IPL Crunch Analytics Dashboard

IPL generates millions of ball-by-ball data points, but traditional platforms only show static, generic stats. They fail to reveal granular patterns—like venue-specific scoring curves or how toss choices mathematically impact wins over 19 seasons. The goal was to build a multi-page analytics engine that turns raw data into actionable strategy for team planning. Process I followed a standard 4-step Data Science Pipeline: 1.Data Audit: Ingested the historic IPL ball-by-ball dataset (2008-2026). 2.Backend Engineering: Cleaned anomalies and structured 0-indexed over structures. 3.UI/UX Framework: Created a multi-page Streamlit layout for structured navigation. 4.Explorer Layer: Developed a dynamic Head-to-Head matchup sandbox Results Production-Ready App: Successfully hosted a live, responsive 6-page analytical dashboard on Streamlit Cloud. Data-Backed Insights: Discovered that despite a 66% bias toward fielding first, the actual toss-to-match win rate is just 50.49% (no real system edge). Dynamic Matchup Sandbox: Enabled users to select any Team vs Team pairing to instantly get live over-by-over run progression curves and match logs. Reflection Integrate an ML Score Predictor: I would build and deploy a real-time Machine Learning prediction engine (using regression models) at the backend to forecast the final innings score based on live inputs like current over, wickets, and venue metrics. Implement Advanced Cricket Metrics: Move beyond basic totals to introduce industry-grade scouting metrics like True Strike Rate (adjusted against stadium baselines) and Dot Ball Pressure Indices. Automate Live API Pipelines: Replace the static .csv file system with a live sports API pipeline to auto-refresh the application with ongoing match data