Over the years, a number of methods have been concocted for predicting the winner of the most widely-watched sporting event in the world – the World Cup.
And this year is no different.
While the task of making such predictions has often fallen to statisticians, sports pundits or clairvoyant animals, researchers at two European universities have used machine learning to predict who will take home the Cup in Russia by simulating the tournament 100,000 times.
WATCH: World Cup’s “Egyptian King” hailed as soccer saviour
The group of researchers from the German Technische Universitat of Dortmund and the Technical University in Munich used several factors, including FIFA rankings, each country’s population and their gross domestic product (GDP), bookmaker’s odds, how many of the national team players play together in a club, the player’s average age and how many Champions Leagues they’ve won, reports Motherboard.
The results reveal that Spain, followed by Germany and Brazil, are most likely to take home the cup this year. The researcher’s methods arrived on Spain as the most likely winner by a margin of almost 20 per cent.
WATCH: Kevin Smith hangs out with World Cup Portugal fans in Calgary
However, should Germany find its way to the semifinals, they’d be more likely to take home the top prize. The researchers explain that this has to do with the structure of the tournament. While Germany is more likely to beat Spain in the semifinals, it has a much harder road to get to that point. Thus, it’s more likely that Germany will be knocked out beforehand.
“Spain is slightly favoured over Germany, mainly due to the fact that Germany has a comparatively high chance to drop out in the round-of-sixteen,” says Andreas Groll, of Technical University of Dortmund, in the study.
Therefore, at the beginning of the tournament, Spain has the best chance of going the distance, but if Germany makes it to the quarterfinals, they become the front-runner, the research states.
While these results are certainly interesting, the researchers themselves warn to take them with a grain of salt.
These researchers aren’t the only AI enthusiasts to take on the results of the World Cup this year. Goldman Sachs also used machine learning to forecast the results of the tournament, by running over 200,000 models.
Sachs used data such as player attributes to help predict specific match scores. Then, Goldman Sachs simulated over one million versions of the tournament.
“We are drawn to machine learning models because they can sift through a large number of possible explanatory variables to produce more accurate forecasts than conventional alternatives,” a group of strategists from Goldman’s international research team wrote in a client note, obtained by Business Insider.
According to this data, Brazil is most likely to win its sixth World Cup title in 2018, defeating Germany in the final round. Unlike the research conducted by Groll and his colleagues, Goldman Sachs predicts that both Spain and Argentina will underperform and lose in the quarterfinals.
WATCH: The World Cup kicks off as North America continues to celebrate 2026 win
Despite the precision involved in making these predictions, football is too unpredictable even for Goldman Sachs to place bets on, which is why the authors of the report included this disclaimer at the end.
“We capture the stochastic nature of the tournament carefully using state-of-the-art statistical methods and we consider a lot of information in doing so,” they said. “But the forecasts remain highly uncertain, even with the fanciest statistical techniques, simply because football is quite an unpredictable game. This is, of course, precisely why the World Cup will be so exciting to watch.”