Abstract
Epidemics are nonlinear adaptive processes in which pathogen spread and human behavior form a tightly coupled feedback loop. Individual decisions about protective measures create strategic interactions. These interactions can either accelerate disease spread or drive collective suppression. We introduce a theoretical lattice-based agent model that fuses SIS contagion with an evolutionary game, systematically exploring how strategy choice and infection pressure co-evolve through comprehensive parameter space analysis. Agents choose between self-isolation and normal activity based on population-wide disease prevalence and perceived costs. Agents then update strategies using a Fermi rule based on global infection prevalence and perceived costs. Infections propagate through contact-based transmission with behavior-dependent probability. We model transmission with a hierarchical probability structure where cross-infection coupling captures risk at behavioral interfaces between strategies. Comprehensive exploration of the four-dimensional parameter space reveals sharp phase transitions between cooperative and defective regimes. These transitions are governed by transmission intensity, recovery probability, risk perception, and economic pressures. A striking paradox emerges: while intense cross-infection coupling drives near-universal isolation adoption, it paradoxically sustains persistent endemic infection, demonstrating that widespread cooperation does not guarantee epidemic control. Modest changes in isolation costs or cross-infection coupling trigger complete phase inversions. This extreme sensitivity characterizes systems operating near critical points. Contact-mediated spread generates persistent spatial patterning in infection status and compartment composition. These findings establish epidemic-behavioral coupling as a fundamentally nonlinear dynamical system exhibiting critical phenomena and emergent spatial organization. Cooperation emergence does not guarantee epidemic control, revealing complex theoretical relationships between individual decision-making and collective health outcomes that require empirical validation for practical application.