Players in the 2014 World Cup tended to be born in the earlier months of the year:
Most countries’ youth soccer teams are broken up such that all players born in a given year are in one age group together. Players born in January are 11 months older than December-born players on their same team. They’re taller and bigger, so they get picked for better teams, get more attention from their coaches, get more playing time, and so on. They’re therefore more likely to make it to an elite level, even though there’s no physical difference left by the time they’re adults.
This phenomenon is known as the relative age effect, and has been observed in Canadian Hockey, Major League Baseball, and even English Grade School Test Scores.
So, how much does this affect the U.S. team’s chances? How important is it that we try to mitigate this effect moving forward?
To start, we know the U.S. has the same problem as the rest of the world because this is the distribution of the 1,300 or so 16-year-old players in the elite US Soccer Development Academy:
At age 16, there are three times more January-born players than December-born players.
The birth month disparity in World Cup players (the first chart) isn’t quite as bad: of the 23 players on a World Cup roster, on average 13.5 were born in the first 6 months of the year. If the distribution matched that of the general population, there would be only 11 born in the first half of the year. So on average 2.5 players “artificially” made the team; they’re not as good as some theoretical players born in later months who didn’t make it because they didn’t get the right chances in life.
Let’s kick them off the team, and add in some theoretical late-in-the-year players to replace them. Let’s assume that it’s equally likely that the “missing” player
would have been the 3rd best or 8th best or 16th best or whatever player on the national team.
So if you add one missing player, the team would have a 1 in 23 (4%) chance of getting a new best player, a 3 in 23 (13%) chance of a new top 3 player, and an 11 in 23 (48%) chance of getting a new starter.
If we extend that to 2.5 players, we have a
- 10% chance of getting a new best player
- 30% chance of getting a new top 3 player
- 50% chance of getting a new top 6 player
So. We’d certainly be better at soccer. And a 30% chance of another Bradley/Howard/Dempsey sounds pretty good. But we wouldn’t be favored over Germany anytime soon. For that, well, we’ll probably need a few more of the next Kobe Bryants and Wes Welkers of the world to grow up loving soccer instead of basketball and football.
It’s not useful to look at any one individual team’s distribution because of the small sample size, so we look at the distribution across all World Cup players. This entire analysis could be applied to any team.
Because of seasonal birth variations and February’s short length, the first half of the year has fewer births generally than the second
One could argue that the very best of the late-month players get discovered and developed no matter what, and therefore the players who would get an additional chance in a fair world wouldn’t be randomly distributed in skill level, they’d be somewhat less likely to be one of the top few players. There’s probably a tiny bit of truth to that, so you could consider the odds calculated from this assumption to be the maximum amount of advantage to be gained from a more equitable system that avoided the relative age effect.