Methods of generating scenarios of enemy behavior in turn­based strategy games

DOI: 10.31673/2412-9070.2024.024851

Abstract

This scientific article is devoted to the methods of generating the scenario of the opponent’s behavior in turn-based strategy games. The opponent in the game is an integral part of the gameplay. The quality of the game as a whole depends on artificial intelligence work, complexity and interaction with the player.
The topic of the article focuses on the analysis of existing systems for generating the behavior of the opponent and their shortcomings, as well as on the general characteristics and features of artificial intelligence in turn-based strategy games. The result of the analysis is the conclusion that the existing methods of generating the behavior of the opponent’s artificial intelligence has both a number of advantages and disadvantages.
During the comprehensive analysis of existing methodologies in the field, critical challenges pertaining to the predictability of artificial intelligence (AI) adversaries have been revealed. These challenges encompass frequent errors arising from incomplete or imprecise information, limitations in the analysis of diverse decision pathways, and the complexity associated with managing a substantial array of potential actions.
Following an exhaustive review of both theoretical concepts and empirical investigations, it is evident that leveraging the foundational principles of RHEA (Rolling Horizon Evolution Algorithm) and MCTS (Monte Carlo Tree Search) holds significant promise for future advancements. The primary objective in forthcoming endeavors involves the refinement and enhancement of the current algorithm designed to generate behavioral scenarios for adversaries within turn-based strategy games. This strategic goal is aimed explicitly at rectifying the deficiencies observed in prior methodologies and augmenting the overall efficacy and quality of gaming experiences.

Keywords: artificial Intelligence; genetic algorithm; turn-based strategy game; generation methods.