The science behind a football match
In 2015, Midtjylland FC, the Danish team of Herning, won the league title for the first time in its history. It has been a significant event: it was the first prize they won using algorithms, statistics and Big Data.
Sports analytics combine performance indicators with scientific models provided by cognitive science and economic research, in order to help athletes, coaches, managers and referees improve their results. Sports analytics have been applied in several American sports. In particular, sabermetrics – which are baseball statistics, developed by the Society for American Baseball Research (SABR) - are widely used for baseball matches, as witnessed by Michael Lewis’ book Moneyball and by the movie starring Brad Pitt. However, football seems to be a different story.
“In football, the scientific method may be useful to create a set of indicators that the team can use to improve performance in an effective manner”, explains Luciano Canova, professor of Behavioural Economics: “But mathematics does not work miracles, because football is a very dynamic sport, with unexpected developments that make it very difficult to build mathematical models to forecast, for example, the final result of a match”. Canova and the philosopher Carlo Canepa are the authors of The science of goals, which describes the growing role of a quantitative approach to sports - especially football - with the ambition to “connect literature with scientific research”.
“Data”, Canova says, “can be useful to debunk many clichés, such as the belief that whoever scores near the end of the first time has a good chance to win the match, because it gives a psychological jolt to the opponents. This is refused by the data”. There is also another important aspect emphasized in the book: leagues are privileged laboratories to analyze facts and phenomena out of the playing field. In this sport the rules are the same for everyone and the performance is more objectively measurable than elsewhere, so it provides the perfect conditions for an experimental study. For example if I want to know the effect of inequality on productivity, or if a multicultural society is more cohesive than one with a single dominant ethnic group, what could be better than taking information directly from the A League? Sports in general are reliable sources from which you can draw information for real life. Statistics, therefore, are very important for the game itself, in order to improve performance and dispel prejudices, but also for society as a whole: the financial fair play, for example, has also been studied in relation to austerity policies. Accurate predictions on results are impossible. “There are too many elements that can’t be controlled”, says Canova: like weather, or social climate - for England, for example, hooligans and Brexit have had a strong impact -; not to mention the important role of a manager.
So, no miracles. “Because”, says Canova: “this is a sport - for me the most beautiful in the world - and the beauty of sports is that strange and amazing things happen all the time. As when Iceland beat England”.