A group of research scientists from the Universities of Dortmund, Munich, and Ghent tried to find the winner of this year’s World Cup using artificial intelligence. Using three different scientific methods and the statistics of World Cup history, artificial intelligence determined Spain as the favorite with a 17.8% probability after 100,000 (one hundred thousand) different simulations. Right behind them with a narrow margin of 17.1% is Germany, who is also the bookmakers’ favorite.

However, not everyone accepts this assessment. The financial firm Goldman Sachs used its own machine learning engines for the World Cup in Russia and found that Brazil would emerge victorious.

Of course, football, and sports in general, have a long history of upsetting both bookmakers and artificial intelligence (AI) models. It is very difficult to account for everything that can affect a team’s performance. Still, it is interesting to look at the statistical methods involved in predicting what will happen in Russia. On the evening of July 15, we will find out if the one hundred thousand artificial intelligence simulations were correct. Update: AI messed up 🙂

Source: https://www.sciencealert.com/ai-runs-100-000-russia-world-cup-simulations-to-pick-winner