Friday, January 31, 2014

Comparing MLS height bias to European leagues

Earlier this week I started to look at the issue of bias within the US soccer system based on a player’s height. While it was clear that tall players are favored over short players, it wasn’t clear whether that was true globally and to what extent. To do that, we'd have to look at other leagues around the world.


I examined players playing within the league of their nationality to remove the potential issue of a league transferring in a bunch of short or tall players from abroad. It is widely accepted that height within a specific region (like a country) is normally distributed. So, I found average heights and standard deviations for 10 different nations in Europe to compare the US to so we can see if this is just an American issue. By using standard deviations, we normalized the data across nations to account for that particular nation’s height profile so it makes sense to compare different nations. For those who are unfamiliar with normal distribution and standard deviation or need a refresher, click here.

What I’ve done is created a category for players within one standard deviation of the mean in each direction and then another category for players who are more than one standard deviation from the mean. What we should see is about 68% of players should be within one standard deviation from the mean. We’ll call them “average players” in terms of their height. This was pretty close to true across the leagues. Where it becomes interesting is when we look at players that are more than one standard deviation from the mean.

Again, I will be using location quotients to compare the US to the European leagues. Smaller values (less than one) mean there are fewer players in that height group than should be expected. Values close to one mean there are about as many players in that category as would be expected. Large values (greater than one) mean there are more players in that category than should be expected.


The results show that, pretty much across the board, there is some bias toward choosing taller players (more than one standard deviation from the mean). This shouldn’t be that surprising as there are obviously some positions (mainly keeper and, to a lesser extent, center back) where it is advantageous to have a bit of height. Interestingly, Portugal and Italy have a pretty strong favor for tall players whereas Finland, Belgium, Austria, and Sweden don’t seem to care all that much about player height.


Alright, so many readers peppered me with questions wondering about how USA stacked up against foreign leagues. Here it is: we have a higher likelihood to see tall Americans in MLS than what is experienced from these ten European nations, on average. On the flip side, we see a much lower likelihood of short players in MLS than in Europe. If we combine all 10 European countries into a single group, we see that it is roughly three times less likely to have a short player in the US than it is in Europe. Even with the inclusion of the relatively more tall-favoring leagues of Italy and Portugal, the USA had a higher concentration of tall players than Europe

Pretty damning evidence that the qualities valued in the US soccer environment are based more on physical characteristics than they are on soccer abilities.

In the next post in this series (likely sometime next week), I’ll be looking at where exactly the short players get weeded out from becoming pros in the top domestic league in the US.

Note: the leagues chosen for this examination were purely based on which nations had the necessary height data available and that I found.

No comments:

Post a Comment