Euro 2016 Projections

by Thomas Negron

For soccer fans, this summer is about as good as it gets for a non-World Cup year. Just as the Copa America is getting interesting, the European Championships are starting today in France, with the home nation taking on Romania. In this post, we will try to mathematically answer the questions that soccer fans around Europe and around the world are asking. Will France be able to win the European championships on home soil, like they did in 1984 (we have them as favorites)? Who will be the biggest dark horse in this tournament (watch out for the Ukraine)? And the question on every Englishman’s lips, is it coming home (maybe! But probably not)

Similar to the Copa America model, this model uses Elo ratings for each team to predict the results of the tournament. The data comes from World Football Elo Ratings, which has Elo ratings for every national team in the world, and is compiled by Advance Satellite Consulting. For those not familiar with Elo, it is basically a rating system that changes based on results in every international match and we can use these ratings to form predictions. If you want a more thorough explanation, the World Football Elo Rating site is a great place to go.

The model specifications remain the same as the Copa model. France is given a boost of 100 Elo points as they are the host for the tournament. 10,000 simulations were run for each group to determine who would come in first, second, and third place. The Euros differ from Copa America in being a 24-team tournament, which means there are six groups of four with the top two in each group advancing to the round of 16. Then the 6 third place teams are ranked the same way as the groups (points, then goal difference, then goals scored) to determine which four teams will round out the round of 16. For this model, coming in third in a group means you have a 2/3 chance to advance. While the actual chance a group sends their third place team may depend on the quality of the teams in the group, this seems like a safe assumption. From here, team’s probabilities to advance to further knockout rounds are based on knockout round matchups and a team’s probability of beating any other team, just like with the Copa America model.

Screen Shot 2016-06-10 at 10.25.51 AM

 

As expected with an expanded field, the top teams in each group are expected to coast to the Round of 16. For many of them, the focus is solely on winning the group and guaranteeing a more favorable draw. France, who many are tipping as the favorites for this tournament, are heavy favorites in their group. The group to watch is definitely Group E, where Belgium and Italy will be fighting for that top spot, with both Sweden and Ireland hoping to play the role of spoiler. Interestingly, this model is not optimistic on Wales’s chance of advancing to the Round of 16. While they climbed as high as 8th in the FIFA rankings last year, their Elo rating remains rather low, and projects them as fairly large underdogs in their group.

Screen Shot 2016-06-10 at 10.27.48 AM

Our model agrees with bookmakers in making France the favorites for the tournament, helped a lot by their relatively easy group and their home field advantage. Defending World Cup champions in Germany are also favorites with defending Euro champion Spain hoping to prove a point after a dismal showing in the World Cup 2 years ago. The model also projects England at about 25% chance to make their first ever final in the European championships. After that it is a steep drop to Belgium, who along with Italy, have a tougher road to the quarterfinals, due to being in the group of death and how the round of 16 matchups are setup.

Again, it is important to note that this model does not take into account anything to do with the actual players selected for each team. For instance, this does not take into account that France will be without Karim Benzema, their starting forward, so France’s chances may be overstated. It is also slow to adjustment, so recent additions or retirements from national squads may not be entirely accounted for.

This model will be updated after each round of matches to give updates on probabilities, as well as winners and losers of each round, so keep coming back to the blog if you are interested in that. We may also put out tweets before or after matches to discuss any potential or actual changes in odds, based on matches, so make sure to follow us here. Enjoy the tournament.

 

About the author

harvardsports

View all posts

2 Comments

Leave a Reply to stevepeterson Cancel reply

Your email address will not be published. Required fields are marked *