A country's chances of making it to the World Cup improve dramatically if it is situated within a thriving footballing region
Every couple of years major global corporations, usually from the accounting or financial services sectors, roll out predictive models to forecast outcomes of major sporting events. The two events that most often fall victim to these ‘predictions’ are the Olympics and the Fifa World Cup. These predictions are usually reported with great alacrity by the business press, before fading away into little more than curiosities as the events take place.
Last month, UBS published their own statistical predictions using “econometric tools, usually applied to assess investment opportunities". And their predictions? That Germany, Brazil and Spain are the favourites to lift the cup. Yawn.
The FT report on the UBS forecast concludes with this wry statement: “Stay tuned for a flurry of other banks and investors also placing their bets in the coming weeks. Spoiler alert: They will all say Germany or Brazil."
This might vindicate the wisdom of football crowds, and Gary Lineker. (Lineker once said that football is a simple game in which 22 men chase a ball for 90 minutes and at the end, the Germans win.) But it reveals perhaps the biggest problem with these predictive models. And that is the predictability of their predictions. Yes, the host country will always do well in an Olympics. And yes, your statistical model is entirely useless if it doesn’t pick Germany, Spain or Brazil to lift the cup. So are these models useless? Not entirely.
Perhaps a more interesting use of these models, especially for India, is if they could be used to predict the nations that will qualify for the tournament.
In June 2014 PwC published The PwC World Cup Index: what can the dismal science tell us about the beautiful game? in which the stat boffins at the firm used “econometric analysis to test the correlation between several variables and World Cup performance". While they forecast Brazil and Germany as favourites to win the 2014 tournament—gasp—the variables they identified as significant determinants of world cup success are well worth mulling over:
• The number of football players available in the country
• Average attendance of domestic top division football matches
• How many times the country has bid to host the World Cup
• Whether the country is from Europe or South America
• Whether the country is the home nation or from the same continent
• Recent form
What if these metrics could determine not just performance in the World Cup, but a nation’s chances at merely qualifying for a tournament?
Now look again at these metrics through this prism of qualification, and then a world map of World Cup appearances, and suddenly an important determinant becomes clear: location.
There are arguably three great hubs of world cup qualification: South America, West Africa, and Europe. Surrounding these are a number of tendrils of footballing merit: North and Central America, North Africa, Korea and Japan.
Why should location matter? Glance at the PWC metrics again. If you are a footballing nation adjacent to other good footballing nations, chances are that you have a strong culture, a good marketplace of players, and therefore a better chance of running a good quality top division with high attendance.
The Japan-South Korea relationship is a case in point. Korea provides the largest contingent of foreign players to the J-League along with Brazil. Their relatively high standard, resources and proximity allowed them to co-host a tournament, in which Korea came fourth.
In other words if the PwC metrics hold true, an Indian team at the football world cup may be at least partly dependent on the other teams in South Asia pulling up their socks. Sunil Chhetri may have to record many more videos in more languages.