Розробка та оптимізація бойового ші в колекційній картковій грі

Abstract

This paper examines the problem of developing combat artificial intelligence for a collectible card game, considering stochasticity and limited resources. The main approaches to building AI have been analyzed, and their advantages and disadvantages in the context of CCG have been discussed. The validity of using Expectimax as the basic algorithm has been substantiated. Suggested an optimization system based on Star2 for branch cutting, Transposition Tables for state caching, and Zobrist Hashing for quick position identification.