Fuzzy Chess Tactics

This paper presents a fuzzy implementation for the tactics of the game of chess. Chess as been traditionally used for explaining fuzzy reasoning, and is often considered a platform of choice to test intelligent systems. However, an extensive search for applications of fuzzy sets, inference or reasoning to the game of chess, gives hardly any concrete results. One can find fuzzy improvements to classic tree search algorithms like, for example, fuzzy forward-pruning (which can obviously be applied to chess), but not a single relevant result in what concerns applying fuzzy to express or implement human chess knowledge. It is a fact that, chess, the ultimate measure of “intelligence” in old times, seems to have shed some of its gloss once it was discovered that you can do it quite well with alpha-beta pruning and a lot of sheer calculating power. However, chess knowledge is still arguably one of the factors that make good human chess players so hard to be beaten by a machine as long as time is not an issue for instance, in e-mail based tournaments, humans have a large advantage over machines and their expertise is the defining winning factor.