Pairwise Comparison of Chess Opening Variations

Abstract

It is demonstrated that a methodology previously described for comparing a number of different test openings with a neutral and perfect opening may also be used to make pairwise comparisons of alternate continuation lines of any chess opening. Such comparisons show, for example, that Accelerated Dragon variations initiated by black are not robust because they depend on white's cooperation. Rather than these specific findings, however, the real purpose of this paper is to further refine, validate, and demonstrate an objective methodology for evaluating the relative merit of chess openings, a subject of great interest to chess players at all levels of play. 1 . This paper is the fourth of a series in a study undertaken to develop a generally applicable methodology for the objective evaluation of chess openings. The proposed methodology uses controlled experiments with chess engines to compare chess openings. In a previous paper, a dual engine experimental design (DED) was used to evaluate a selected set of queen's pawn openings (Munshi, Comparing Chess Openings Part 3, 2014). The dual engine test is more comprehensive than single engine analysis because a greater number of opening variations are explored and the test is therefore more thorough (Munshi, Comparing Chess Openings, 2014). In this paper we describe and demonstrate a further application of the DED methodology by showing that it may be used to make pairwise comparisons 2 of two alternative continuation lines of any chess opening without the need to identify a neutral and perfect opening as a standard. For example, in the Yugoslav Attack should white play 9.Bc4 or castle queenside instead? In all such cases, pairwise comparisons may be used to determine whether one of the two variations being compared offers a statistically higher probability of winning ceteris paribus.