In the women’s individual gymnastic competition at the Olympics, there were a couple of controversies over scoring that experts believe cost the US gymnast a gold. The problem is that officials from the top countries cannot be judges if a citizen from their country is in the competition, so the remaining judges are from countries with no gymnastic tradition. Is there an efficient way to solve this problem?
Doney Joseph, Los Angeles, CA
In a sport whose entire aim is to win the approval of the judges, error is unavoidable. Yet (as the sports economist Stefan Szymanski reminds me) there is a trade-off here between two sorts of error: bias and variance. The decision to use disinterested but inexperienced judges should reduce bias but increase the variance of the outcome. That is a painful trade-off.
National pride is not the only source of bias. Judges have friends; they may know athletes, or their coaches. It was for this reason that the great US orchestras began to introduce “blind auditions”, where musicians performed behind a screen. The idea was to eliminate favouritism towards the students of particular teachers. The unexpected result – shown by the economists Claudia Goldin and Cecilia Rouse – was to disarm sexism as well as favouritism: many more women succeeded in the blind tests.
Alas, gymnasts can hardly perform behind a screen, although I suppose they could compete in balaclavas. All I can recommend is an alternative reward system for judges. Each should assign scores in isolation, and then be paid only if the other judges agreed. The judges will need to co-ordinate on a “focal point” if they wish to be paid, and the obvious focal point is their honest opinion of a gymnast’s performance.
This is surely worth a try. If the judges cannot agree, of course, they will be paid exactly what they deserve: nothing.
Also published at ft.com.