Comparative Analysis of Cricket Batsmen Using Survival Analysis

Abstract

This study presents a comparative analysis of exponential and Weibull distributions for modeling batting performance in T20 international cricket, using survival analysis to evaluate player consistency. Leveraging individual innings data from 10 elite batsmen (including Babar Azam, David Warner, and Virat Kohli) sourced from ESPNcricinfo, the research employs maximum likelihood estimation (MLE) to fit both distributions and compares their goodness-of-fit via Akaike (AIC) and Bayesian (BIC) information criteria. Results demonstrate that the exponential distribution, characterized by a constant hazard rate, provides a superior fit for most players, supported by lower AIC/BIC values and chi-square goodness-of-fit tests (p > 0.05). Survival probabilities at key thresholds (10, 30, 50, 80, and 100 runs) reveal Babar Azam as the most consistent performer, with a 75.6% chance of surpassing 10 runs and a 24.7% probability of reaching 50 runs. Conditional survival analysis further highlights his ability to convert starts (57.1% chance of progressing from 30 to 50 runs), outperforming power-hitters like Glenn Maxwell, who exhibit higher volatility. The Weibull distribution, while flexible, proves unnecessarily complex for T20 batting data. Practical applications include optimizing batting orders, real-time tactical decisions, and player development frameworks. The study underscores the exponential model’s utility in cricket analytics while advocating for its integration into team strategies. 

Keywords: Survival Analysis, T20 Cricket, Exponential Distribution, Weibull Distribution, Batting Consistency, AIC/BIC, Maximum Likelihood Estimation, Babar Azam, Conditional Survival Probabilities, Player Performance