System Dashboard
Complete overview of the IPL Win Probability architecture.
Stacking ML Ensemble deployed via FastAPI. Combines Random Forest, Gradient Boosting, and XGBoost with a Logistic Meta-Learner.
Live Match Intelligence
Real-time ball-by-ball win probability powered by stacking ML.
Simulation Lab
Counterfactual "What-If" Analysis to quantify strategic impact.
Quantifies the exact shift in win probability caused by the scenario intervention.
Momentum Timeline
Ball-by-ball win probability and pressure index.
Model Explainability (SHAP)
Understanding model predictions through feature attribution.
The model dynamically assigns importance based on match context. While Required Run Rate and Overs Remaining generally dominate global importance, local ball-by-ball analysis reveals that Momentum and Pressure Index strongly influence predictions during critical death-over chases.
MLOps Governance
| Time | Action | Drift Flag | Perf Flag |
|---|
Franchise Intelligence
Strategic insights and export tools for teams.
Phase Dominance
Tactical Intelligence Brief
Run analysis to generate strategy insights...