Does it matter when NBA players were born?

I once read something about how professional athletes were more likely to be born at the beginning of the year. I don’t know the provenance of this. The source is probably questionable. But I was thinking about it today while discussing birthdays with a friend.

So, I decided to check it out. For one sport anyway. I scraped birthdays on Basketball Reference, giving me several thousand ABA/NBA players going back to the 1940s.


So, this isn’t too promising. As I remember, the theory says players born in the beginning of the year would have some advantage as children. Someone born in January would have ten months more physical maturity than someone born in October, for instance. This could be significant when you’re eight years old, setting you up for life.

This doesn’t seem to bear out by the time they get to the NBA. There might be some noise in here from international players – seasons and school terms differ, especially between hemispheres. But international players are a small minority, especially over the life of the ABA/NBA.

Some interesting research suggests that peak birth rates are influenced by latitudes. The bulk of the United States lands between July and September. As we can see, NBA players don’t seem to follow this pattern.

So, I decided to slice the data a few other ways. Maybe the theory might hold among elite players? Let’s look at NBA Hall of Famers.


This is getting closer to the distribution we’d expect in the United States. But we’re only playing with about 150 people here, so let’s not read too much into it. Anyway, it doesn’t really support the theory. The peak comes around the beginning of the basketball season in America.

The last notion I have is that older players are throwing us out. Over the past couple of decades there’s been a lot of infrastructure built to target young prospects – camps and tournaments etc. It stands to reason those born in the past thirty years could be more affected by birth month.


So, here are all the players born since 1990. Again, this distribution looks a bit more promising (for the theory) than the entire dataset. But we’re still only dealing with about 600 players. I want to look more into this.