Synthetic Players A Quest for Artificial Intelligence in Computer Games

Synthetic players are the computer-controlled actors in a computer game. In this paper, we describe both the implementation issues and the behavioural expectations related to synthetic players. We recognize the role of synthetic players by analysing the components of games in general and computer games in particular. In 1962 students at MIT competed against each other in the first real-time graphical computer game Spacewar (Graetz 1981). Probably none of them could have dreamt how realistic and complex computer games would turn into in four decades and how large a business would grow around them. However, it took ten years before commercial arcade games such as Pong and Space Invaders created the business in the 1970s and before home computers brought computer games within the reach of all enthusiasts in the 1980s. Since then game development and programming have turned from small amateur enterprises into a more professional and larger-scale industry. Nowadays, the typical time-span of development from an idea to a finished product is about two years and demands the work contribution of 20–50 persons. Game industry is usually compared to film industry, and in 2001 the global market of computer games finally surpassed global film box office. The current estimates of the annual revenue generated by computer games are around €25 billion and the annual growth is predicted to be over ten per cent over the next few years (Game Developers' Association of Australia 2003).