Frequency distribution of contextual patterns in the game of go

Abstract

In this paper, we present two statistical experiments on the frequency distribution of patterns in the game of Go. In these experiments, we extract contextual patterns of Go as spatial combinations of moves. An analysis of a collection of 9447 professional game records of Go shows that the frequency distribution of contextual patterns in professional games displays a Mandelbrot fit to Zipf's law. Additionally, we show that the Zipfian frequency distribution of Go patterns in professional games is deliberate by rejecting the null hypothesis that the frequency distribution of patterns in random games exhibits a Zipfian frequency distribution. © 2008 Springer-Verlag Berlin Heidelberg.