Creating engaging artificial characters for games

Abstract

Game developers and researchers aim to model human behaviour in order to create more engaging, entertaining and satisfying artificial characters for computer games. It is a popular belief that intelligent behaviour is the key to creating better game AI. However, as yet there is no empirical evidence to support this theory or to indicate whether other attributes, such as social interaction, realistic behaviour and communication, should also be considered. This study aimed to find out which attributes people desire in team members and opponents in computer games. The study employed a questionnaire, administered to a group of university students, directed towards ascertaining the importance of different aspects of player behaviour in games. It was found that there are two different, non-homogenous groups, each with separate needs and wants that game developers should consider. Firstly, it was found that people who prefer playing computer games with other humans tend to value intelligent behaviour and social interaction more than people who prefer computer players. Secondly, it was found that people who prefer computer players do so for convenience, practice and a preference for games that can only be played individually. It is recommended that games developers should aim to model intelligent behaviour for the first group and that the second group require an in-game learning environment for skill-development.