The Gold Medal Chase

I have a wager with a friend on which country will win the most gold medals. It’s not very patriotic, but I have China, and he has the United States.

Dartmouth economist Andrew Bernard, who has built an econometric model to predict Olympic medals, thinks it’s a coin toss.

Bernard’s model couldn’t be much simpler: the main variables that go into the prediction are a country’s population, GDP per capita, how many medals they won in the last Olympics, and whether they are the host country. Historically the model has predicted well. (As an aside, if you ever need to predict anything and you are only allowed one variable, using what happened last time is probably the best variable you will ever find.)

When it comes to predictions, however, I put more faith in markets than I do in any individual economist. Just before the start of the opening ceremonies, Tradesports had China twice as likely as the United States to win the most gold medals. The volume was extremely low — only $1,000 in trades total — but still that gave me confidence.

I have to believe that when a country with as much at stake in these Olympics as China sets its mind on finding ways to beat the anti-doping rules, they will succeed.

So far, the results are consistent with my conjecture. China leads the U.S. 11-7 in gold medals and Tradesports now has China as 78 percent likely to win the most gold medals.


Leave A Comment

Comments are moderated and generally will be posted if they are on-topic and not abusive.



View All Comments »
  1. totallydawesome says:

    I thought Andy Bernard went to Cornell??!

    Thumb up 0 Thumb down 0
  2. Mike B says:

    Why does everyone assume that the winner of the Olympic Games is the one with the most first place finishes? Multi-team, multi-event sports rarely give the championship to the team that simply wins a plurality of the events.

    The Olympics should calculated on either a 3-2-1 or a somewhat better 5-3-1 system with 5pts for a gold, 3 for a silver and 1 for a bronze. Moreover, to prevent countries from gaming the games by targeting sports with high medal density, like China is doing with their focus on weightlifting, team sports with tournaments should use a 10-5-3 (or other proportionately weighted) point payout.

    After the games the countries should be ranked by point totals and then cuts of the television and merchandising revenue would be distributed to the top ranked NOC’s in a fashion similar to a golf tournament. The heads of state for the top three countries would also be presented with some sort of cool championship medals at the closing ceremonies.

    Thumb up 0 Thumb down 0
  3. Pierre says:

    It find it weird for an economist to say such thing as “it might not be patriotic, but my bet went to China”.

    Doesn’t it create huge arbitrage opportunities / distortions if we can expect everyone to bet for their own country ?

    Thumb up 0 Thumb down 0
  4. Gary says:

    So what happens if in the future, the results are changed at some point in the future by medals being revoked? This year we’re keeping samples to be tested for 8 years into the future. Suppose something in Michael Phelps 10,000 calorie a day diet proves problematic? Perhaps several Chinese athletes had some tainted lo mein, explaining their unnatural strength?

    With China and the USA running so close in total medals (albeit with China winning many more golds) the total medal count could be altered by just 1 or 2 juicing athletes getting caught.

    Thumb up 0 Thumb down 0
  5. science minded says:

    a missed point. Why is China likely to win the most gold metals. I think that they have two things going for them 1)tradition and 2) society. Did you see the opening of the olympics? The circular like movements- that goes back around 4-5,000 years to their fighting approach of “rhythmic disciplined archery.” And generally speaking, men do better when they play on teams.

    So my guess is that America is likely to win when the women don’t play on teams- but compete alone.

    Thumb up 0 Thumb down 0
  6. Bernardo A says:

    @ Pierre:

    probably the other way round: with marginally decreasing utility (which is always reasonable), the expected payout of betting against your home country is higher than betting for your country. In other words, if I bet against my own country and my country loses, at least I win some money and viceversa.

    Thumb up 0 Thumb down 0
  7. Ben says:

    I didn’t think medal counts by country were an official olympic statistic, just something the media feeds us to bolster interest & national pride.

    Thumb up 0 Thumb down 0
  8. Jessica says:

    Am I missing something? His model does predict China will win the most gold medals; 37 of ‘em to only 36 for the US.

    Thumb up 0 Thumb down 0