Measuring players' performance efficiency index using Fuzzy logic in multiplayer sports: A basketball study

Abstract

In multi-player sports, measuring a player's performance according to position precisely and accurately is one of the topics worth researching. A player's efficiency scores show how well the player meets the requirements of the position she plays and at what level she contributes to the basketball team. The Performance Index Rating (PIR) is one of the most fundamental scoring and mathematical system used by the Euroleague Basketball Company's first and second tier competitions. This research aims to construct a fuzzy performance index system using fuzzy logic in basketball sports to increase players' awareness of their own performance and to evaluate their performance more sensitively and efficiently. The proposed system is presented seperately for three different positions in basketball: Guard, Forward, Center. Three different fuzzy inference systems are formed for each position: offensive action, defensive action and negative action systems. In these inference action systems, parameters used are point, assist, total successful throw, offensive rebound, steal, block, defensive rebound, total missed throw, turnover, faul, block against. For each position, the fuzzy performance index is generated through an upper fuzzy inference system where the results of the action systems are taken as input. For the proposed approach, detailed professional basketball data from the 2017-2018 to 2022-2023 seasons are used, taking into account that players must average at least 20 minutes of playing time per game. By comparing the obtained new fuzzy index values and the famous PIR values, it is observed that the proposed system is better, more sensitive and more effective both in position evaluation and in general. In future studies, applying this method to other multiplayer team sports is aimed.