Statistical modelling of cricket data

Abstract

Within the dynamic world of cricket, a game celebrated for its record-setting performances across various platforms, numerous questions arise in the minds of viewers before the start of any series. Specifically, the anticipation surrounding the selection of players for a country’s team prompts our statistical modelling approach to provide answers. Factors such as player’s overall stats, team performance, and last five performances are considered. The approach focuses on individual player batting and bowling performances, career statistics, and player independent factors. Experimental analysis using Hadoop and Hive for Indian players shows up to 92% accuracy in prediction of selection of players.