Attempt to implement Sigames Football Manager simulation engine without manual control.
Live Demo - https://open-football.org
The demo simulation is quite resource-intensive, so I disabled the continue button.
If you want the full simulation, download and try it locally
The project is a simulation of soccer leagues and clubs, including match simulation.
The project is NOT a game, it is a simulation without the possibility of control.
The goal is to get as close as possible to a real soccer simulation and based on this data:
- Predict match results
- Predict the success of player transfers Example: send Kylian Mbappe to Italian Bari and run simulation for 5 years and watch results
What do I need to do?
- Make a simulation of the match
- Simulate the rest of soccer activities (training, transfers, etc.)
- Change the synthetic player database to a real one and run the simulation.
- Use the simulation
Currently available nation for simulation - England
- Local run
// run frontend (Angular)
cd ui
npm install --force
npm start
...
// run backend
cargo run
...
open chrome at http://localhost:18000- Run in Docker
cd open-football
docker build -f .\build\Football.Dockerfile -t open-football .
docker run -d -p 18000:18000 --name open-football open-football
open chrome at http://localhost:18000Match page example (click on any goals)
Player page example (click on any player)
/src/core - Core Rust app logic (including match)
/src/database - Simulation data source logic
/src/dev/graphics - Dev utils for instant match development (src/core/src/match)
Match dev looks like (cross-platform)
/src/dev/neural - Dev utils for training NN for using in match, etc
/src/neural - Core Burn neural network data /src/server - HTTP server for running API for Angular UI
/ui - Angular app that you can see in Live Demo
Apache License 2.0