Self-segregation versus clustering in the evolutionary minority game

Complex adaptive systems have been the subject of much recent attention. It is by now well established that members ("agents") tend to self-segregate into opposing groups characterized by extreme behavior. However, the study of such adaptive systems has mostly been restricted to simple situations in which the prize-to-fine ratio R equals unity. In this Letter we explore the dynamics of evolving populations with various different values of the ratio R , and demonstrate that extreme behavior is in fact not a generic feature of adaptive systems. In particular, we show that "confusion" and "indecisiveness" take over in times of depression, in which case cautious agents perform better than extreme ones.