Design and analysis of an unbiased intelligent recommendation system for all-rounders in cricket based on multiple criteria decision making

Article Type

Research Article

Publication Title

Engineering Applications of Artificial Intelligence

Abstract

Designing and analyzing an unbiased intelligent all-rounder recommendation system in cricket is a critical and complex decision-making task, where performance prediction itself is a crucial issue. The majority of existing works on cricket focus on batsmen, bowlers, and teams. However, performance analyses of all-rounders are hardly found. Hence, the motivation of this paper is to propose an artificial intelligence (AI)-based method for assessing the performance of all-rounders utilizing various multiple-criteria decision-making (MCDM) techniques. With this goal in mind, effective attributes are considered for evaluating all-rounder bowling and batting performances. Case studies using a set of 20 all-rounders have been taken from the recent International Cricket Council (ICC) all-rounders list to determine their ranking. The performance of various MCDM techniques is investigated using ICC rankings, with the Spearman Rank Correlation Coefficient applied to the One Day International (ODI) format. The results indicate that the Criteria Importance Through Intercriteria Correlation (CRITIC)-VlseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR) method yields quite promising insights, achieving a higher correlation with ICC rankings (Spearman Rank Correlation: 0.794) and hence can be used as an intelligent AI-based all-rounder recommendation system. Finally, we have also conducted a sensitivity analysis on the ranking outcomes of all-rounders to examine the utility and robustness of our proposed MCDM approach. The proposed approach is beneficial for team selectors, analysts, and fantasy sports platforms, offering a fairer and more reliable ranking of all-rounders.

DOI

10.1016/j.engappai.2025.112197

Publication Date

12-9-2025

Link to Full Text

Share

COinS