Attribute Selection and Verification for Effective Hockey Player Ranking in the Sri Lankan Context

Abstract

Sri Lanka's hockey domain has been hindered by the lack of a standardized dataset and the absence of a well-defined set of attributes for player performance assessment, limiting the application of advanced AI technologies. This study aims to identify and verify a key set of attributes specific to Sri Lankan hockey players. The methodology involved three phases: First, an extensive literature review of 30 research papers was conducted to identify a preliminary list of 10 attributes. Second, these attributes were calibrated through face-to-face interviews with coaching experts, involving 25 national-level hockey players, resulting in the refinement of 8 attributes. Finally, these attributes were verified through a structured questionnaire survey conducted with hockey coaching experts, leading to the finalization of 9 key attributes: Front Goals, Penalty Strokes, Penalty Corners, Green Cards, Yellow Cards, Red Cards, Card Issue Time, Goal Time, and Skill Accuracy. The next phase of this research will focus on developing a mathematical rating equation with weighted attributes to calculate performance scores for each player.