On Finding the Most Compatible Batting Average

Prodip Kumar GOUR
Dibyojyoti BHATTACHARJEE

Keywords

Data Mining in Sports, Performance Measurement, Cricket, Batting Average

Abstract

Batting average is the most commonly used measure of batting performance in cricket. It is defined as the total number of runs scored by the batsman divided by the number of innings in which the batsman was dismissed. Generally, the innings of a batsman comes to an end due to his dismissal, yet there are some cases in which the batsman may not get dismissed due to sudden termination of the batting innings of the team. The sudden termination may take place due to bad weather or victory or injury of the batsman or for running short of partners etc. In case, there are several not out innings in the career of a batsman, the batting average may get overestimated. To overcome this problem of over estimation, several authors proposed different modifications to the existing formula of batting average or defined new measures. Though each method expressed its advantages over the existing batting average, yet none of them are universally accepted as the most efficient replacement of the existing formula. This paper makes an attempt to study the existing solutions to the problem and then to evaluate the best or at least the most compatible alternative. For the purpose of quantification, data from the ICC Cricket World Cup played in Australia and New Zealand in 2015 is considered.