Remember me

London Olympics—USA Predicted to Win Back Medals Title From China

July 26, 2012

Every couple of years I get the opportunity to work with the Wall Street Journal to help predict the medal winners in the upcoming Olympics. It's great fun, and it's especially cool because we take an actuarial approach to the process.

Here is a link to the excellent article on the Wall Street Journal website (wsj.com).

Instead of predicting the medalists in each event, the WSJ and I assigned probabilities that each contending athlete might win. For example, instead of simply counting a gold medal for the favorite in the 100-meter dash, Jamaica's Usain Bolt, we assign a probability that he will win. In this case, we assigned Bolt a 40 percent chance. We gave his teammate, Yohan Blake, a 20 percent chance, USA's Ryan Bailey a 20 percent chance, and Trinidad's Keston Bledman a 10 percent chance, with the final 10 percent going to another Jamaican, Asafa Powell. We use the probabilities in two ways: 1) we ran simulations to see who would win medals, and 2) we summed up the probabilities to predict a final medal count by country.

Here is a summary of the results:

Based on the probabilities provided by the Wall Street Journal for each of the 302 Olympic events for the 2012 Summer Olympic Games in London, we ran 1,000 simulations. Each notable competitor was assigned a probability of winning his/her event and a separate probability of winning a medal.

Countries with the Most Total Medals

The United States won or tied for the most total medals 998 times out of the 1,000 simulations. China tied or won the most medals three times. USA and China tied for the most medals once.

Countries with the Most Gold Medals

Based on our probabilities, the United States is the overwhelming favorite to win the most total medals, but it's far less certain that they will garner the most gold medals as a country. In our 1,000 simulations China broke through with the most golds 304 times. That's quite a bit fewer than USA's 746 times, but there is still a significant chance China will repeat as the country with the most gold medals in the Summer Olympic Games. Russia also has a chance to win the most gold medals in the 2012 Olympic Games. They took home the most gold seven times. There were 57 times where the USA, China, and/or Russia tied for the most gold.

Range of Medal Counts

In the 1,000 simulations, the most total medals the United States won in any one simulation was 118 medals. The fewest was 89. Here are the results for the four leading countries:

Highest and Lowest Medal Counts in 1,000 Simulations
Country Highest Lowest
United States 118 89
China 100 70
Russia 89 60
Great Britain 70 48

The most gold medals for the United States in any one of the 1,000 simulations was 52. The fewest was 25. Further results:

Highest and Lowest Gold Medal Counts in 1,000 Simulations
Country Highest Lowest
United States 52 25
China 48 22
Russia 38 16
Great Britain 35 7

Predicted Medal Counts

During the 2012 Summer Olympic Games, 302 events are scheduled with 302 gold medals to be awarded and a total of 937 medals overall. Based on summing the probabilities for each competitor's country the predicted medal counts by country are:

 
Country Gold Medals Total Medals
United States 40 108
China 38 92
Russia 29 83
Great Britain 22 66
Germany 16 49

 

 
 

COMMENTS (4 Comments, most recent shown first)

Pale Hose
Well, the top three nations were: 1) US with 46 gold; 29 silver; 29 bronze; for a total of 104 medals. 2) China with 38 gold; 27 silver; 23 bronze; for a total of 88 medals. 3) Great Britain with 29 gold; 17 silver; 19 bronze for a total of 65 medals. Very close on your predictions John.
2:36 PM Aug 17th
 
taosjohn
Some of the matchplay events,like boxing, award two bronzes rather than having a third place match. Also there often are a few ties for bronze.
7:31 PM Jul 26th
 
wjwinship
I get 906 total medals assuming one gold, one silver, and one bronze for each event. Where do the additional 31 medals come from?
4:44 PM Jul 26th
 
areuss44
It seems like the heavy lifting here is done in assigning the probabilities, but there's no clear explanation here or in the WSJ article of the method used. Are they just "guesstimates"?
2:21 PM Jul 26th
 
 
©2024 Be Jolly, Inc. All Rights Reserved.|Powered by Sports Info Solutions|Terms & Conditions|Privacy Policy