Skip to main content

The Globe and Mail

A financial take on predicting the World Cup

With the kickoff of soccer's World Cup, market watchers are predicting calmer waters for financial markets. Participants will be watching the pitch more than the trading desk.

"Risk on, risk off, Europe, U.K., U.S., fiscal consolidation, equity market performance, all manner of economic data - as of today, all such considerations merely become mere inconveniences apart from during half-time breaks and between matches," wrote David Watt, senior fixed income and currency strategist at RBC Dominion Securities. "Take a trip to any trading room and I challenge you to find any television set not turned toward footy."

Well, if traders are going to focus more on the beautiful game than the ugly markets, we see no reason why Market Lab shouldn't, too. In fact, some people are way ahead of us in transferring their analytical skills from picking stocks to picking World Cup winners.

Story continues below advertisement

Predicting the Cup

Quantitative researchers at UBS AG have created a model to predict which countries are likely to succeed on the World Cup 2010 stage. Having analyzed past World Cups, they have narrowed down the vast array of variables to just three key factors, which form their quant model: Past performance, being the host nation, and Elo Rating.

The Elo Rating is an "objective quantitative measure" of each team's strength, adapted by the world of soccer from a system developed by American physicist Arpad Elo to rank chess players. It looks at each team's game results in pre-tournament play, layering in other elements such as the margin of victory, the importance of the game (e.g. World Cup qualifier vs. exhibition match), the location of the game and the ranking of the opponent.

The model suggests that host nation South Africa has the highest probability of all teams to advance to the Round of 16 - reflecting the fact that despite the team's low world ranking, a host nation has never failed to make the Round of 16.

As for the overall Cup winner, the model leans toward Brazil - yet even as the favourite, its chances are less than 1-in-4.



Sometimes it works, sometimes …

UBS's soccer quant model has been tested twice before: in World Cup 2006, and in the 2008 European Championship.

Story continues below advertisement

In 2006, the model gained fame for correctly identifying 81 per cent of the final 16 teams, 75 per cent of the final eight, and half of the final four. But it flopped in Euro 2008 - correctly predicting only 63 per cent of the final 16, and none of the final four.

"The moral of this story is that one needs to be humble about the predictive power of one's models," wrote UBS chief economist Andreas Hoefert. "Successful forecasting can often depend as much on luck as skill, which is a lesson that is too often forgotten when it comes to quantifying the future."

Report an error Editorial code of conduct Licensing Options
As of December 20, 2017, we have temporarily removed commenting from our articles. We hope to have this resolved by the end of January 2018. Thank you for your patience. If you are looking to give feedback on our new site, please send it along to feedback@globeandmail.com. If you want to write a letter to the editor, please forward to letters@globeandmail.com.