Sushant Singh

IPL Match-Winning Insights Analysis

The objective of this project was to analyze IPL ball-by-ball data to uncover patterns behind match victories. I wanted to investigate whether winning the toss truly provides a competitive advantage, identify which match phase contributes most to winning, and determine the top-performing batters and bowlers across multiple IPL seasons using data-driven analysis. Process I began by exploring and cleaning the IPL dataset using Pandas. The dataset contained nearly 289,000+ ball-by-ball records with mixed datatypes and missing values in wicket-related columns. I standardized season formats, handled logical null values, and structured the data for analysis. For toss analysis, I compared toss winners with actual match winners to calculate win percentages. For phase analysis, I divided innings into Powerplay (1–6), Middle Overs (7–15), and Death Overs (16–20) to compare average scoring patterns between winning and losing teams. To identify top performers, I aggregated batter runs and bowler wickets across five seasons while excluding run-outs from bowler wicket counts for more accurate cricket analysis. Results The analysis revealed that toss advantage had very little impact on overall IPL match results, with toss winners winning only slightly more matches than toss losers. The most significant insight was that death-over scoring had the strongest relationship with victory, as winning teams consistently outperformed losing teams during overs 16–20. The project successfully transformed raw ball-by-ball cricket data into meaningful insights using data cleaning, aggregation, statistical analysis, and visualization techniques. Reflection If I had more time, I would extend the project by building an interactive dashboard using Streamlit or Power BI for real-time exploration of IPL insights. I would also incorporate advanced analytics such as win probability prediction, team-wise strategy comparisons, and machine learning models to predict match outcomes based on live match situations. Additionally, I would include more seasons and deeper player performance metrics like strike rate impact, economy under pressure, and partnership analysis.