YASHWIN SAI

IPL Ball-by-Ball Analysis — Toss, Phase & Venue Intelligence Across 1,218 Matches

IPL analysts and fans rely on gut feel and commentary narratives — "Chinnaswamy is a chasing ground," "winning the toss matters," "powerplay sets the game up." But none of these assumptions had been stress-tested against ball-by-ball data across all 1,218 IPL matches. The gap: no one had quantified whether these widely-held beliefs actually hold up statistically, or whether they collapse under scrutiny when you control for variables like target size, venue name variants, and no-result matches. Process Downloaded ball-by-ball IPL data (289,673 deliveries, 2007–2026) and loaded it into Python using pandas. First cleaned all venue name variants — Cricsheet uses 3 different strings for Chinnaswamy alone — then excluded no-result and tie matches from all win-rate calculations. Tested three core questions: toss impact, phase-level run rates, and top players. For the surprise finding, I started with a flawed hypothesis (death-over finisher SR predicts wins) but stress-tested it and found survivorship bias — innings with fast finishers were already winning. Rebuilt the analysis around venue-conditional chasing and found the 52-point swing at Chinnaswamy. Every number was verified by cross-checking raw match counts before being submitted. Results Toss winners win just 51.6% — near coin flip — but choosing to field after winning adds a 9.3-point swing. Death overs show the largest phase gap at 1.54 RPO between winning and losing innings. The headline finding: Chinnaswamy's chasing advantage is entirely target-conditional — 76% chase rate under 170 collapses to 24% at 200+, a 52-point swing. Verified across 100 clean matches. Chinnaswamy actually has the lowest 200+ defending rate (76%) of major venues — lower than Kotla (85%) and Eden (84%). Reflection I'd add innings-level context earlier — my first "finisher SR" finding looked strong until I checked whether those innings were already in a winning position, which killed the claim. Lesson: always control for game state before claiming causation. I'd also incorporate weather and time-of-day data to properly test the dew hypothesis rather than reasoning from cricket logic. And I'd split the analysis by era (pre-2016 vs post-2016) to check whether the toss decision trend towards fielding is getting stronger over time as captains have learned from the data.