An Overview of Gambling-Related Psychological Features of Loot Boxes

Background/Aims: Loot boxes are randomised in-game monetisation mechanisms that share similarities with gambling. Although previous studies summarised the intricacy of loot box design based on random reward mechanism, there is no specific framework linking the characteristics of loot boxes with the gambling-related psychological factors. Therefore, this study aimed to explore loot box features and categorise them into different domains to analyse how these design elements incorporate gambling-related psychological principles to enhance player engagement and monetary expenditure in video games.Methods: Based on the Loot Box Features Model and researchers’ lived experience, we developed a codebook identifying 23 gambling-related features capturing six domains: (1) monetary manipulation & investment, (2) psychological manipulation, (3) sensory gambling cues, (4) fear-of-missing-out mechanics, (5) reward structure & probability systems, and (6) social dynamics. We conducted a coding exercise on the first hour of gameplay content from the 20 mobile games based on their grossing, excluding social casino games. Two researchers co-coded an initial subset separately and refined the framework before applying it to the remaining titles.Findings: All six domains appeared in the sample, with the highest prevalence in monetary manipulation, psychological manipulation, and fear-of-missing-out. Typical tactics included VIP schemes, limited-time discounts, variable-ratio reinforcement, rarity, and pity systems. The classic gambling resemblance was also found in some games. Social dynamics features showed the lowest prevalence.Conclusion: This study developed a framework to summarise gambling-related psychological features of loot boxes and initially explored the prevalence of these features in the highest-grossing games. We conclude with a discussion of potential association between these identified features of loot boxes and their potential risks on game players. The limitation of this study is the limited number of sampling games, and the first one-hour recordings are not enough to present full views of these games. Future work should audit longer play windows and validate the framework across regions.