Pygskin - Football Analytics Platform
Software AI/ML Web

Pygskin - Football Analytics Platform

Django web application using ML to predict college football play calls, deployed on DigitalOcean with Nginx/Gunicorn

View on GitHub Pygskin Website

Overview

Pygskin is a comprehensive football analytics platform that leverages machine learning to provide insights into college football play calling patterns. The application predicts what plays teams are likely to call based on historical data and game situations.

Key Features

  • ML Model Inference: Real-time play call predictions using trained Scikit-Learn models
  • Data Pipelines: Engineered data processing workflows with Pandas and NumPy
  • Historical Analysis: Comprehensive play-by-play dataset analysis
  • Web Interface: Clean, responsive Django application
  • Production Deployment: Hosted on DigitalOcean with Nginx/Gunicorn stack

Technical Stack

  • Backend: Django, Python
  • ML/Data: Scikit-Learn, Pandas, NumPy
  • Deployment: DigitalOcean, Nginx, Gunicorn
  • Database: PostgreSQL

Key Achievements

  • Processed and analyzed thousands of historical play-by-play records
  • Achieved meaningful prediction accuracy for play call tendencies
  • Deployed production-ready web application with high availability

Learn More

Visit the live application at pygskin.com to explore college football analytics and predictions.