An Overview of Underlying Gambling-Related Psychological Features of Loot Boxes

This study develops a framework linking loot box features in video games to established gambling-related psychological mechanisms, positioning itself as a complement to the existing LoBoF v0.1 model, which describes loot box systems procedurally rather than by their psychological function. Motivated by evidence that loot box engagement is associated with problematic gambling and that non-gamblers' loot box spending can predict later gambling initiation, the authors built a codebook of 23 features grouped into six psychologically grounded domains: monetary manipulation and investment, psychological manipulation, sensory gambling cues, fear-of-missing-out mechanics, reward structure and probability systems, and social dynamics. The codebook was pre-registered and drew on the LoBoF framework, gambling psychology literature, and the researchers' own experience.

The framework was applied to the first hour of gameplay in 20 top-grossing mobile games on Google Play, excluding social casino titles, with two coders establishing inter-rater reliability on an initial subset before one reviewer coded the remainder. All six domains appeared in the sample. Currency conversion and monetary investment were present in every game, while limited discounts, VIP schemes, variable-ratio reinforcement, and rarity tiers were each found in 85 to 90 percent of titles. Losses disguised as wins and near-miss effects were also common, and features frequently co-occurred, suggesting that these games deploy several psychological mechanisms cumulatively rather than relying on any single one.

The authors position the framework as a tool for theory-driven hypothesis generation rather than descriptive auditing, and they note that it is designed to be applicable across time, platforms, and cultures. Its limitations include a one-hour observation window that may miss features unlocked through extended progression or seasonal events, an inability to assess haptic feedback from video recordings, and the fact that the psychological functions assigned to each feature remain untested. The authors call for future work to examine whether these mechanisms actually shape player cognition and spending, to use longer play sessions, and to validate the framework across regions.