Math brain predicts World Cup winners

Ian McHale of the University of Salford in England produces a vast statistical analysis in order to determine which country will actually lift the most prestigious sporting prize in the world.

Chris Matyszczyk
3 min read

So much money has already been wagered on the World Cup finals, which started Friday in South Africa.

So few of these bettors, however, will have turned to the Institution of Engineering and Technology's magazine before risking their savings, their house, or the money they were keeping back to pay the divorce lawyer.

The magazine, you see, has helpfully published an analysis of the World Cup, written by a man for whom numbers say so much. You might experience conflict with your potato chips when I tell you that the University of Salford has an Economics of Gambling degree course. Lecturer Ian McHale creates mathematical models in order to help peculiar humans understand their true odds.

Naturally, McHale's expertise is in much demand for the World Cup. So I can only hope that by passing along some of the professor's knowledge, I might save a marriage or two. Or at least a weekend.

The scene at the Opening Ceremony in South Africa. Yes, Switzerland is in the World Cup finals.

Bookies don't only use mathematical models to set the odds, McHale told the IET magazine. "An expert odds-setter is employed to adjust the model-generated odds, given any extra information," he explained. Now that would be an interesting job for a vast left brain.

McHale created his own World Cup model, basing his calculations on the outcomes of almost 9,000 international matches. He used a programming language called R and another called Matlab. I am sure they are both very fine and helpful.

"Predicting the winners of the tournament overall is not just a case of picking the best team. One needs to take into account the effect of the tournament structure," McHale explained to the magazine. Essentially, the first round is a group phase, in which two out of four teams go through. Only then does it become a knockout tournament.

So McHale created 100,000 simulated tournaments in order to reach his conclusions. I know you will pour through his table and calculations, which are presented on the IET's site (together with a video featuring McHale and his charming Liverpool accent). You just want to know the results.

All right, here are the predictions for the quarterfinals:

QF1: Holland vs. Brazil

QF2: France vs. England

QF3: Germany vs. Argentina

QF4: Italy vs. Spain

How depressing that the dour Italians might get so far. Also, it seems like the mathematics don't favor any of the African teams, which perhaps doesn't give true weight to the influence of the hosting continent. In 2002, for example, when South Korea and Japan were the hosts, the supposedly unheralded South Koreans made the semifinals.

One other thing that McHale pointed out was that the best value bet seems to be on the French, who staggered into the finals on two illegal handballs and a vast dollop of myopia from a Scandinavian referee.

So the semifinalists? According to the machines, you will see Brazil, Spain, Argentina, and France. And Brazil will meet Spain in the final--with Spain coming out victorious.

Strangely, this analysis concurs with so many pundits and experts. This means that either the World Cup results are entirely obvious, or that machines are making considerable and heartening progress toward becoming truly astute, sensitive humans.

(Disclosure--and ad: I will be, and am already, blogging about the World Cup for CBS News.com. Please come over and say hello. For Americans, I have already posted a special World Cup guide to help you gain enthusiasm for the coming weeks of uncontrolled world joy.)