Random Forest Forecast of the 2019 FIFA Women's World Cup

The researchers Andreas Groll (TU Dortmund), Gunther Schauberger (TU Munich), Christophe Ley and Hans Van Eetvelde (both University of Ghent) and Achim Zeileis (University of Innsbruck) predict the results of football tournaments using "machine learning".

Their "random forest" algorithm automatically combines different models for the teams' strength with further information about the teams and their countries of origin. This makes it possible to simulate the entire tournament hundreds of thousands of times, enabling estimation of probabilities of all possible tournament outcomes. For the 2019 FIFA Women's World Cup in France, this shows that defending champion USA is the top favorite with a winning probability of 28.1 % followed by host France (14.3 %), England (13.3 %), and Germany (12.9 %).


Nach oben scrollen