page 1  (15 pages)
2to next section

An earlier version of this paper appears in: H.J. van den Herik and L.V. Allis, editors, Heuristic Programming in Artificial Intelligence 3 { The Third Computer Olympiad. Ellis Horwood, 1992.

METAGAME: A New Challenge

for Games and Learning

Barney Pell1

University of Cambridge

Cambridge, UK

E-mail: bdp@cl.cam.ac.uk

Abstract

In most current approaches to Computer Game-Playing, including those employing some form of machine learning, the game analysis mainly is performed by humans. Thus, we are sidestepping largely the interesting (and difficult) questions. Human analysis also makes it difficult to evaluate the generality and applicability of different approaches.

To address these problems, we introduce a new challenge: Metagame. The idea is to write programs which take as input the rules of a set of new games within a pre-specified class, generated by a program which is publicly available. The programs compete against each other in many matches on each new game, and they can then be evaluated based on their overall performance and improvement through experience.
This paper discusses the goals, research areas, and general concerns for the idea of Metagame.

1Parts of this work have been supported by RIACS, NASA Ames Research Center [FIA], and a British Marshall Scholarship.