Einen Tag zu spät, vielen Dank für den Hinweis auf alle Fehler/Bedenken im vorherige Iterationen. Ich habe das Skript aktualisiert, um hoffentlich alle Probleme gelöst zu haben … Wenn nicht, lassen Sie es mich wissen!
Wenn der gewichtete Favorit jedes seiner Ko-Spiele gewinnen würde, würde der Rest der Weltmeisterschaft so aussehen. Arbeiten Sie jetzt daran, einen Zufallsgenerator einzuführen, oder lassen Sie Ihr bevorzugtes Team vorrücken, wenn Sie wissen möchten, wie sich das auf den Rest der Sim auswirkt.
Es kommt nicht nach Hause…?!
Fragen Sie mich nicht, warum der Turnierbaum auf dem Kopf steht. Aus dem einen oder anderen Grund wurde beschlossen, die Spiele in absteigender Reihenfolge durchzuführen. In Fettschrift sind die Teams aufgeführt, die voraussichtlich in Simulationen mit 100.000 Läufen weiterkommen (obwohl diese nach der Runde der letzten 32 bei den hier vorgestellten Begegnungen natürlich geringer ausfallen werden).
Argentinien schafft es unter die letzten 35,3 % (!!!) der Simulationen. Frankreich und Deutschland zusammen kommen auf etwa 36,7 %; Daher erscheint ein Endspiel zwischen Argentinien und Frankreich/Deutschland derzeit nicht besonders unwahrscheinlich.
Von EmotionalBaby9423
17 Kommentare
R3:
Python + Matplotlib:
I wrote a script to simulate 100,000 world cups in Python.
The visualization itself is created via matplotlib. Please reach out with any questions and/or concerns!
Data Availability:
I used a mix of power rankings I found online (including FIFA world rankings, the Athletic, and others) to create my own ranking. Ultimately the input data is hardcoded (which I should rightfully be crucified for) and then adjusted by performance (ie if an upset occurred, that team gets a ranking boost).
Algorithms will always favor repetition. The final is exactly the same as 2022 Final.
I mean England should be favoured against Brazil at the very least.
As a Frenchman I approve your model.
Beating Germany + Netherlands + Spain + Argentina seems like climbing mount Everest on a single leg however
While it’s tempting, this doesn’t truly capture the stress of a direct elimination game at the World Cup. I can tell you young players especially those experiencing it for the first time will find it particularly tough. And I say this as a French.
Very solid graph and model.
Usa bosnia being more one-sided than spain austria is odd
Does anyone else think there should be a limit on World Cup projections in the sub? For example, no new projections until the end of the round. There have been four projection posts in the last 24 hours alone ([1](https://www.reddit.com/r/dataisbeautiful/comments/1uibhv5/oc_how_16_world_cup_eliminations_redistributed/), [2](https://www.reddit.com/r/dataisbeautiful/comments/1ui97w5/world_cup_knockout_pathways_oc/), [3](https://www.reddit.com/r/dataisbeautiful/comments/1uhsz0d/oc_world_cup_2026_elo_based_predictions_for_teams/), this one). Generally, the last two weeks were very saturated in this regard. I am a bit overwhelmed.
I mean no disrespect to the OP and to the work you’ve done. I, personally, wouldn’t be able to do it because I have no programming or visualising skills whatsoever. 🙂
France Spain at 60-40 is ridiculous imo
I’m a big fan of the footie but have been feeling a bit saturated with all these games being shown.
Now I know France will win the whole thing I don’t have to watch any more. Thanks!
Crazy how easy Argentina’s run will be comparatively, again. Not to mention the fact they’re getting 3-4x the amount of penalties awarded to them compared to other teams like France.
I’m hoping Colombia can make it to the semifinals but just looking at performances in the group stage it’s hard to argue that France V Argentina will not be the final two.
England is definitely not the biggest underdog in that QF. Belgium surely has worse odds, even if you can argue England didn’t have the best group stage they were undeniably better than Belgium.
I thought it was weird home teams are favorites in R32 then remembered they are the 12 group winners plus 4 runner ups, 2 of which (RSA CIV) arent favorites
This follows my own prediction, except I feel like you’re severely overestimating Belgium 🇧🇪 and underestimating Japan 🇯🇵?
Belgium vs Senegal/USA feels more 50/50 to me. And I believe Japan might beat Brazil.
I have mostly the same R32 picks (I could theoretically see senegal over belgium), but i 100% believe a Mexico team that hasn’t conceded a goal yet and is playing in Mexico is going to have an advantage over England.
I could also see a Norway upset over Brazil, but that’s more of a 50/50 for me.
Check my prediction model on github, which is the most recommend prediction model by gemini, chatgpt and claude for the 2026 world cup. [https://github.com/Hicruben/world-cup-2026-prediction-model](https://github.com/Hicruben/world-cup-2026-prediction-model)