the sheet to
It wasn’t magic, it was method: this is how we predicted Spain’s World Cup victory two days before the tournament began
How a prediction written two days before the World Cup turned out to be correct — and what this tells us about the right way to predict football
20 JUL 26
Translated by AI
Photo: LaPresse
On 9 June, we wrote that the World Cup would be won by the team “that knows what to do with the ball even when it’s heavy”. It wasn’t clairvoyance: it was a prediction based on separating talent from method, and suggestion from probability. On 9 June, two days before the World Cup began, we wrote a sentence that today has the advantage of being verifiable: “Spain will win the 2026 World Cup.” We had also predicted a Spain–France final, so let’s start with the part that dispels any myth: we correctly predicted the winner but got the other finalist wrong. France were knocked out in the semi-finals, Argentina made it all the way to the final, and yesterday Spain beat them. The prediction was correct. It wasn’t perfect. How did we manage it? By not trying to predict every result. Football is too unpredictable to be treated like an equation: an injury, a penalty, a deflection can destroy even the most sophisticated prediction in ten seconds. The real task wasn’t to say who would score the decisive goal, but to understand which team had the best tools to navigate the tournament without having to reinvent itself every time. The first criterion had been to distinguish the strongest team from the one best suited to winning. France perhaps had the most complete squad, Argentina had experience and Messi, England had talent, and Brazil had Ancelotti and unpredictability. But all of them needed the World Cup to take the form most favourable to their identity. Spain, on the other hand, could change its style without changing its identity. This was the crux of the prediction. Spain was no longer a caricature of tiki-taka, possession turned into an aesthetic religion. It had retained control whilst adding verticality, pace down the flanks and pressing after losing the ball. Yamal was the marvel, Nico Williams the burst of pace, the midfield the structure, but the real advantage lay in the collective: every player seemed to know the next move. The second criterion had been to reduce dependence on the ‘exception’. The most enticing national teams often wait for the star player to solve the problem. Spain, by contrast, thrived on distributed solutions: it could win through passing, through high pressing, through a winger, through a run into the box, or through a substitution. It was not merely a team with better players. It was a team in which the players seemed to become better thanks to the system. The World Cup demonstrated this theory almost as a lesson. Opponents often had a star player capable of turning the impossible into the possible. Spain had a structure capable of preventing the impossible from even arising. Not the miracle of the soloist, but the depth of the system. Not a team condemned to always win in the same way, but a team capable of finding a different way each time to achieve the same result. The third criterion was to remove from the prediction what often makes predictions foolish: fan bias, nostalgia, the allure of history. You see Messi and you imagine the final triumph. You see Brazil and you think of a resurgence. Artificial intelligence is not immune to errors, but it can be less susceptible to the narrative we wish to see and more attentive to the structure we observe. This does not mean that we knew Spain would win. It means that, amongst many possibilities, we had found the Spanish scenario to be the most robust. The difference is important: guessing is a result; predicting is a method. This time, the method and the result coincided. Perhaps because Spain won the World Cup in exactly the way we had tried to predict: not by shouting louder than the others, but by filtering out more noise than the others.