Strategy research based on chess shapes for Tibetan JIU computer game

The game of NoGo needs to be optimized its theoretical system and evaluation methods since it is been proposed in recent years. At present, most of the programs of NoGo are based on Monte-Carlo Tree Search (MCTS), but its programs are rarely based on static evaluation algorithm. In this paper, according to the rules and characteristics of NoGo, a static evaluation method has been put forward. In the method, the points on the chessboard of NoGo are classified by the summarized information of the points around them firstly and then the empty points are estimated. The valuation of an empty point is according to its type and the types of the points around it. The sum of the valuations of all empty points on the chessboard is the total evaluation of the chessboard. The experimental results show that the static evaluation method can accurately estimate the value of each state in NoGo.

Research on static evaluation method for computer game of NoGo

This paper has studied the game rules, victory or defeat rules, and the key techniques of Dots-and-Boxes, and has designed and realized one 6×6 Dots-and-Boxes based on the representing method of Strings and Coins. By corresponding Dots-and-Boxes chessboard to strings and coins chessboard, the proposed representing method is not only intuitive and convenient in chessboard representation, but also has other 3 merits which include: (1) can reduce the size of representing data largely, and simplify the chessboard stored matrix, thus easy to store and analysis the data, (2) help to judge the key chains and rings among elements of chessboard, and (3) help to calculate the number of chains, and apply the long-chain rule. In the 4th machine gambling tournament took in 2010, the realized Dots-and-Boxes system gained the second best, which indicates the feasibility of the proposed chessboard representing method and the validity of the Dots-and-Boxes system.

/pdf/research-and-implementation-of-dots-and-boxes-game-system-9u8x3z5uo4.pdf

Research and Implementation of Dots-and-Boxes Game System

Doudizhu poker is a very popular and interesting national poker game in China, and now it has become a national competition in China. As this game is a typical example of incomplete information game problem, it has received more and more attention from artificial intelligence experts. This paper proposes a multirole modeling-based card-playing framework. This framework includes three parts: role modeling, cards carrying, and decision-making strategies. Role modeling learns different roles and behaviors by using a convolutional neural network. Cards carrying can calculate reasonable rules especially for “triplet” by using an evaluation algorithm. Decision making is for implementing different card strategies for different player roles. Experimental results showed that this card-playing framework makes playing decisions like human beings, and it can to some extent learn, collaborate, and reason when facing an incomplete information game problem. This framework won the runner-up in the 2018 China Computer Game Competition.

/pdf/study-on-the-strategy-of-playing-doudizhu-game-based-on-5fccpnmpz0.pdf

Study on the Strategy of Playing Doudizhu Game Based on Multirole Modeling

Computer Game is an important and challenging research direction in the field of artificial intelligence. It is the best experimental carrier for studying human thinking and realizing machine thinking. Checkers (100) are regarded as a complex two-player game, and their game tree search based on traditional expert knowledge has been difficult to deal with the changing search conditions of chess games. The emergence of UCT algorithm has broken this search bottleneck. First, The MCTS-based UCT algorithm was applied to computer Checkers in this paper, and then the UCT is improved aiming at the selection of expandable nodes irregularly and the characteristics of Checkers’ moves. The sorting strategy of prioritizing the optional nodes is presented before the UCB algorithm choosing a node, which enhances the selection and expansion of the search tree. The experiments show that the proposed algorithm improves the performance of UCT algorithm, and both the chess power and the search efficiency are improved significantly.

Application and Improvement of UCT in Computer Checkers

A long-term goal of computer game programmers is to make computers able to play board and card games well enough to challenge Super Master players.In order to speed up the development of new domains of research that could yield breakthrough results,it is necessary to understand what is holding back computer game research.This paper describes key problems in the areas of academic concepts,actual games,theoretical developments,military applications,international competitions,cultivation of the creativity of students and so on,all of which must to be quickly solved in order to get more results in this challenging and important field.

MA Zong-min

Papers

Challenging issues facing computer game research