≡ Menu

In 2006, Doug looked into how much of a role the month in which you were born could play into your chances for athletic success later in life. Doug didn’t just ponder this out of thin air: a bit more research has been spent on this topic than you might think. Steve Levitt, of Freakonomics fame, found some evidence indicating that “older” kids in the same level of play — older by as much as 365 days, I suppose — tended to be more likely to become professional athletes. Basically, if you’re the oldest kid in your travel soccer team or 8th-grade basketball team, chances are you will be better than the other kids. This leads to a snowball effect, where you might be more likely to receive more personal coaching and your confidence should increase.

J.C. Bradury graphed the birth-month of over 16,000 major league baseball players. Take a look:

Birth-month of every MLB player through 2005

Bradybury notes that the cut-off date for participation in Little League baseball is July 31. The graph would seem to support this theory: kids born in August, September and October would be “older” than their Little League teammates born in May, June, and July. And it just so happens that August/September/October are the birth months of a much higher percentage of Major League Baseball players than May, June, and July.

Doug did a quick study of active players he had in his database and saw the same general pattern. But this was seven years ago, and we have much more data now. So let’s test this theory.

Here are all players to enter the league since 1990.

Month#Perc
17329.3%
26067.7%
36558.3%
46057.7%
56187.9%
66197.9%
76578.4%
86568.4%
97039%
106488.3%
116678.5%
126888.8%

May and June aren’t great months, but they’re not really outliers, either. This looks like a pretty random distribution to me, but what if we go back another twenty years? The next table looks at only players to enter the league from 1970 to 1989:

Month#Perc
16389.3%
26068.8%
36449.3%
45347.7%
55407.8%
65167.5%
75788.4%
85648.2%
95828.4%
105848.5%
115347.7%
125758.3%

How about every player with a rookie year between 1950 and 1969?

Month#Perc
13289.2%
23018.4%
33229%
42727.6%
52807.8%
62657.4%
73188.9%
83269.1%
92978.3%
102928.2%
112817.9%
122928.2%

For this sort of exercise, tables aren’t as helpful as graphs. The black line shows births for players entering the league from 1990 to 2012, the blue line shows players in the 1970 to 1989 period, and the red line is for players who entered the league between 1960 and 1969.

age distribution 2

The NFL isn’t Major League Baseball in many ways; I don’t think there’s a cut-off for football the way there is for Little League baseball (although I don’t know anything about Little League baseball, either; I’m assuming there is a national cut-off, or more importantly, that there used to be). Most football players receive their training through school programs, and I think the cut-off dates for grade levels vary pretty significantly throughout the country. So even *if* there was an actual effect here, it might not show up.

Of course, here’s something else to keep in mind: July/August/September are the most common months for birthdays in the US. To the extent there is something up with the data, my hunch is it has to do with the high number of January birthdays. January is an outlier for all three periods in the study while also being a below-average month for births. To the extent that there are football-related activities with December 31 cut-offs, then this might make some sense. It also might vary by position: we already know Valentine’s day is for passing lovers.

{ 5 comments }