There are many ways of approaching picking a winner, and you are sure to have your own theories. One way of doing so is to use historical data and statistics, and that is what we shall look at in this article.
Stats Perform (www.statsperform.com), a company dedicated to carrying out sports analytics using Opta Data, has used an AI model to do just this and came up with the following predictions:
- France – 20.5%
- Belgium – 15.7%
- Spain – 11.3%
- Germany – 9.8%
- Portugal 9.6%
- Italy – 7.6%
- Netherlands – 5.9%
- Denmark – 5.4%
- England – 5.2%
- Switzerland – 2.3%
- Sweden – 1.5%
- Croatia – 1%
- Russia – 1%
- Poland – 0.8%
- Ukraine – 0.8%
- Wales – 0.6%
- Turkey – 0.4%
- Czech Republic – 0.2%
- Austria – 0.2%
- Finland – 0.1%
- Hungary – 0.1%
- Scotland – 0.1%
- Slovakia – 0.04%
- North Macedonia – 0.02%
France to Be Crowned Champions?
As can be seen, the model gives France a roughly one in five chance of winning, which is not bad when there are 24 teams taking part. France will be hoping to become only the fourth team to win back-to-back World Cup and European Championships, something managed by just West Germany (Euro 1972, World Cup 1974), Spain (Euro 2008, World Cup 2010, Euro 2012) and France themselves (World Cup 1998, Euro 2000).
Furthermore, if France does manage it, it means that Didier Deschamps would become the first person ever to win the World Cup and European Championship as both a player and a manager. Furthermore, he is also the only manager at this year’s tournament to have won it as a player.
The model gives France a 46.8% chance of winning their group, which his extremely impressive considering that Group F has been labelled the “Group of Death”, containing France, Germany, Portugal and Hungary. By comparison, Spain has been given a 70.3% chance of winning Group E, the Netherlands have a 65.9% chance of winning Group C, England 64.6% in Group D, and Belgium are given a 63.5% chance of winning Group B.
Can Belgium Win Their First Trophy?
As you will have noticed, Belgium, managed by Roberto Martinez, are the second favourites according to the model. It is a truly talented team that had an excellent qualification campaign for the tournament, winning all ten of their games. If they were to lift the trophy, it would be their first-ever tournament title, with the exception of the Olympics. At present, Belgium is the only national team to top ranking without having won a World Cup of European Championship.
France and Belgium are the only European teams to have made it to the quarterfinals in all three of the last major tournaments, so it seems likely that both will make it to the later stages of the Euro 20. The model has given both teams a 63.9% or better chance of reaching the last eight.
From 2008 to 2012, Spain were on top of the footballing world, winning the World Cup in between two Euro titles. However, consecutive exits in the last 16 of the last two tournaments mean that the model has given the team just an 11.3% chance of winning. On the other hand, this still makes them third favourites, and their impressive 6-0 win against Germany last November should definitely not be forgotten.
Tackling the Group of Death
Portugal won the last Euro championships and they are coming to this tournament with another excellent team, and of course Cristiano Ronaldo. Ronaldo holds a number of Euro records, such as most appearances (21), and joint most goals (9), and he is sure to be determined to help his side win once again. However, the model does not appear to think it is very likely, placing them just behind Germany.
Much will depend on which teams make it out of the dreaded Group F. The model has given Portugal a 19.4% chance of topping the group, a 27.7% chance of finishing second, a 35% chance of third place, and a 17.9% chance of finishing bottom. This translates into an 82% chance of reaching the Round of 16, a 57.7% chance of making the quarterfinals, a 34.6% chance of reaching the semi-finals, and an 18% chance of making it to the finals.
Germany and Spain have the joint record for the most European Championship wins, three each. However, once again, being in Group F means that the model has given them a relatively small chance of winning it again this time round.
Even Italy and Netherlands are a long way behind the favourites, at 7.6% and 5.9% respectively, while Denmark and England, at 5.4% and 5.2% have not been given a very good chance at all, despite there being a lot of speculation about them in the press. Denmark did surprise everyone when they won the 1992 Euros, but they have only reached the knockout stages of the tournament once since then, when they made the quarterfinals in 2004.
What About England?
If you have been following the build up to the tournament in the news then you will have seen that England are thought to have an excellent chance. Indeed, here at mr.play Sports they are second favourites to win with odds of 5/1, behind Spain (9/2), and ahead of Belgium (6/1).
The model will have been considering a number of factors when it gave England just a 5.2% chance of winning. To begin with, England has played more games than any other team at the European Championship without ever reaching the final (a total of 31). The team has a good chance of winning their group this time around, but they have a very difficult path to an unlikely final.
Assuming England top their group, they would then face the second-placed team from Group F, which is likely to be France, Germany or Portugal. If England finish second, then they would probably play Spain, which goes a long way to explaining the model’s prediction.
The model makes use of conditional probabilities to examine what could happen in each round, and England have a high chance of facing Spain, Germany or France in the quarterfinals. If they were then to make it to the semi-finals the model has them facing Belgium, Italy or the Netherlands. In other words, after the group stage, every match is likely to be tough.
However, England does still have a far better chance of winning than many other teams, such as Switzerland, Sweden and Croatia. While Croatia may be the runners-up from the 2018 World Cup and have been given a 76.7% chance of reaching the Round of 16, from there on they too would face a number of extremely tough matches on their way to the final.
How The Modelling Works
The model used by Stats Perform estimates the likelihood of each match outcome using a mixture of betting market odds and Stat Perform’s own team rankings, which are based upon historical and recent team performances. The model assess the quality of the opposition and the difficulty of reaching the final by using the match outcome probabilities with the composition of the groups and the seedings going into the knockout stages. It then went on to simulate the remainder of the tournament 40,000 times. The model analysed the outcome of each simulation and then returned each team’s chances of progressing at each stage of the tournament.
Stats Perform is not the only place to have used statistics and computer simulations to predict the outcome of the tournament. A research team together with the University of Innsbruck in Austria did something very similar, simulating the tournament 100,000 times. This model also placed France as favourites to win with a 14.8% chance. However, there things diverge, as France is followed by England (13.5), Spain (12.3%), and then Germany and Portugal on an equal footing (10.1%).
This model actually predicts a far more competitive tournament than Stats Perform’s does. Furthermore, the university has used modelling to predict the winner of the tournament correctly a number of times in the past. However, the researchers stress that even if they give France a close to 15% chance of winning, it still means that there is an 85% of them losing.
Ultimately, computer modelling and statistics can give fans an insight into the upcoming tournament, but the results should never be treated as certainty. Nonetheless, you can use them to help plan your betting strategy, and as the tournament progresses, it will be interesting to see how the models hold up.