Made by Krshs90
ReadMe By Gemini 3.1
Bugfixed by Google Antigravity
WorldCupML is an AI-powered desktop application built in Python that accurately predicts FIFA World Cup match outcomes. It fetches live tournament data and compares it against decades of historical match results, running complex machine learning algorithms and millions of mathematical simulations to forecast match winners and exact scorelines.
- Real-Time Match Data: Pulls live tournament fixtures, current live scores, detailed team lineups (starting XI vs subs), and group standings straight from the ESPN API.
- XGBoost Machine Learning: Dynamically calculates Expected Goals (xG) by engineering advanced statistical features like Elo ratings, recent team form, and venue context.
- Monte Carlo Simulations: Runs up to 100 million Poisson-distributed parallel universe match scenarios in milliseconds using vectorized Numpy arrays to calculate precise probabilities.
- Premium UI/UX: A sleek, beautiful, fully responsive dark-mode GUI engineered with CustomTkinter.
You don't need to be a programmer or have Python installed to use WorldCupML!
- Head over to the Releases Tab on this repository.
- Download the latest
WorldCupML.exefile from the V1.0 Release. - Double click it to launch the app instantly!
If you'd like to run the code from the source, clone the repository and install the required dependencies:
git clone https://github.com/Krshs90/WorldCupML.git
cd WorldCupML
pip install -r requirements.txt
python app.py- Python (Core Language)
- CustomTkinter (Frontend Interface)
- XGBoost & Scikit-Learn (Predictive ML Regressors)
- Pandas & Numpy (Data Analytics & Simulation Engine)
- ESPN API (Live backend payload)