Business Analytics Blog

Business Analytics Blog

Opinions expressed on this blog reflect the writer’s views and not the position of the Capgemini Group

Mathletics

Category : Sport

There are 153 days left before the Olympic Games in London. Although the founder of the modern Olympics Pierre de Coubertin says the Olympics spirit emphasises participation and not winning, athletes may disagree. In the lead up to the event, sports enthusiasts may be asking themselves, how many medals can Great Britain win on home soil? Can GB maintain its top four position in the Olympic medals table?

Many of these questions will be answered with varying degrees of accuracy in bars and pubs across the nation. Usain Bolt will probably win three gold medals for Jamaica if he does not false start; the American basketball team will probably retain the men’s basketball gold medal; Michael Phelps might not win seven gold medals in London, but only a brave person will bet against him winning a handful; you’d probably expect Jessica Ennis to win a heptathlon medal for Team GB. Drawing on analytical methods can however provide an alternative basis for an estimate for overall medal performance .

Using an econometrics model can help to predict Olympic performance by investigating factors or variables that could have an influence on a countries’ medal count. The influences of the identified variables are estimated by adopting the Classical Linear Regression Model based on the 2008 Beijing Olympics medal count, shown below.

Y = β0 +1.95G +0.003P +39.2H + 0.91C +ε Below are an explanation of the variables that are used in the Classical Linear Regression Model.

Economic Resources (G) - A cursory look at the Beijing 2008 Olympics medal table will quickly reveal that the top four countries by gold medals won are permanent members of the UN Security Council; France the last permanent member came in the top 10. It seems safe to infer that economic resources are a determinant of performance. Richer countries are able to afford to spend more on better training regimes. As a proxy for economic resources we will adopt the GDP per capita.

Population(P) - Additionally, a large population increases the potential group of athletes and enhances the chances of a nation winning more medals. An example is China with its population of well over a billion people improves the odds of producing an Olympic champion. Conversely India has a large population but poor performance in the Olympics. This variable will be the total population of each country.

Host Nation effect(H) - One can also not ignore the host country effect: host countries are allowed to participate in all events, and the majority of the crowd at the events will be rooting for the home athletes. There is also the possibility of extensive practice at the designated venues by the same home athletes. China demonstrated the power of home advantage by topping the medals table in their backyard four years ago. The host nation will be represented by dummy variables, 1 for host nation and 0 for all other countries.

Past performance(C) - Finally, the past is a good indication of the future and past performances should reflect expectations of the future model. The medal count from the Athens Olympics will be included as an explanatory variable.

Prediction(Y) – The number of medals predicted to be won.

The Classical Linear Regression Model results shows the contribution that each variable has in predicting a countries total medal count. It estimates that just being the host nation, contributes to 39 of the total medals that would be won. By applying all of the variables, the model makes the following predictions for 2012.

• Great Britain, 68 medals • Australia, 46 medals • France, 31 medals

This is obviously good news for the GB team. We also ran the model for the French and Australians (We didn’t have time to predict them all). Now it’s just down to the athletes to stay fit and prove the prediction right! Come on synchronised dividers and gymaths. (Sorry).

About the author

Nigel Lewis
Nigel Lewis
Nigel leads the Capgemini Consulting’s 35 strong Business Analytics team, which delivers analytical, operational and strategic modelling solutions to clients. He has 18 years consultancy experience as well as 8 years experience in the UK gas industry. Nigel has successfully managed complex projects in both the public and private sector, including capacity modelling, simulating supply chain operations, strategic business modelling to support future policy decisions, and implementing complex demand forecasting systems. Nigel is currently focussing on the development of Capgemini’s customer analytics and analytics advisory services.

Leave a comment

Your email address will not be published. Required fields are marked *.