Ranking Model Of Players Based On ELO Method

The ranking of table tennis players is conducive to evaluating the strength of table tennis players,helping the association to scientifically select talents,contributing to the development of China's table tennis career and improving the overall level of table tennis competition in China.In order to make up for the imperfections of several popular ranking algorithms in the current table tennis,this paper innovatively introduces the ELO scoring method originally applied in the field of chess into the ranking model of table tennis players,and under the real competitive competition data.It confirms the validity of the model.The basic ELO method ranking model was established by selecting the real game data of the 2016 Brazil Rio Olympic Games,the 2012 London Olympic Games,and the 2008 Beijing Olympic Games Table Tennis Men's Singles Competition.First,assign the same initial points to all participating athletes.Secondly,the outcome of the match between the athletes is used as an indicator of pairwise comparison.Finally,the results of the competition are substituted for calculation.From the results of the model,the basic ELO system is applicable to the table tennis player ranking model.This provides a new direction for the establishment of a ranking model for table tennis players with Chinese characteristics.Establish a dynamic ELO method ranking model.The ranking model for the basic ELO method can not reflect the athlete's previous strength level,the ranking has local stability,and the ranking in the back of the ranking list is not clear.This paper attempts to select the world points of each athlete as the starting point of the model operation,which will reflect the score of the player's strength.Introduced into the ranking model,and assigned different weights to each stage of the competition to establish a dynamic ELO method ranking model.The research results show that the dynamic ELO method ranking model is more realistic than the basic ELO method ranking model.A comparative analysis of the two models.This paper takes the world rankings of the athletes given by the ITTF as the evaluation standard,and firstly uses the frequency distribution map and the bipartite graph to visually compare the two visual visualizations.Then,the classic Kendall correlation coefficient and the measure of Spearman in statistics are selected for accurate numerical analysis.The comparison results show that the dynamic ELO method ranking model is more accurate than the basic ELO method ranking model.