25 Comments
User's avatar
jerry chen's avatar

Impressive. I also have built a prediction site. Please check. https://cup26matches.com/

behlanmol's avatar

Can you please share the dataset used for prediction? Or Maybe a guide for beginners to web scrap the data correctly.

Dom's avatar

did you apply your model against the last two tournaments?

I built a model myself and also had it applied to the last to WCs:

Also interesting: I added player's data to the simulation, which had meaningful impact on the predictions (let's see if for the better or the worse :)

Frank Andrade's avatar

Yes, it did well in 2022, but not in 2018. How about yours?

Dom's avatar

yes, same here, it got Argentina on 2nd in 2022, but France only on 5th in 2018.

And I came across this source (goalimpact, but paid data) and https://www.linkedin.com/posts/seidel_worldcup2026-footballanalytics-prediction-activity-7469662933204238337-xw-S

Dom's avatar

yes, I'd love to peek under the hood, how does the model actually work?

Frank Andrade's avatar

I just added the section at the end!

Jo Capel's avatar

I always wish for Spain to win, and maybe this is a good omen and they will! Excellent tool. Thanks, Frank. How many tokens did this use?

Frank Andrade's avatar

I can't tell. I used Opus 4.8 at max effort, though

Fabrice Talbot's avatar

Oh my god! I was just going to build my own prediction with Claude when I read this.

I’m using a different approach. Will share when ready.

I’m actually doing it because I’m betting on an app with my friends and I came last four years ago. Hope I can do better.

Don’t need Claude to pick the winner.

Allez les Bleus!

Frank Andrade's avatar

Nice! Feel free to share the link here to check it out

Jay's avatar

Without running any models, my prediction is that the final will be between a South American and a European team. That much seems fairly certain.

Whether it's France or Spain representing Europe doesn't make a huge difference in my view, as Argentina would likely start as favourites because they're better equipped to adapt to the conditions.

If Brazil make it beyond the quarter-finals, however, they'll have the momentum behind them. With Carlo Ancelotti in charge, they could have every chance of going all the way and winning the tournament.

MJ Alavi's avatar

How does it work?

Frank Andrade's avatar

I just added the section at the end!

Sarah's avatar

how does it work?!

Frank Andrade's avatar

I just added the section at the end! Didn't expect so many to be interested in the technical side

Mike SAD's avatar

Please let me know how this works 8

Frank Andrade's avatar

I just added the section at the end! Didn't expect so many to be interested in the technical side

Ivan Taslim's avatar

Im interest in how the model actually works

Frank Andrade's avatar

I just added that section

Austin Marton's avatar

The flaw in your model is you, the human. In the Q&A in question 2 I feel you've made some questionable decisions to guide Claude. As much as I would like England to win, I think France will beat Spain in their semi. Your model should somehow take into account Champions League form of the players in National teams as let's face it the best players in the World are in Europe playing champions League with French and English players being present in the the semi finalists and finalists. Your model should also somehow take into account historical grudges so as much as I love Messi, there is no way England will lose to Argentina, again. So France to beat England in the final 2-1.

Frank Andrade's avatar

Well, we'll see after the final on July 19th ;)

Austin Marton's avatar

Yes, and the flaw in my model is me :) ! And love being human, Regards, Claude