A general technique is proposed to deal with the formalization of intuition and human-oriented concepts in competition thinking games like chess, such as defensive play, attack, tactical play, etc. We present a manner of transferring these intuitions, which are in general ambiguous and not well defined, into formal definitions, and then directly to use them in game play. Among other concepts we define notions of attack, threat, defensive play and strategic play. Experiments are made in computer chess to empirically evaluate this technique. There are applications to machine learning, such as the possibility of combining different evaluation functions into a single one. There are also important applications in education such as teaching and evaluation of human players. Keywords chess, evaluation function, heuristics, intuition, null-mover