World Cup predictions: Picking the winner in every game of the entire tournament - ESPN
Everyone is using artificial intelligence to do, well, everything. With the World Cup starting on June 11, you can't scroll for more than a couple of minutes without hitting another post or video or reel of someone telling you how they used AI to predict the World Cup.
So, I decided to use my own supercomputer to predict every game of the 2026 World Cup — the supercomputer is called «my brain.»
There will be 104 matches at the 2026 World Cup, and… OK, fine, I lied. I have predicted only 103 of them. I skipped the third-place game because self-care is important.
This is already the biggest World Cup ever, with the field having been expanded to 48 teams, which means we will have 38.5% more matches than we did four years ago. It's a lot.
With that, here is what happens when a human being predicts the outcome of every game of the 2026 World Cup.
Mexico: 1800 Elo rating (ranked 14th of 48), 95% chance of advancing
South Korea: 1754 rating (20th), 77% chance of advancing
Czechia: 1691 rating (31st), 60% chance of advancing
South Africa: 1526 rating (45th), 35% chance of advancing
The ratings and predictions you see above come from the DTAI Analytics Lab at the Catholic University of Leuven in Belgium. It's run by Jesse Davis, an American from Wisconsin, and the lab is consistently producing the most cutting-edge, public-facing analytics work in the soccer world. Every four years, it also tries to predict the World Cup, using a model that has outperformed bookmakers in previous tournaments.
But do not take that as betting advice! The model has outperformed the implied probabilities derived from bookmaker odds once you remove the vig (the sportsbook's commission), not the actual odds you would've been able to place a bet


