Advantages and disadvantages of minimaxing versus product-propagation back-up rules for game tree searching have been intensively discussed in the literature. So far, examinations have almost exclusively been carried out through experiments, demonstrating slight superiorities for one or the other of these back-up rules, depending on the particular game chosen or on assumptions underlying simulations on abstract game trees. In contrast to these purely quantitative investigations, we aim at elaborating differences in strength of these back-up rules by characterizing properties of critical situations in which the differences between these back-up rules prove relevant. Evidence from the examinations carried out suggest that degrees of quality of the static evaluation function used, ranges of the heuristic values associated with the positions compared, and frequencies of value combinations are major properties influencing the suitability of either back-up rule. The results provide hints for assessing degrees of competence of minimaxing and product propagation, which can be exploited beneficially for motivated combinations of the two back-up rules, according to game tree properties observed in a particular game.
Keywords: Search, Automated Reasoning
Citation: Helmut Horacek: Towards Understanding Conceptual Differences Between Minimaxing and Product Propagation. In W.Horn (ed.): ECAI2000, Proceedings of the 14th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2000, pp.604-608.