• 6/11/2021
  • Reading time 3 min.

Comprehensive statistical modeling using machine learning

EM forecast sees France as the favorite

After having been postponed by one year due to the COVID-19 pandemic, the 2020 European Championship begins today. The (statistical) favorite this time is France with a winning probability of 14.8 percent. This has been determined by an international research team from the Technical University of Munich (TUM), the Technical University of Dortmund, the University of Ghent (Belgium), the University of Innsbruck (Austria) and the University of Molde (Norway) with the help of machine learning.

Image of a soccer stadium. After having been postponed by one year due to the COVID-19 pandemic, the 2020 European Championship begins today. A research team's forecasting model sees France as the favorite. Pexels.de / Juan Salamanca
After having been postponed by one year due to the COVID-19 pandemic, the 2020 European Championship begins today. A research team's forecasting model sees France as the favorite.

In order to identify the most probable winner the entire European Championship was simulated 100,000 times using match results randomly generated by the model, match by match, following the tournament draw and all UEFA rules. 

The forecast combines several statistical models for the teams' playing strengths with information on team structure (such as market value, number of Champions League players, club match performance of individual players) and socioeconomic factors of the country of origin (population and gross domestic product).

This results in probabilities for the advancement of all teams in the individual tournament rounds and ultimately for the European Championship victory. The favorite this time is France with a probability of winning of 14.8 percent, followed by England (13.5) and Spain (12.3). Of course, the tournament can still hold surprises – this is also suggested by the relatively narrow gaps in the win probabilities at the top, as well as the already low probability of even the top nations.

Correct predictions in the past

So far, however, past forecasts have been quite successful: Prof. Achim Zeileis' Innsbruck model, which is based on adjusted odds from the betting providers, was already able to correctly predict the EURO final in 2008, as well as World and European Champions Spain in 2010 and 2012, among others.

This year, it will be part of the more comprehensive combined model developed by the teams led by Dr. Gunther Schauberger, researcher with Prof. Stefanie Klug, professor of epidemiology at the Technical University of Munich, Prof. Andreas Groll (TU Dortmund) and Prof. Christophe Ley (Ghent University), which even outperformed the betting providers' forecasting performance at the 2018 World Cup.

Sound statistical modelling

"I've been involved in these simulations since 2014 ahead of major tournaments such as the men's European or World Cup or the 2019 Women's World Cup," explains Dr. Schauberger. "We have steadily improved on the statistical side over time and have been able to add more and more new information and methods."

In the German group, the forecast sees a probability of 85.3 percent for both Germany and Portugal to make it to the round of 16. For France, that probability is slightly higher at 89.7 percent. Germany's 10.1 percent probability of becoming European champion is well below that of the favorites and on a par with Portugal.

"Based on the solid statistical modeling, it will be really exciting to see whether the forecasts come true and how the 2020 European Championship will turn out in the end," says Prof. Klug. "Of course, I hope that the statistics are wrong and Germany becomes European champion!"

Further information and links

The prognosis was developed by Dr. Gunther Schauberger (Chair for Epidemiology, TUM), Prof. Andreas Groll and Franziska Popp (both TU Dortmund), Prof. Christophe Ley and Hans Van Eetvelde (both University of Ghent/Belgium), Prof. Achim Zeileis (University of Innsbruck/Austria) and Prof. Lars Hvattum (University of Molde/Norway).

Technical University of Munich

Corporate Communications Center

Contacts to this article:

Dr. Gunther Schauberger
Chair of Epidemiology
Technical University of Munich
Georg-Brauchle-Ring 56, 80992 Munich, Germany
Tel.: +49 89 289 24955
E-Mail: gunther.schaubergerspam prevention@tum.de

Back to list

News about the topic

HSTS