This study investigates the competition between two cognitive -behavioral mechanisms-exploration an‎d imitation-within the framework of complex networks. To this end, the results of the reference paper were first reproduced on undirected networks, including complete, ran‎dom, scale-free, an‎d small-world topologies, in order to validate the simulation procedure. The networks were then modeled as directed using three different schemes (ran‎dom orientation, upper-triangular orientation, an‎d fully one-directional orientation), an‎d the dynamics based on soft imitation an‎d ran‎dom exploration were applied to them. The variance analysis of strategies shows that, in undirected networks, all topologies eventually converge to a steady state with a fixed variance, although the final values differ across network structures. Specifically, the complete network exhibits the lowest steady-state variance, while the small-world network displays the highest. In contrast, in directed networks, convergence still occurs, but the steady-state variance increases significantly. This increase is minor in more regular networks (such as the complete network) an‎d substantial in more asymmetric structures, particularly small-world networks. Furthermore, examining the soft-imitation parameter q∈ [0,1] reveals that the midpoint q=0.5 consistently produces the minimum variance an‎d plays a decisive role in system stability. Overall, the findings indicate that edge directionality can enhance strategy diversity an‎d give rise to distinct emergent patterns compared to the undirected case.

كيوان آقابابائي ساماني

فرهاد فضيله

فرهاد شهبازي دستجرده , اسماعيل عبدالحسيني سارسري