that nba example

OK, so the NBA analogy I invoked the other day has prompted some confusion. I’m in Australia and have two ASA papers to finish and several more room-escape-puzzles to try, so I don’t have time for a longish post about this, but:

Everybody agrees and understands that height is important for success in basketball. Why does height matter? Is it because players heads are higher? A little, for purposes of vision, but mostly height matters because of what it implies about how far people can reach, both vertically and horizontally. Taller people have longer arms, and plus their shoulders are higher so they can reach farther up with those arms.

You know what else is unusual about NBA players? The length of their arms, relative to their height. The NBA scouting combine explicitly measures arm length, because it’s something NBA teams are very interested in when evaluating the potential of players.

Can you name a basketball whose ratio of arm-length to height was actually lower than what that ratio is for the general population? I can: Yao Ming. There are NBA players more than half a foot shorter than Yao Ming who have longer arms. But, Yao Ming was 7’6″, and that kind of height made up for not having longer arms.

Larger point: if height is really important for NBA success, then you’d expect that, among NBA players, height is going to be inversely correlated with a lot of other things that matter for basketball success, even if these things are uncorrelated in the general population. You’d be expect shorter people who make it to the NBA to be better pure shooters, for instance. You might expect them to be quicker on average, and better ball-handlers. Or, another way to put the matter would be to say: how could a slower and less-coordinated athlete possibly make it into the NBA? You’d think the most likely scenario would be someone who is crazy tall.

Any dots that can be connected from this example to GRE in graduate admissions is left as an exercise for the reader.

Author: jeremy

I am the Ethel and John Lindgren Professor of Sociology and a Faculty Fellow in the Institute for Policy Research at Northwestern University.

3 thoughts on “that nba example”

  1. Well, for those who were not hyper-nerds from birth and played basketball, there is a real advantage to being able to shoot over people on offense and not have people shoot over you on defense. However, this is offset by the effects of height on dexterity. I’m not sure what you mean by GRE scores either–inversely related to imagination and creativity, maybe?


  2. I’m not saying there isn’t a relationship between height and agility, but if you use your eye-test of watching highly competitive sports for judging how strong that relationship is, you’ll get an exaggerated view–because shorter people have to make up for being shorter somehow to be observed in elite competition, and exceptional agility is one such way.


  3. This is Pearl’s whole point about collider variables. If you condition on a dependent variable, for example only consider the successful, then independent variables that are not associated in the population will be, often though not necessarily, negatively associated in the subgroup. Basically, the point is if you condition on an endogenous variable you will change the relationship between its ancestors which can totally screw up identification and estimation. Elwert and I discuss these issues at length in our paper, “Endogenous Selection” just published in ARS.Lots of substantive examples.


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.