Who can win the Cup? Statisticians update each team's title chances After the elimination of the Brazilian team, the hexa's dream was until 2030. However, the biggest event in world football continues to mobilize fans and arouse curiosity as to who will be this year's World Cup champion. With the end of the round of 16, which country has the best chance of winning? The interactive platform 'Sports Forecast', made by statisticians from five universities, including the University of São Paulo (USP) and the Federal University of São Carlos (UFSCar), together with a company, updated the results for the final stretch of the competition. ? Follow g1 São Carlos and Araraquara on Instagram Before the start of the competition, statistical data showed that Brazil, Germany and Portugal - teams considered strong - were among the 10 most likely to become champions. However, all three teams have already been eliminated. Among the eight countries that advanced to the quarter-finals: Morocco, France, Norway, England, Spain, Belgium, Argentina and Switzerland, the Sports Forecast indicates that the French team has a 33.5% chance of being this year's World Cup champion (see below for the complete list). France MAGN IMAGES via Reuters/Vincent Carchietta ? See the countries with chances of winning * According to the Bayesian edition - alternative probabilistic estimates using advanced Bayesian modeling: France - 29% Argentina - 26.6% Spain - 12.7% England - 11.9% Norway - 10.9% Switzerland - 6.5% Morocco - 1.7% Belgium - 0.7% Already considering the model strength, which are estimates based on the historical FIFA Elo ranking and classic strength simulation: France - 33.5% Spain - 19% Argentina - 18.4% England - 15.3% Norway - 6% Morocco - 3% Belgium - 2.5% Switzerland - 2.3% * The above predictions were made before the start of the World Cup quarterfinals. However, the models available through Sports Forecasting are dynamic and, as games take place, the real results are inserted into the system, updating the probabilities. SEE UPDATES. Erling Haaland of Norway and his teammates celebrate Norway's qualification to the World Cup quarter-finals after the match. Dylan Martinez / Reuters Methods used ? The two models, the Bayesian and the force, use statistical methods to estimate how many goals each team is likely to score in a match, but the main difference between them is in the information used to make the predictions. Bayesian model: combines objective indicators, such as the FIFA ranking score, and subjective indicators, such as experts' guesses about the scores of the games to be played. Strength model: calculates the potential of each team exclusively based on objective indicators, such as the FIFA ranking, Elo ranking, recent performance of the team, history in World Cups, host status and market value of the squad. Sports Forecasting Project ? Sports Forecasting Website Reproduction Francisco Louzada, professor at the Institute of Mathematical and Computer Sciences (ICMC) at the University of São Paulo (USP), in São Carlos (SP), explained that the project was consolidated in the 2010 World Cup, although researchers had already been carrying out studies since 2006. "What has changed drastically since then is the power of our structure. If data is the fuel, today it is much richer and detailed, and computers are the engine, now infinitely faster, our statistical modeling is the GPS that guides all of this", he stated. Louzada said that in 2010 the team worked with the equivalent of a paper map, which was useful but static. Now, they operate with an intelligent GPS that recalculates the route for each game, learning from what happens at each stage to show the most likely path to the title. "This evolution is vital for the 2026 World Cup, as the new 48-team format required a total recalibration of our route. What remains unchanged is our essence: combining academic rigor with the national passion for football," he said. Football - 2026 FIFA World Cup - Germany x Curacao - Houston Stadium, Houston, Texas, USA - June 14, 2026. Reuters How do the calculations work? ? According to Louzada, algorithms based on Bayesian Inference and Monte Carlo Simulations are used. "The model does not try to 'guess' the score, it works by studying the attack and defense strength of each team to calculate the chances. Furthermore, we have the simulation structure. We run the entire tournament thousands of times on the computer. If Brazil wins the World Cup in 150 thousand out of 1 million simulations, we say that it has a 15% chance of winning the title", he said. Louzada prefers the Bayesian methodology, which works with an objective approach to data and history, and a subjective approach with the opinion of experts. "It allows you to combine these two sources of information. It is a structure that 'learns' from the tournament: we start with an initial belief and update it as new data appears, dealing much better with the uncertainties of a World Cup", he explained. The project is a collaboration between researchers from UFBA, USP, UFSCar, UFRJ, UFMT and Neoma Business School (France). In addition to Louzada, the team includes Adriano Kamimura Suzuki, Diego Nascimento, Fernando Moraes, Lilia Costa, Nailton Santos and Paulo Henrique Ferreira, among others. Analysis of the results ? According to Louzada, the results for this year's world tournament, before the start of the competition, showed that Spain, France, England and Portugal are among the favorite countries to win the World Cup. According to him, due to the depth of the squad and recent performance. "Brazil appears in the elite squad, floating between 5th and 6th position together with Argentina. Notoriously, the first four enter the tournament with odds based on historical and technical performances, as is the case with other teams, such as Brazil, Argentina and Germany, for example, which appear with a slightly lower intensity", he said. Spain National Team Reproduction However, the professor highlighted that, during the championship, at each stage, with the original data on the results of the competition, in addition to the help of the modeling used in the project, the results increasingly come closer to those of the true champion teams. "In statistics, 'hit' means that the events occurred within the predicted probabilities. In 2010, our model was awarded nationally for its accuracy. In the 2014, 2018 and 2022 World Cups, the champion was always among our four favorites", said Louzada. World Cup ? The 2026 World Cup started on June 11th. This edition is the first in history to be held in three countries — the United States, Mexico and Canada. In total, 16 cities will host tournament matches, the majority of which are in the USA. Who can win the Cup? Statisticians calculate selections with the most chances REVIEW EPTV VIDEOS: See more news from the region on g1 São Carlos and Araraquara