University of Oxford World Cup Round 16 AI Prediction: Korea is not there: Dong-A Science

The results of a study using artificial intelligence (AI) at Oxford University in the UK to predict Belgium’s victory at the 2022 World Cup in Qatar have been published.

According to the international journal Nature on the 22nd (local time), as a result of predictions using mathematical modeling based on AI developed by Oxford University epidemiologists, Belgium was most likely to win. Brazil followed suit.

The AI-based mathematical model used in this prediction analyzes data such as points scored and conceded in similar situations in previous matches in each country. Based on this, we evaluate each team’s offense and defense to predict the chances of winning. A ‘double Poisson model’ is applied to predict the situation by calculating the probability of a specific event based on past data.

Oxford University researchers had previously correctly predicted that Italy would defeat England at Euro 2020 using a double Poisson model. Six of the countries that progressed to the quarter-finals also made it.

According to the University of Oxford, France had the highest chance of winning this year’s World Cup in Qatar, after Belgium and Brazil. It was followed by Argentina, the Netherlands, Germany, Spain, Switzerland, Portugal and Uruguay. Portugal and Uruguay are in Group H, along with South Korea. Denmark, England, Poland, Croatia and Mexico were also nominated for the round of 16.

The importance of data is gradually increasing in AI models that predict the outcome of sports matches. Real football players wear wearable vests in training and calculate all their movements as data. Famous football teams employ mathematicians or experts to carefully analyze the collected data.

The results obtained from data analysis are also changing the way games are played. Strikers shoot from less distance than before and wingers make short passes instead of crosses. Coaches will focus on increasing possession.

Nell Memert, a researcher at the Cologne Sports University in Germany, said, “Big data has opened a new era in football,” and found, “It has changed the philosophy and behavior of teams, the way they analyze opponents, and the the way they scout players.”

Data and statistics are actively used not only in football but also in other sports. Michael Lewis, in his 2003 book Moneyball, detailed how Billy Beane, general manager of the Oakland Athletics, an American major league team, relied on statistics to win the league on a shoestring budget in 2002. General manager Bean is famous for devising a new strategy by paying attention to the batsman’s on-base rate.