This repository includes a scrapy spider to build a consistent database about high elo League of Legends games, as well a simple algorithm do provide game predictions based on the team composition alone.
Just run the container:
docker compose up --build
Update your database:
scrapy crawl opgg -o db.json
The program currently uses a json file as database, there is a woking mongodb pipeline that can be used for that, feel free to send a pull request 😉
Then you just have to enter de worker container and run the predict script when everyone have selected their champions.
docker exec -i -t worker /bin/bash