Dear All, does anybody know of works on modeling agents who would be willing to harm other agents while not taking into account what could be the resulting consequences for themselves ("mean" agents, so to say)? I would appreciate every piece of information regarding works related to that. The underlying motivation is that such agents could be responsible for interesting systemic effects in a society consisting of the majority of "normal" and a minority of "mean" agents. Specifically, I do not imply here that "mean" agents are born as such, but just that in prevalent number of circumstances their actions are interpreted as such, for whatever reason (frustration, desperateness, competition, ...). I apologize in advance for any inconvenience this message could possibly induce. With kind regards! Josip Stepanic Department of Quality Faculty of Mech.Eng & Naval.Arch. - Univ. of Zagreb