A method of soccer's player rating based on machine learning

The system for rating players after the match has great commercial value, and it is also a hot topic that has been discussed in academia. This paper envisages “standing on the shoulders of giants”, using existing rating as a basis, using machine learning methods to find a balance among different rating systems, and deriving a player's post-match rating system based on the outcome of the soccer match. This not only solves the problems of high data acquisition difficulty and multiple dimensions, but also integrates the existing rating system. The experimental results also show the feasibility of this method.