Predicting Euro 2024 (Quarter Finals)

Predicting Euro 2024 (Quarter Finals)

In the last weeks I have built 2 different models to predict the Group Stage results of Euro 2024. You can check the previous articles in my blog to understand how the model works. In this article I explained how the model predicts results. While in this article I explained how to predict the total goals.

Last week I published an article where I updated my model and predictions for both results and goals in the Round of 16.

Let’s look at the result to see how the model behaved in the Round of 16 and let’s see the model predictions for the Quarter Finals.

Round of 16 predictions (results and goals)

See below the table with the predicted and actual results and goals of all matches (after 90 minutes). Together I show the potential wins considering the odds from oddsportal.

Date Match Result Prediction (Win) Over/Under 2.5 (Odds)
29-06-24 Switzerland-Italy 2-0 1 (+2.45) Over (-1.00)
29-06-24 Germany-Denmark 2-0 1 (+0.62) Under (+0.80)
30-06-24 England-Slovakia 1-1 X (+3.29) Over (-1.00)
30-06-24 Spain-Georgia 4-1 X (-1.00) Under (-1.00)
01-07-24 France-Belgium 1-0 X (-1.00) Over (-1.00)
01-07-24 Portugal-Slovenia 0-0 X (+3.57) Under (+1.05)
02-07-24 Romania-Netherlands 0-3 X (-1.00) Under (-1.00)
02-07-24 Austria-Türki̇ye 1-2 2 (+3.64) Over (+0.90)
Total     5/8 (+10.57) 3/8 (-2.25)

The model correctly predicted some big results, like Switzerland and Turkey winning, and England and Portugal drawing. It failed to predict the victory of Spain and Netherlands, which were also priced very high.

Overall, we have 5 out of 8 correct results, a 62.5% accuracy. Simulating a betting strategy of 1 unit per match, this would have won 10.57 units.

In general, the model was not very accurate in predicting the goals scored, especially predicting Overs. This might be due to the fact that teams tend to risk less when playing to go through, which seems reasonable.

The total correct predictions are 3 out of 8, just a 37.5% accuracy. Simulating a betting strategy of 1 unit per match, this would have lost 2.25 units.

Let’s have a look at how the model performed in each round so far.

Matchday 1X2 prediction (Win) Goals prediction (Win)
Match 1 4/11 (-4.04) 5/11 (-0.65)
Match 2 7/12 (+5.38) 3/12 (-6.62)
Match 3 8/12 (+14.15) 5/12 (-2.39)
Round 16 5/8 (+10.57) 3/8 (-2.25)
Total 24/43 (+26.06) 16/43 (-11.91)

Again we can see that the model performs well in the 1X2 market, but it has a poor performance in the Over 2.5 market, where it loses against the bookies in every matchday.

Betting on both markets with a stake of 1 unit for each match, would bring a profit of 14.15 units. With a conservative approach of a 50 units bankroll, with 2% of it stacked on each bet (1 unit) this means an ROI of 28.3%. Insane ROI if you ask me. It means turning $1000 into $1280 in the span of 3 weeks.

Quarter Final predictions

For the QF predictions, we still take into account the Qualifiers as well as the Group and Round of 16 performances. But we tweak a bit the weights of the model, to give 65% importance to the Final stage and 35%% to the Qualifiers. These are the predictions for 1X2 and Over/Under 2.5 markets. All the predictions are after 90 minutes.

Date Match Prediction (Odds) Over/Under 2.5 (Odds)
05-07-24 Germany-Spain 1 (2.83) Over (2.07)
05-07-24 Portugal-France X (3.01) Over (2.53)
06-07-24 England-Switzerland X (2.99) Over (2.65)
06-07-24 Netherlands-Türkiye 1 (1.61) Under (2.08)

It turns out that these predictions are quite independent of the weights. Only the latest match could end up in any of the three results, if we tweak the weights in the right direction. I am (again) a bit perplexed by the Over/Under predictions. It seems especially unlikely that there will be three or more goals in the England or France game, give that these teams have shown a very poor scoring form in the latest matches. Let’s see.

All the code to run the simulation is available in my latest book, where I explain how to build a betting model for Euro 2024. I have also written a few books where I go into the details of how to get the data, visualize and train a model to predict football results for the Premier League, La Liga, Serie A, Bundesliga and the other major European national tournaments, complete with code examples.

Antonio
Antonio Author of Code a Soccer Betting model in a Weekend and Soccer Betting Coding
comments powered by Disqus