Detecting Fortresses in Chess
Matej Guid, Ivan Bratko
Artificial Intelligence Laboratory, Faculty of Computer and Information Science, University of Ljubljana
Abstract.
We introduce a computational method for semi-automatical detecting fortresses in the game of chess.
It is based on computer heuristic search and can be easily used with any state-of-the-art chess program. We
also demonstrate a method for avoiding fortresses and show how to find a break-through plan when one exists.
Although the paper is not concerned with the question whether it is practical or not to implement the method
within
the state-of-the-art chess programs, the method can be useful, for example, in correspondence chess or
in composing chess studies, where a human-computer interaction is of great importance, and the time available
is significantly longer than in ordinary chess competitions