Abtract

This paper presents a novel approach to analyze human decision making and decision
quality in a setting in which fully rational decisions are ruled out due to cognitive constraints:
chess. The identification approach contrasts human behavior with a computational
benchmark of cognitively bounded rationality based on the algorithms contained in modern
chess engines. The results document that behavioral deviations are not necessarily
associated with worse performance. Additional evidence shows that behavior is related to
asymmetric evaluation, time pressure, fatigue, and complexity. Faster decisions are
associated with more frequent deviations, but also with better performance, suggesting that
intuition and expertise might be important behavioral factors.

More informations

Location
Faculté Jean Monnet
54, Boulevard Desgranges 92330 SCEAUX
RITM, Room Gaudemet