A Swiss university lab has harnessed the power of artificial intelligence to predict the outcome of matches during the Euro 2016 football tournament that begins on Friday. The researchers’ unique model uses individual players’ performance as a measure.
The website kickoff.aiexternal link was launched on June 7 by three PhD students at the Lausanne Federal Institute of Technology (EPFL) and is targeted at football lovers.
“All three of us are football fans and we wanted to apply machine learning to a new data set,” Victor Kristof, one of the brains behind kickoff.ai, told swissinfo.ch.
While several other football results prediction models exist, kickoff.ai incorporates two special features that help with better accuracy.
The first one is modelling individual players’ performances, which ensures more variables are incorporated when predicting an outcome of a match. Most traditional models only take into account the performance of the entire team and not that of its constituent team members separately.
The second feature is the use of Bayesian inference which is a statistical method to examine the probability of a hypothesis. It gives an estimate of the degree of confidence in a prediction by taking into account any uncertainty like teams that have never played each other before or a new player.
The team tested their model on 2012 Euro football tournament and found that the results were quite similar to the odds offered by bookmakers.
“They used the wisdom of the crowd while we crunched numbers instead,” says Kristof.
The trio, which also includes Lucas Maystre and Antonio González, plan to extend their model to every major football championship in the future.