The efficiency of individual optimization in the conditions of competitive growth

The paper aims to discuss statistical properties of the multi-agent based model of competitive growth. Each of the agents is described by the growth (or decay) rule of its virtual “mass”, with the rate affected by the interaction with other agents. The interaction depends on the strategy vector and mutual distance between agents, and both are subjected to the agent's individual optimization process. Steady-state simulations yield phase diagrams with the high and low competition phases (HCP and LCP, respectively) separated by critical point. Particular focus has been made on the indicators of the power-law behavior of the mass distributions with respect to the critical regime. In this regime, the study has revealed a remarkable anomaly in the optimization efficiency.