High-performance cognition under uncertainty: lessons from poker and digital games

Abstract

Research on skilled performance has traditionally focused on domains where performance quality and outcomes are tightly coupled, such as chess or music. Many domains, however, involve genuine uncertainty — poker, digital games, and sports betting among them — where good decisions can produce poor outcomes, while poor decisions may occasionally be rewarded. I propose that high-performance cognition under uncertainty can be understood as a sequence linking relatively stable, latent decision-making capacity , its state-dependent expression in decision quality , and the observed outcome that follows. A regulatory layer comprising emotion regulation and metacognition influences how consistently capacity is translated into decision quality when outcomes provide unreliable feedback. Poker provides a model system in which epistemic uncertainty (hidden information) and aleatory uncertainty (randomness) can be distinguished, whereas digital games offer experimentally tractable environments for manipulating challenge and feedback. Within this framework, tilt illustrates regulatory disruption that can impair the expression of decision-making capacity. Flow is proposed more tentatively as an absorbed state through which successful regulation may help preserve decision quality. Because outcomes can influence subsequent emotional and metacognitive states, the framework is recursive across repeated decisions. Finally, absorbed states may serve different functions: task-oriented absorption may support performance, whereas avoidance-oriented absorption may be associated with maladaptive gaming or gambling. The framework shifts attention from how skilled performance is acquired to how decision quality is maintained when outcomes provide unreliable feedback.

The authors' abstract, as published at the source. Current Opinion in Behavioral Sciences, 2026 · DOI ↗