Euro 2024: we asked ChatGPT to predict 7 Euro games - here's how well it did

  • We asked ChatGPT to predict results for first seven games at Euro 2024.
  • It also tried to figure out who would score the first goal in each of the games.
  • Predictions included Scotland’s opening game and England’s encounter with Serbia.

A severe case of Euro fever broke out across the continent this weekend as tournament action kicked-off in Germany. England, Spain, Germany, Italy and Scotland all started their campaigns over the first few days - some more successfully than others. 

It was a full action slate - at least for the first five or so games - and it took until Sunday (16 June) evening before there was a match with less than at least three goals in it. We asked ChatGPT to attempt to predict the results - and the first goalscorers - for the first seven matches of the tournament.

Hide Ad
Hide Ad

But with so many goals flying in from Munich to Stuttgart, how well did the AI-tool do? Let’s find out.

Game 1: Scotland vs Germany 

The opening night of the tournament kicked-off in Munich on Friday with a thumping 5-1 win for the hosts Germany against Scotland. In the build up to the game we asked ChatGPT to predict the result (link to article) and here’s what it said: “Overall, given their form, squad strength, and home advantage, Germany is expected to secure a comfortable win over Scotland in this opening match of Euro 2024.”

It turned out to be very prescient - as it was indeed a comfortable win for the hosts. However ChatGPT did back German striker Niclas Füllkrug to open the scoring but in fact it was Florian Wirtz who scored the first goal of Euro 2024 - but Füllkrug did grab the fourth. 

The match prediction was correct - but the first goalscorer was wrong. 

Game 2: Hungary vs Switzerland 

The Euro 2024 action resumed on Saturday lunchtime as Switzerland took on Hungary in Cologne. ChatGPT had a hard time with this game and described it as “challenging” to pick a winner. 

The AI-tool elaborated: “Overall, while it is difficult to predict a clear winner, the match is expected to be competitive with possibilities for either team to shine.” The Swiss won the game 3-1. 

ChatGPT backed Dominik Szoboszlai for Hungary and Granit Xhaka or Xherdan Shaqiri for Switzerland to be the first goalscorer. But it was in fact Switzerland’s Kwadwo Duah who opened the scoring. 

The match prediction was incorrect (as it didn’t predict a winner) - and the first goalscorer was wrong as well. 

Hide Ad
Hide Ad
ChatGPT tries to predict first seven games at Euro 2024. Photos by Getty ImagesChatGPT tries to predict first seven games at Euro 2024. Photos by Getty Images
ChatGPT tries to predict first seven games at Euro 2024. Photos by Getty Images | Getty Images

Game 3: Spain vs Croatia 

It was one of the pre-tournament favourites Spain who were in action next, facing Croatia (a repeat of the round-of-16 clash from Euro 2020). ChatGPT predicted: “Given the competitive nature of both teams, a draw seems a likely result, but either side has the capability to snatch a win.” 

Spain ran out 3-0 winners on the day, with all the goals coming in the first half. The AI-tool also backed Alvaro Morata to open the scoring for Spain - and that did come true. 

The match prediction was correct (but it was a bit of an on-the-fence prediction) - and for the first time ChatGPT got the first goalscorer correct.

Game 4: Italy vs Albania 

The final match of Day 2 at the Euros saw holders Italy kick-off their campaign against Albania, who topped their group in qualifying. ChatGPT was bullish in its prediction in favour of the Azzurri, writing: “In summary, Italy is the favourite to win, with a predicted scoreline ranging from 1-0 to 2-0 in their favour.”

It was a tricky game for Italy but after falling 1-0 down, they came back to claim a 2-1 victory in Dortmund. ChatGPT also backed Gianluca Scamacca to open the scoring - but it was Nedim Bajrami who scored first. 

The match prediction was correct - but the first goalscorer was wrong. 

Game 5: Poland vs Netherlands 

The first match on Day 3 of Euro 2024 started with Poland facing the Netherlands in Hamburg. ChatGPT predicted that the Dutch would run out as winners, writing: “The predictions generally favour a victory for the Netherlands, with some suggesting a scoreline of 2-1 in favour of the Dutch​.” 

It managed to correctly predict not only the result but also the final score. Holland came from behind to win 2-1 thanks to Wout Weghorst’s late winner. 

Hide Ad
Hide Ad

Memphis Depay and Adam Buksa were backed as potential first goalscorers. And Buska did indeed open the scoring on Sunday afternoon. 

The match prediction was correct - and ChatGPT got the right score. First goalscorer was also correct.

Game 6: Slovenia vs Denmark 

The middle game of the day provided the first draw of the tournament as Slovenia and Denmark shared the points at the Stuttgart Arena. ChatGPT had backed the Danes to get the job done, writing: “Denmark is favoured to win against Slovenia in their Euro 2024 opener.” 

Denmark’s Rasmus Hojland was backed by the AI-tool to be the first goalscorer, but it was in-fact Christian Eriksen who opened the scoring. His goal game three years after he collapsed on the opening weekend of Euro 2020 (held in 2021). 

The match prediction was wrong - and the first goalscorer was wrong. 

Game 7: England vs Serbia 

The final game on Day 3 of the Euros saw the Three Lions kick-start their campaign with a win. Jude Bellingham’s goal in the first half sealed a 1-0 over Serbia. 

ChatGPT had backed England, saying: “The consensus among multiple sources is that England has a stronger team and is more likely to secure a victory.” 

The AI-tool also had Harry Kane and Dusan Vlahovic as candidates to score the first goal in the game. However it was Bellingham that scored the first - and only - goal of the game. 

Hide Ad
Hide Ad

The match prediction was correct - but the first goalscorer was wrong. 

So how did ChatGPT do overall? 

Out of the 14 predictions for the opening weekend of Euro 2024, ChatGPT managed to get 7 (and a half right, extra half point for correct score in Poland vs Netherlands) correct, so just over 50% right. Not a bad batting average. 

The excellent Not the Top 20 podcast has changed focus from the EFL to the Euros for the summer and had an instant reaction pod following England’s win over Serbia. If you were a bit worried by the Three Lions performance, the episode could help calm the nerves a bit - you can find it on Spotify

Related topics:

Comments

 0 comments

Want to join the conversation? Please or to comment on this article.