx of Latin Americans. But the two leagues aren’t always following the beat of the same drummer. The NL integrated more quickly than the AL, and expansion has taken place in different leagues at different times. Here are the averages for the two leagues, given in five-year chucks to smooth out the year-to-year variations:

During the first decade of the 20th century, the NL was extremely imbalanced, but the AL was more moderate. That changed in the 1920s; from the late ‘20s until World War II, the AL was consistently more imbalanced than the NL. They reversed positions during the war, but in the postwar years, the AL reached levels of imbalance that have not been seen since.

Part of this was the Yankee hegemony, but only part: you also had teams like Cleveland and Boston at the top, and bottom-feeders like the Senators, Browns, and A’s. And you’ll also notice that the AL became dramatically more balanced in the late ‘50s, prior to the fall of the Yankee empire in the mid-‘60s (and prior to the expansion of 1961). How dramatic was this? Here are the CBS figures for the AL from 1954-1958: 125, 94, 83, 65, 36.

This was not a fluke; it was a real, substantive shift of competitive balance. And it was at a time when the Yankees were still winning pennant after pennant. What were the causes? Several factors, I think. In the ultra-balanced year of 1958, the Indians had four key black players (Minoso, Doby, Vic Power, and Mudcat Grant). But the AL was slower to integrate than the NL, so that factor can’t account for everything. The Browns/Orioles of the early fifties were the worst team in the league at producing talent, but in late 1954, Paul Richards took over as GM/manager, and the franchise began a steady ascent. White Sox GM Frank Lane made a number of good moves, putting the team in a position to contend. I’m sure these aren’t the only factors, but they’re the most obvious.

Anyway, the other notable thing about that graph is that the recent disparity between the AL and NL. While the NL has remained steadily balanced since the late 1960s, the AL has seen a major jump in imbalance in the new millennium. Using rolling five-year averages, the AL has been more imbalanced than the NL by double digits since 2003. The last time the two leagues were consistently this far apart in terms of balance was the 1950s.

One big difference between the leagues today is payroll disparity. In 2008, the standard deviation in AL payroll was $42.8 million, while the same figure for the NL was just $29.3 million. Contrast that with, say, 2000: the AL standard deviation was $26 million, the NL $24 million. Payroll isn’t the only factor, of course; Tampa Bay is proof enough of that. Still, salary differences are probably the biggest reason why the AL is so much more imbalanced than the NL these days.

What degree of balance is desirable? Obviously, a league like the 1909 NL isn’t much fun for the majority of fans, with a few dominant teams and a few awful teams, and hardly anyone in between. But neither is it all that exciting to watch a battle of mediocrities. A good league, in my view, is one that has a healthy mix of great, good, mediocre, and bad teams. Parity is good in the sense of turnover: it’s nice to have new contenders cropping up every couple years. But a league full of .500 teams just isn’t all that exciting.

Sometimes, very imbalanced years produce great pennant races. The AL races in 1920 and 1948, and the NL races in 1908, 1941, and 1962 (just to name a few) all happened in seriously imbalanced seasons. But other great races came in rather balanced leagues, like the NL in 1951, 1959, 1964, and 1991; or the AL in 1967.

Finally, I would propose that the question of competitive balance is relevant when we talk about all-time great teams. The Cubs of the early 1900s dominated a league that had a few really great teams and not much else. So too the 1948-56 Yankees – though in that case, the franchise continued to be successful for many years afterwards. The ’27 Yankees played in a league with a CBS of 98.

On the other hand, the Big Red Machine was a dynasty in an otherwise very balanced league; the same can be said for the late ‘90s Yankees. During the Braves’ long string of division titles, the National League was quite balanced. Personally, I’m inclined to give more weight to a team that was able to dominate a balanced league than one that racked up lots of wins against hopeless competition.