On November 23rd, Croatia and Morocco played their first match of the 2022 World Cup group stage against one another in Group F. Neither side could grab much advantage and the match ended 0-0. At the time, this draw fostered some doubt amongst soccer pundits that either team could escape the group stage. Now of course we know why they played such an evenly matched game — they represented two of the four best teams in the entire competition.
Now the 2022 World Cup semi-final draw is set and will commence Tuesday, December 13 with a match between Croatia and Argentina. The next day, Morocco plays France. The winners will meet in the World Cup final on Sunday.
Unanimous AI assesmbled a Swarm AI system consisting of fans of international soccer to predict winners of these matches in the final week of the tournament.
As the chart above shows, the Swarm liked Argentina more than the sportsbooks, who have France as a heavy favorite to win it all. The magnitude of the choice of El Albiceleste is even greater when you consider that the Swarm saw Argentina vs Croatia as being a close match, while finding that France would have a comparatively easier time winning their semifinal with the Moroccans.
Overall, the 2022 World Cup has been a rollercoaster of upsets and unpredictable results. One measure of just how difficult it has been to make predictions on the competition can be seen by examining the Mathematical Model computed at Oxford University.
The Oxford model shown is heavily based upon statistics and data and extended model runs. This model picked 9 of the 16 teams to make the knockout rounds. Both teams it predicted in the finals are already out – including Belgium, who never advanced past the group stage.
The Swarm AI predictions made on November 17 did better. It correctly picked 11 of 16 teams to advance past the group stage, and it still has one of its top two teams (Argentina) in play for the finals. This is fascinating when you consider that Swarm AI connected 20 sports fans together but used NO actual data about sports. On the other hand, Oxford used no humans, but processed massive amounts of sports data.
If you’re interested in more Swarm AI-generated sports picks, you can check out Sportspicker AI here.