Abstract Online gaming has grown from simple recreational activity into a vast, algorithmically engineered digital ecosystem that shapes human behaviour through immersive design, reward uncertainty, social hierarchies, and habit-forming engagement structures. With more than three billion active players globally and problematic gaming prevalence estimated at 6–12% among youth, the issue has moved from marginal concern to public health priority. This paper presents a comprehensive socio-psychological framework for understanding how immersive gaming environments interact with psychological characteristics across age groups, with particular sensitivity during adolescence. Drawing from behavioural psychology, developmental science, media studies, historical analysis, and emotional intelligence research, the paper identifies core engagement mechanisms including variable reward schedules, identity attachment, social hierarchy dynamics, habit-forming routines, and circadian disruption. Three original theoretical contributions are advanced. First, a dual-loop reinforcement model distinguishes stimulation-driven engagement — linked to sensory intensity and emotional arousal — from uncertainty-driven engagement — linked to probabilistic rewards and anticipation. Second, a multi-dimensional game taxonomy classifies games across environment, energy type, player count, and money involvement, introducing the physical energy thesis — that games requiring bodily effort preserve sportsmanship, while games requiring none allow entry without spirit — and the solo-AI danger thesis, which identifies single-player machine-mediated gaming as the highest-risk configuration. Third, a monetization corruption curve traces how gambling, as distinct from prize-based reward, systematically destroys the sportsmanship inherent in any game across historical eras. A central preventive argument proposes that Emotional Quotient (EQ) development, rather than exclusive focus on intellectual achievement, is the most sustainable protection against digital addiction. The paper demonstrates that high-IQ individuals with low EQ are paradoxically among the most vulnerable to structured reward systems, possessing the analytical capacity to optimize their engagement while lacking the emotional architecture to feel the relational and social costs of their choices. By shifting focus from individual blame to system-level interaction, this work contributes to interdisciplinary discussions on digital well-being across the lifespan and offers a structured foundation for future empirical research and policy development.