English football fans are hoping their team will return home to win the World Cup, but a supercomputer predicts otherwise, indicating that Brazil is the favorite to win the 2022 World Cup.
And if you want to bet, the supercomputer predicts that Brazil will play Argentina in the final on December 18 – but be careful, because a similar prediction for the outcome of the 2018 World Cup was wrong.
This also led to the selection of five-time champion Brazil to win, as France emerged victorious by defeating Croatia in Moscow.
An international team of researchers made predictions for the 2022 FIFA World Cup using a hybrid model that combines data from three advanced statistical models.
After being knocked out in the quarter-finals four years ago, Brazil were once again favorites with a 15% chance of winning. It is followed by Argentina, the Netherlands, Germany and France, according to a panel of experts from the universities of Innsbruck, Ghent and Luxembourg, as well as the technical universities of Dortmund and Munich.
The forecast combines several statistical models of playing strength with information about team structure (eg market value or number of Champions League players) as well as socio-economic factors of the country of origin (population or GDP).
“For scientific reasons, we decided to use our own machine learning approach, which we have successfully used in previous tournaments, to predict probabilities,” said Achim Zillis from the Department of Statistics at the University of Innsbruck.
With the values expected from the researchers’ model, the entire World Cup was simulated 100,000 times match after match.
This resulted in all teams advancing to various rounds of tournaments and eventually winning the World Championship.
Of course, the championship is far from predetermined, which is reflected in the relatively low probability of winning the top teams.
The probability of winning Argentina is 11.2%, the Netherlands – 9.7%, Germany – 9.2% and France – 9.1%.
The 2022 World Cup seems interesting to researchers from a scientific point of view because of the unusual date – the tournament had to be postponed to the winter months due to extremely high temperatures in Qatar in the summer.
“During the winter months, all major football leagues in Europe and South America are forced to interrupt their regular matches in order to adjust to the tournament,” said Achim Zillis.
This gives national teams less time to prepare and less time for players to recover before and after the World Cup. Combined with harsh weather conditions, this also increases the risk of injury.
So having a team with many players in international competitions like the Champions League and Europa League this year could do more harm than good.
The researchers’ calculations rely on four sources of information: the first two are a statistical model of each team’s playing strength based on all international matches over the past eight years, and another statistical model of teams’ playing strength based on betting odds. from 28 international bookmakers. As well as additional information about the teams, such as the market value and their countries of origin, such as population, is the third, and the fourth is a machine learning model that combines different sources and improves them step by step.
Previously, researchers trained the model on historical data.
Source: Daily Mail