IPL Crunch'26

We were to do ball-by-ball data analysis to identify match patterns in IPL games and understand which factors contribute most to winning matches. Specifically, I analyzed whether winning the toss actually improves match outcomes, which innings phase (PowerPlay, Middle Overs, or Death Overs) has the strongest relationship with winning, and who the top-performing batters and bowlers were across seasons. The dataset contained detailed delivery-level IPL records, requiring aggregation and transformation to generate match-level insights and visualizations. Process I cleaned & structured the IPL ball-by-ball dataset using SQL Server by correcting datatypes, handling NULL values & resolving duplicate match-level calculations using distinct match IDs. I used SQL queries to analyze toss-win conversion rates, phase-wise scoring trends & player performance statistics. For phase analysis, I categorized the game into PowerPlay (1–6), Middle Overs (7–15) & Death Overs (16–20), then aggregated runs scored by winning and losing teams separately to avoid misleading ball-level averages. The cleaned dataset was imported into Tableau Public, where I built visualizations - toss impact charts, phase comparison charts and top batter/bowler. A surprising insight was that middle overs showed a larger scoring gap between winning and losing Results The analysis showed that teams winning the toss had only a moderate advantage in overall match outcomes, suggesting in-game performance mattered more than toss results alone. A key insight was that middle overs produced a larger scoring gap between winning and losing teams than death overs, indicating that teams often establish control before the final phase of the innings. The dashboard successfully transformed raw ball-by-ball data into clear visual insights using SQL and Tableau, improving inter-pretability of match patterns and player performance trends. Reflection If I continued this project, I would expand the analysis by adding advanced metrics such as strike rate and wicket rate per over. I would also integrate Python for deeper statistical analysis and predictive modeling, such as win probability estimation and feature importance analysis. Additionally, I would improve the dashboard interactivity by adding filters for seasons, teams and phase to make the analysis more exploratory and user-friendly.