This is a Streamlit web application that displays sports betting predictions and analytics for the NFL and College Football. The app fetches live data from a Supabase backend to show game predictions, player props, and model performance. Visit Pivot Bets and the Streamlit Web App
- League Selection: View predictions for NFL, NBA, College Football, and NFL Player Props.
- Model Accuracy: Displays the historical accuracy of the prediction models for Moneyline, Spread (ATS), and Totals (O/U).
- Game Predictions: Shows predicted winners, scores, and probabilities for upcoming matchups.
- Player Props: (NFL) Displays player-specific projections for yards, TDs, and boom/bust probabilities.
- Dynamic Filtering: Users can filter predictions by specific matchups
- Frontend: Streamlit
- Backend: R, GitHub Workflow, Supabase Database
- Languages: R, Python, SQL
- Core Libraries: pandas, supabase-client, tidyverse, cluster (k-means)
- Custom Models: Bayesian Inference, Monte Carlo integration