CONTROL AND DECISION MAKING

Algorithm for choosing an NPC behavior modeling method in video games

K. I. Shutov, A. A. Lobanov

MIREA — Russian Technological University, Moscow

Abstract: The article presents an algorithm for selecting NPC behavior modeling methods in video games. The proposed approach combines principles of the SHNUR method – one of the verbal decision analysis (VAR) techniques – with elements of the MAUT. The paper considers traditional approaches (FSM, behavior trees) and modern technologies (PANGeA, GANs, CiF-CK), providing a comparative analysis of their capabilities, advantages, and limitations. A method for selecting the optimal algorithm based on project criteria is proposed, which simplifies the decision-making process for developers.
The study aims to develop an efficient and clear tool for selecting the optimal NPC behavior modeling method based on the project's characteristics, available resources, and specific requirements. The results can be useful for game developers.

Keywords: game design, video games, NPC behavior, artificial intelligence, multi-criteria decision-making, VAR, SHNUR, MAUT, gaming industry.

DOI: 10.14357/20718632250408

Bibliographic databases:

•