Analysis of 2000 Predicted Medal Tally

There is just one published prediction that we found for the medal tally at the 2000 Sydney Olympics. Now that the actual results are known, the accuracy of the prediction can be assessed. We are using our own analysis method to calculate a single percent accuracy score. It basically calculates the average percentage difference between their prediction and the actual results for the final top 5 countries.

In 2000, the leading country on the medal table at the end of the Olympics was the USA, as predicted by Andrew B. Bernard of the Tuck School at Dartmouth. He predicted 39 medals, only one less than the actual result.

His prediction for the other top countries was not quite so accurate, particularly for Germany where his prediction was nearly double what they actually achieved.

Based on my analysis method, he achieved a percentage accuracy score of only 66% (average percentage difference of 34%), mostly dragged down by largely overestimating the number of gold medals by Germany.

TABLE of top 5 countries with predicted TOTAL GOLD medals won

Final Rank Country Actual Gold Medals Bernard Prediction percentage difference (%)
1 USA 40 39 -2.5
2 Russia 32 22 -31.3
3 China 28 19 -32.1
4 Australia 16 14 -12.5
5 Germany 13 25 +92.3
      percent accuracy score 66%


Prediction modelling by Andrew B. Bernard of the Tuck School at Dartmouth. He used a forecasting model incorporating four factors: measures of available resources, population and per capita income, as well as the share of medals in the most recent Summer Olympics and a host effect. See source: Andrew B. Bernard

Related Pages