Player Ranking in Taekwondo: A Bayesian Elo Rating System

This study provides a novel statistical ranking system for Taekwondo players in Sri Lanka. The world Taekwondo ranking function increases monotonically, which is discouraging for players who might win challenging games with higher ranked players. There is no current system for ranking Sri Lankan players, and the current world ranking system cannot be adapted to Sri Lanka, because only a few Sri Lankan players have yet reached even the lowest type of the tournaments, organized by World Taekwondo Federation, for which the rankings are applicable. A Bayesian Elo rating model which can be modified with the type of the tournaments is introduced for ranking Sri Lankan Taekwondo players. JAGS programming is used to estimate the scaling parameter in the predicting function of the rating model. The updated ranking point is then assigned for each player at the end of each tournament. The strategies to increase the rank of a player are illustrated by applying the proposed model to a tournament. The proposed model can even be used as an alternative for the current world ranking system.