I wonder if my complete lack of interest for football can be brought to a truce, at least for the duration of the World Cup. So, when I learned (thanks to the launch of its Italian translation) of the existence of Soccernomics, a book that promises to connect football with statistics, I hoped that this could drive my interest towards the subject of 9 out of 10 water cooler discussions in the next few weeks.

A first doubt has come to my mind when I’ve read about their prediction for the finalists: Brazil and Serbia. I’d be thrilled, but it’s a curious prediction to be coming out of a formula that uses GDP (Serbia is 75th in the world, according to the IMF), population (86th) and experience. How was experience computed in this case? According to one of the authors, Serbia has been given a bonus for having been the backbone of the former Yugoslavian team – making me think that this is a largely subjective parameter. At the end of the day, a prediction itself – concealed as a numeric parameter.

In general, I believe that applying statistical analysis to football has two major problems:

– the outcome of a match depends on a low score, therefore random events might have a great impact on the result;

– actions are complex, and composed of a long series of small events: having a complete record of these would be necessary to even begin analysing their impact on the outcome of a match.

In conclusion, I’m ready to celebrate Serbia as a finalist but, if that would happen, I’d be willing to congratulate the authors for their ability to make predictions but not for the quality of their formula.

And, in my quest for a reason to be interested in the world cup, I’ll have to look elsewhere…