Overview of Amazon chess’ game playing algorithms

Abstract

Amazon chess is a machine game with high dimension branches, in which the first step is more than 2000 actions, so it is an ideal test platform for search algorithm and Monte Carlo tree search. This paper analyses the rules and common game algorithms of Amazon chess based on its environment, and compares the advantages and disadvantages of different algorithms. Finally, the improvement direction of Monte Carlo tree search algorithm is reviewed from four aspects of evaluation function, database, search strategy and parallel computing, and the research prospect of Amazon Monte Carlo game algorithm is prospected