IPL Data Analytics 2008-2026 — Analysed 293,308 Deliveries to Find What Really Wins Matches

Everyone has IPL opinions but very few back them with data. Three questions needed answering using real ball-by-ball data: 1. Do toss winners actually win more matches? 2. Which phase — Powerplay, Middle or Death Overs — wins matches? 3. Who are the top 5 batters and bowlers across all seasons? The challenge was finding verified complete IPL data and converting 1,233 raw JSON files into meaningful analysis. Process Downloaded 1,233 JSON files from Cricsheet.org. Wrote Python code to parse all files into Pandas DataFrames. Cleaned data by merging renamed teams like Delhi Daredevils to Delhi Capitals. Removed run outs from bowler stats for accuracy. Classified every ball into Powerplay, Middle or Death Overs phase. Compared winning vs losing team runs per ball in each phase. Built HTML dashboard with 4 charts and 2 tables to show findings. What didn't work: Kaggle datasets had wrong player names so switched to Cricsheet.org for verified accurate data. Results 1. Toss winners win only 51.7% — barely better than a coin flip 2. Death overs most decisive — winners score 1.753 vs 1.457 for losers 3. Kohli leads all time with 9,213 runs across 19 seasons 4. Chahal leads wickets with 231 — spinner beating all pacers 5. SURPRISE: Narine has best death economy 7.29 beating every pacer Dataset: 1,233 matches, 293,308 deliveries, 19 seasons 2008-2026 Reflection 1. Expand venue analysis to cover all 20 IPL stadiums with batting first vs chasing rates for each ground 2. Add player form tracking season by season not just all time totals 3. Build match win probability predictor using 19 seasons as training data 4. Add team auction spend vs wins to find best value franchises 5. Add head to head team comparison — which team beats which most