After 36 years the FIFA World Cup competition returns to South America. And as in all previous South American FIFA World Cup events, a South American team is again expected to take home the victory – this time the host Brazil. These are the results of a study carried out by statistician Prof. Achim Zeileis from the University of Innsbruck and his two colleagues Dr. Christoph Leitner and Prof. Kurt Hornik from the Vienna University of Economics and Business (WU Vienna). The scientists have applied a model that proved reliable in forecasting the results of the 2008 and 2012 EURO events and the 2010 FIFA World Cup: the bookmaker consensus model. To forecast the winner, they obtained long-term winning odds of 22 online bookmakers, which in combination with complex statistical models allow for the simulation of all possible courses of the tournament and results. According to this model, Brazil is the most likely to win the tournament with a probability of 22.5 percent. The odds to win the tournament for the three runner-ups are comparably high but the winning probability is clearly lower than the host’s: Argentina 15.8 percent, Germany 13.4 percent and Spain 11.8 percent. All other competitors have considerably lower winning probabilities, with Belgium being the insider tip of the ‘best of the rest’ with a predicted 4.8 percent winning probability. The bookmaker model can also predict the likely teams for the final game, however, the uncertainty is much higher here: The most probable final is a match between Brazil and its neighbour Argentina (6.5 percent). “We not only model the winner but also every possible course of the tournament. According to these models, Brazil emerges as the only team with 80 percent probability to beat almost any other team in the tournament. We are able to determine the probabilities of each team in the eight groups to proceed to the next stage in the tournament, and, eventually, the two teams in the final,” explains Achim Zeileis.
Statistical expertise and bookmaker knowledge
“Bookmakers base their odds on the most likely results. As experts they not only take historical data into account but also short-term events such as injuries,” explains the statistician. The odds are assigned in such a way that they predict the real results as accurately as possible on the one hand but also ensure profit for the bookmakers on the other hand. This system is an excellent basis for the model developed by Prof. Zeileis, Dr. Christoph Leitner and Prof. Kurt Hornik. “However, the quoted odds of the bookmakers have to be adjusted for profit margins. After that we can derive probabilities.”
In two further steps the statisticians determine the probable winner: The quoted winning odds show general winning probabilities for each team. In the next step, the scientists calculate the probability of each team to win against another team. “The tournament schedule was already known at the time the bookmakers calculated their odds, and, therefore, the winning probability of each team within its group has to be taken into account,” says Zeileis. Together with the expectations of the bookmakers, the pairwise winning probabilities are added to the computer model, which then runs a simulation of every possible course of the tournament. “Compared to other models, ours has the advantage that it also yields ‘survival’ probabilities of each team over the course of the tournament,” explains Zeileis. “However, we are still a long way from predicting the outcome with 100 percent certainty,” adds Zeileis. Thus, we can still look forward to an exciting football tournament until the final whistle blow in Rio de Janeiro on July 13th.