Gucci Little Piggy

Tyler Cowen has a good article with Kevin Grier at Grantland on the economic strategies countries employ to boost their medal counts.  In it he cross-references Nate Silver’s “Medalball” piece which Steve Sailer has also discussed.

Besides population size, program investment, and a focus on individual and easy-to-win events, Cowen makes the argument that countries can boost medal counts by playing to their comparative advantage:

Along these lines, countries can beat the system by focusing on their comparative advantages. Countries with a lot of coastline often have an advantage in swimming, surfing, and boating compared to landlocked countries. Countries with high altitudes have an advantage in training long-distance runners compared to sea-level countries.

Sailer previously touched on one strategy that would be easiest to exploit which boils down to the fact that women’s events would bring the most bling for the buck.  Sailer wrote:

The most obvious strategy is one followed by East Germany and China: it’s much easier to win medals in women’s events. Outside of gymnastics and a few other sports, the number of girls who, deep down inside, really want to do what it takes to win is smaller. So, focus on macho sports for women, such as women’s weightlifting.

Of course, being a first-mover here would be important.  A lot of new events have come online and many of them to bring women up to parity in terms of Game participation.  If I care about medal count I’m shifting more money to obscure women’s events because the competition won’t be that stiff.  Just as Billy Beane’s Moneyball strategy was quickly exploited by many other teams, there will be a short window of opportunity.

If the distribution of medals of the U.S. and China for the 2012 Games are any indication of this trend, of China’s 34 total medals (as of today), 21 were won by women (62%) while the U.S. is almost at parity – 18 medals for men versus 19 medals for women.

A couple of other thoughts on Cowen’s piece.  He writes:

A couple of nations greatly exceeded their economics-based medal forecast in the 2008 Games. Jamaica was predicted to win only four medals by Forrest, Sanz, and Tena (they wrote a well-known forecasting paper for medals) but actually won 11, while Kenya won 14 medals against its prediction of five. If you think those totals are not impressive, consider that Mexico, despite its higher income and larger population, only won a grand total of three medals in Beijing.

The obvious is not mentioned.  This is the Great ‘R’ – not ‘residual’ but ‘race’.  Of all nations we’d expect those with high black populations of West African descent to confound predictions based upon income or overall population totals.  This would be because of a more ‘natural’ athleticism which wouldn’t show up in these economic models.  This would be like trying to model per capita income without considering IQ.

Cowen also makes predictions:

Medal totals will become more diversified over time. The market share of the “top 10″ countries will continue to fall (it was 81 percent in 1988) as economic and population growth slows in the rich world. The developing world has greater room for rapid economic growth, and most parts of the developing world also have higher population growth. The Olympic playing field will get more and more level.

This prediction will probably pan out in the long-run, but it seems that the insertion of the 1988 data point is out of context.  First, remember that the Soviet Union was intact and huddled together many nations that now obviously compete by themselves.  I compiled a table of more data points which suggests that Cowen’s prediction does not seem to be panning out.  1992 is also an aberration because of the breakup of the USSR and the Unified Team which consisted of Russia and former Soviet republics.  If we look at medal distribution today we’d have to control for the dismantled USSR.   From 1996 (good starting point?) to these games, the distribution seems to be narrowing.  (table below compiled from Wikipedia entries for each Summer Olympiad)

Keep in mind though that there are different ways to measure the distribution.  The “top 10” make up a smaller percentage of countries going forward compared to the past, yet they seem to be increasing their gold medal dominance.  And the distribution of dominance for the top 10 should also take into account the top 10′s changing participation distribution.