Integrating Curiosity and Uncertainty in Game Design

The document argues that curiosity, long treated as a single construct in game design, should instead be understood as five distinct forms: perceptual, manipulatory, complexity/ambiguity-based, conceptual, and adjustive-reactive. Drawing on Loewenstein's information-gap theory, it proposes that each form responds to a different kind of uncertainty, and that games can deliberately trigger these forms through hidden information, incomplete maps, ambiguous mechanics, and narrative anticipation. The central thesis is that existing player-engagement models, such as Lazzaro's Four Keys and the MDA framework, miss this differentiation and would benefit from integrating curiosity theory with Costikyan's typology of uncertainty.

The authors distinguish stable trait curiosity from situational state curiosity and contend that both shape how players respond to uncertainty. A key finding is that tolerance for uncertainty depends less on the size of a knowledge gap than on a player's confidence that the gap can be closed. Feedback mechanisms, such as "warmth" indicators in puzzles, therefore make uncertainty more enjoyable by raising that confidence. Examples across the analysis include Pokémon Blue, Halo 3, Monument Valley, Don't Starve, Civilization III, World of Warcraft, Clue, NetHack, Dragon Age: Origins, and Waking Mars, spanning a range of genres and platforms.

The work's conclusions are framed with appropriate caution. Perceptual and adjustive-reactive curiosity are identified as especially useful for making exploration satisfying and for guiding puzzle solutions. The authors propose that games could eventually detect players' trait and state curiosity levels and adjust uncertainty accordingly, but they present this adaptive approach as a proof of concept requiring further empirical validation. The scope is primarily theoretical and design-oriented, drawing on the academic literature on curiosity, flow, and game enjoyment rather than on original player data.