During the World Cup in 2010, an octopus called Paul picked the winning team in 12 out of 14 football matches, which also included the final match. However, eight years later an octopus is no longer used, but replaced with an Artificial Intelligence algorithm designed to predict who will win the football cup 2018.
This year’s World Cup is to be held in Russia and football fans from around the world are impatiently looking forward to the matches and who will win the football cup 2018. A team of researchers from Belgium and Germany collaborated to predict the course of matches before they actually occur. The researchers from the German Technische Universitat of Dortmund, the Technical University in Munich, and the Ghent University in Belgium created a machine learning based algorithm which takes several factors into account, including the FIFA rankings, each country’s population and their gross domestic product (GDP), bookmarkers’ odds, as well as factors like how many of the national team players are playing together in the club, the average age of the players and finally, the Champions League trophies won.
The team collected all these factors together and taught it to the machine learning algorithm, which then simulated the soccer tournament 100,000 times. The algorithm, during its simulation calculated each team’s probabilities to pass to advance beyond the group stage, and then pass each next round on up to the finals. Spain was the most likely candidate to win, followed by Germany.
The researchers also told Motherboard that with “the myriad of possible constellations this exact tournament course is still extremely unlikely.” Also, the machine learning system from the researchers couldn’t predict that there would be the possibility of Spain’s coach being fired and replaced with another one only two days before the team starts playing, which caused a huge uproar among the players and fans in the country.
The tables showing the team’s odds of winning can be found on Motherboard’s website if you want to check it out.