Resources‎ > ‎

Scientific papers

A short-list of papers relevant to AI-based game design.

Game AI Revisited. Yannakakis, G. N. (2012).  Proceedings of ACM Computing Frontiers Conference.
In this paper Georgios (my colleauge at UoM) recognises that AI in the games industry and in academia respectively have been quite different things, but that there has, recently, emerged four key areas in game research that might change this. These areas are, according to Georgios:  (1) computational modeling of player experience, (2) the procedural generation of content, (3) the mining of player data on massive-scale,  and (4) the alternative AI research foci for enhancing NPC capabilities.

AI-Based Game Design: Enabling New Playable Experiences, Mirjam P. Eladhari, Anne Sullivan, Gillian Smith, Josh McCoy, University of California Santa Cruz, Technical Report, December 2011
Anne, Gillian, Josh, and myself wrote this report after having had lots of discussions about how each of us approach AI in relation to game design in our respective game prototypes that we have designed and implemented. We saw some common challenges and ways to approach them. 

Expressive AI: Games and Artificial Intelligence. Mateas, M. 2003. DiGRA (2003).
In this paper Michael (who was the advisor for my PhD) outlines Expressive AI as field that takes both artistic values into account as well as technological ones, to be applied to games (and other media). It is recognised that Game AI (including AI for Chess and Go, but applied to digital games generally) has focus on both the audience experience and on human authorship. (That is, there is more to game AI than to find the optimal algoritm, there is the design of the play experience too to take into account.) 

Research Directions for AI in Computer Games. Fairclough, C., Fagan, M., Namee, B. Mac, & Cunningham, P. (2001). Development, 333–344.
A decade ago Fairclough et al. noted that the interest for applying AI to games was growing among researchers, outlining research concerning non player characters and interactive story telling. Fairclough notes that "Some might say that research into AI for computer games is not the most noble undertaking."  I remember when I went to a conference for the first time, to the Computers and Games Conference in Alberta in 2002. I was, then, working with at story logics in narrative games, but that was seen as a bit obsqure. Most of the other papers were on deep search techniques for AI opponents in Chess and Go - that was, back then, what most academics thought of as "Game AI".