General video game evaluation using relative algorithm performance profiles

Download: PDF.

“General video game evaluation using relative algorithm performance profiles” by Thorbjørn S. Nielsen, Gabriella A. B. Barros, Julian Togelius, and Mark J. Nelson. In Proceedings of the 18th Conference on Applications of Evolutionary Computation, 2015, pp. 369-380.

Abstract

In order to generate complete games through evolution we need generic and reliable evaluation functions for games. It has been suggested that game quality could be characterised through playing a game with different controllers and comparing their performance. This paper explores that idea through investigating the relative performance of different general game-playing algorithms. Seven game-playing algorithms was used to play several hand-designed, mutated and randomly generated VGDL game descriptions. Results discussed appear to support the conjecture that well-designed games have, on average, a higher performance difference between better and worse game-playing algorithms.

BibTeX entry:

@inproceedings{GVGprofiles:Evostar15,
   author = {Thorbj{\o}rn S. Nielsen and Gabriella A. B. Barros and Julian
	Togelius and Mark J. Nelson},
   title = {General video game evaluation using relative algorithm
	performance profiles},
   booktitle = {Proceedings of the 18th Conference on Applications of
	Evolutionary Computation},
   pages = {369--380},
   year = {2015}
}

Back to publications.