When mathematics comes to football

KE NAKO!  For those who do not know it this is the slogan for the 2010 FIFA World Cup currently underway in South Africa. It is a Sotho phrase meaning, “It is time.”

To the South Africans and indeed the rest of Africa this World Cup is the continent’s time to shine and cast better images to the world than those pictures of poverty, civil war and tyranny usually associated with it. Despite earlier worries of slow preparation, prompting rumours that the tournament would be moved to another country, the competition in fact commenced in high spirits with a gig on the eve of the kick-off. The opening match was riveting with the hosts South Africa drawing one-all with Mexico.

There has been isolated whinging by players about the official match ball, Jabulani, especially its unpredictability regarding when it will drop and its fast speed. Another controversy surrounds the vuvuzelas, the horns that are being blown at matches. A few players have said the noise disrupts their communication on the pitch while some supporters have resorted to using ear plugs to block the sound. Be that as it may, the celebration of Africa’s first ever World Cup continues to reverberate across the continent and the event is even attracting interest from some unlikely followers in the world of mathematics and finance.

Weeks before the kick-off, mathematics whizzes from three top investment banks on Wall Street publicised results of their quantitative models for the 2010 World Cup. The three banks are UBS, JP Morgan and Goldman Sachs (GS). UBS geniuses were formerly hailed after they correctly predicted that Italy would win the 2006 World Cup.

 

However, probably reflecting that the prediction was rather an act of luck than science, their model dismally failed to foretell the outcome of the 2008 European Championship. Spain won the tourney while the predicted favourite, the Czech Republic, did not even proceed beyond the group stages. Worried by these two contrasting results, UBS chose to be more vigilant this time round. It warned its readers to take the current prediction “with caution and a pinch of good humour”.

Its model relied on three factors, namely, past performance, whether or not a team is hosting and the strength of each team in the past three months. The semi-finalists, according to UBS, will be Brazil, Germany, Italy and Netherlands with Brazil winning the trophy. Host South Africa is the only African team expected to qualify for the round of 16 although that looks unlikely considering their 3-0 drubbing by Uruguay on Wednesday.

Goldman Sachs expects Brazil to win with other close contenders for the title being Spain, Germany and England while Nigeria is the only African team expected to reach the round of 16.

Probably the most fascinating forecast came from JP Morgan’s quantitative analysts, or “quants”’ in short. Unlike UBS and GS models which focused only on semi-finalists, the JP Morgan nerds even predicted winners for each of the 64 matches to be played at the tournament. Brazil is the strongest team, they declared, but England will win the competition on account of easy fixtures. England is expected to beat Spain in the finals through penalty kicks because it has “an impressive score on our Penalty Shoot-out metric”. Netherlands will win the third place play off against Slovenia. Cameroon will be the only African team in the round of 16.

In coming up with these results, the analysts at JP Morgan said they used the same model they use for stock picking. Input data used included probabilities to win from bookmakers and exchanges, official Fifa team rankings and results from previous World Cup tournaments together with qualifying competitions. The model has successfully predicted eight out of 17 games played so far and this should be upsetting to the analysts.

On Wall Street quants are both esteemed and loathed. They are appreciated for transforming finance through blending it with computing and mathematics. They designed and implemented complex models that allowed financial firms to price and trade securities in order to boost profits. Financial engineers devised not only the exotic, mortgage-backed securities, but also the mathematical models of risk that suggested these securities were safe.

 

It is on the last aspect that they created enemies. They assured everyone that their models were foolproof and for a period it appeared correct. But their creations such as collateralised debt obligations are largely held responsible for initiating and precipitating the global financial crisis. However, the quants also blame the regulators along with other risk-takers and they are adamant that they are as much victims as other market players.

With this tainted reputation on their CVs it is difficult for anyone to believe their World Cup forecasts. Soccer is different from quantum physics or any of their crazy math stuff because it includes a lot of the human element in it. For instance, their models did not predict that England goalkeeper Robert Green would fumble a seemingly harmless shot into the net. In the end England drew with USA whereas JP Morgan expected the Three Lions to win that game.

While quants probably deserve their recognition for spearheading the evolution of finance, their skills may not be valuable in football prediction. On this one, their views are no different from many others from sangomas and other laypersons. It is a guesstimate that should not be taken as true prophecy. The winner will only be known on the July 11. In the meantime, let Africa enjoy its time in the sun.

 

By Ranga Makwata

Top