Model Predicting who will be the winner of FIFA World Cup 2022
In this notebook I implemented a model that predicts who will be the winner of the 2022 FIFA World Cup.
It includes:
- Calculating xG of a team by avering their score in the latest matches
- Adjusting the xG of a team according to the previous opponent's ranking
- Adjusting the xG of a team according with who they are playing in the next match
- Adding time decay (latest matches have heavier weights)
- Simulating matches based on the xG of two teams
- Implementing the whole structure of the tournament (group stage and knockout stage)
Model is fully modifiable, that means:
- We can use that model in future tournaments - we only need to download the dataset with newest matches and ranking
- We can predict the winner of one match - with or without draws
- We can change parameters like date from which matches are taken into account
- We can modify the amount of iterations
We can simulate the whole tournament:
Or just the knockout stage:
Or just one match:
In the notebooks everything is clearly explained. If you have any questions don't hesitate to ask.
Used datasets:
- Match Results - https://www.kaggle.com/datasets/martj42/international-football-results-from-1872-to-2017
- Elo Rating - https://eloratings.net/