A model built by researchers from the University of Liverpool’s Management School places England as second favourites to win Euro 2024 behind old rivals Germany.
Using the latest machine learning technologies, this state-of-the-art forecasting model can predict match results based on the quality of individual players and how they are likely to interact with each other on the pitch.
After running 10,000 simulations, Professor Ian McHale and Dr Benjamin Holmes from the Centre for Sports Business and Dr Kamila Zychaluk from the Department of Mathematical Sciences, have predicted the results of each match at the UEFA European Championship.
See the Euro 2024 prediction tree: Euro 2024 predictions
On the question of whether this time “it’s coming home”, the model offers a promising forecast for Gareth Southgate’s team, as second favourites after Germany, with a 22% probability of lifting the Henri Delaunay Cup.
In contrast, reigning champions Italy only have a 1% opportunity of winning the tournament, after a dramatic penalty final shootout win against England at Wembley in 2021.
According to the model, the most likely final is Germany vs England, with 51% and 42% chance of making it to the showdown match respectively, ahead of France, Portugal and The Netherlands.
The other home nation interest lies with Steve Clarke’s Scotland, who have a 59% chance of reaching the knockout stages.
Rule changes from Euro 2016 mean third place teams can make it out of their respective groups, and a spot in the last 16 would be a historical first for Scotland having failed to make the knockouts in each of their previous Euro and World Cup appearances.
The model, published in the International Journal of Forecasting, makes adjustments based on which players are on the pitch and updates predictions as the tournament progress, for example if players suffer injury or are suspended.
Professor Ian McHale said: “Having measured player performances over the last 10 years, the machine learning model takes into consideration the abilities of actual players on the pitch.
“Our model knows how good each player is, and how they will interact with teammates and opponents.
“While bookmakers make subjective adjustments to probabilities to account for changes in lineups, our model uses advanced performance metrics at the individual player level and machine learning to recalculate predictions.
“The end result is a forecasting tool that has been shown to beat the market.”