ADITYA PRASAD

PL Crunch'26: Deconstructing the Anatomy of a Victor

T20 cricket strategy is often driven by subjective 'conventional wisdom'—specifically the belief that winning the coin toss is a major determinant of match outcomes. The industry lacks a clear, data-backed consensus on whether this toss bias is an empirical reality or a psychological myth. I identified a gap in existing analytics: while surface-level stats exist, there was a need to deeply analyze phase-by-phase performance (Powerplay, Middle, and Death Overs) to isolate the actual statistical drivers of victory versus perceived advantages. Process My approach prioritized data integrity before visualization. I began by standardizing a massive, multi-season dataset, using Python to resolve mixed data-type errors across season formats. I then engineered a feature-extraction framework to categorize over 200,000 individual deliveries into discrete tactical phases, ensuring accurate phase-specific run-rate analysis. Throughout the project, I iterated on dashboard design: early attempts at simple bar charts failed to communicate the momentum shifts clearly. I pivoted to an integrated line-chart visualization to better represent phase-wise performance volatility. I prioritized a 'Tiled' dashboard architecture over floating elements to ensure a clean, professional, and mobile-responsive layout for the final output. Results The analysis empirically debunked the 'toss advantage' myth, revealing a near-perfect 50.4% vs 49.6% win split. Most importantly, I isolated 'middle-over run-rate differentials' as the primary statistical driver of victory. The final product is a production-ready BI dashboard that allows stakeholders to filter player contributions by match outcomes, providing an immediate, actionable scouting and strategy tool. Reflection If I were to scale this, I would incorporate player-specific strike rates under pressure situations (e.g., chasing a high target in the final 5 overs) to deepen the personnel evaluation. Additionally, I would explore adding a predictive modeling layer—using machine learning to calculate 'win probability' at any given point in a game—to evolve this from an explanatory analysis into a live, real-time strategic decision-support engine.