Skip to content

Repository files navigation

Turkey Süper Lig Predictions

Turkey Süper Lig Predictions

AI-powered match predictions for Turkish football with bookmaker odds comparison


Overview

This app predicts outcomes for Turkey's Süper Lig matches using machine learning models trained on historical data. It compares model probabilities against real-time bookmaker odds to identify potential value bets.

Features

  • Match Predictions — Win/Draw/Loss probabilities for upcoming fixtures
  • Bookmaker Comparison — Real-time odds from Bet365, Unibet, and more
  • Value Detection — Highlights bets where the model disagrees with the market
  • Team Analysis — Deep dive into team performance and statistics
  • Live Standings — Current league table and historical trends
  • Model Lab — Explore model performance and feature importance

League Info

  • League: Süper Lig (Turkey)
  • Teams: 19 clubs including Galatasaray, Fenerbahçe, Beşiktaş, and more
  • Data Sources: football-data.org, ESPN, Odds-API.io
  • Season: 2026-27

Quick Start

  1. Clone the repository
  2. Copy .env.example to .env and add your API keys
  3. Run the app:
pip install -r requirements.txt
streamlit run predictions.py

For detailed setup, configuration, and development workflows, see TECHNICAL.md.

How It Works

The app uses an ensemble of machine learning models (XGBoost, neural networks) trained on:

  • Historical match results and statistics
  • Team form and head-to-head records
  • Stadium coordinates and weather data
  • Bookmaker odds (for comparison, not prediction)

Predictions are generated before each matchday and cached for fast loading.

API Keys

You'll need free API keys from:

Add them to your .env file. See TECHNICAL.md for detailed setup, configuration, and development workflows.

License

This project uses the Pitch Oracle core library. See pitch-oracle-core for licensing details.

About

Super Lig (Turkish futbol/soccer) Data Analytics and Betting

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Contributors

Languages