Loot boxes and disguised gambling: the hidden risk in video games

Introduction: From Random Monetization to Concerns About Digital Gambling

Loot boxes are digital items purchased to obtain randomly determined rewards. Their resemblance to conventional gambling stems from the combination of payment, uncertainty, and the possibility of receiving rare or valuable items. Health and regulatory concerns arise when this mechanic is integrated into games aimed at minors, involves real money, or allows rewards to be transferred outside the game. On Steam, 35 games were identified that featured paid loot boxes and transferable content with monetary value, even though they likely violated gambling laws in several jurisdictions (Xiao & Henderson, 2024). Recent literature links their purchase to problem gambling, financial difficulties, and psychological distress, although opinions differ on whether the boxes cause subsequent gambling or reflect a pre-existing vulnerability (Vuorinen et al., 2025; Wilson et al., 2024). The issue therefore requires distinguishing between structural similarity, statistical association, and causality.

The Psychological and Economic Architecture of Loot Boxes

Randomness, variable rewards, and psychophysiological responses similar to those of slot machines

The architecture of loot boxes replicates several behavioral principles associated with slot machines: uncertain outcomes, rewards of unequal rarity, sensory cues, and the expectation of an exceptional prize. An analysis of 20 top-grossing mobile games classified 23 gambling-related features into six domains, including monetary manipulation, psychological manipulation, sensory cues, fear of missing out, probability systems, and social dynamics (Zheng et al., 2025). In that sample, monetary and psychological mechanisms appeared most frequently, along with limited-time offers, VIP schemes, variable-ratio reinforcement, rarity, and progressive reward systems (Zheng et al., 2025). The analysis describes the design but does not demonstrate that each feature causes harm to players.

Experimental evidence suggests a different connection. In a study involving loot box users, gamblers, and controls, opening loot boxes and spinning a simulated slot machine increased electrodermal activity in the first two groups, while heart and respiratory rates did not change significantly (Sanmartín et al., 2024). The arousal in response to both activities was similar among gamblers and loot box users, while the control group reacted to the slot machine but not to the loot box (Sanmartín et al., 2024). The sample was small, consisting of 40 participants, and the primary indicator was electrodermal activity, not subsequent spending behavior. Therefore, the finding supports a limited psychophysiological similarity, not clinical equivalence between the two activities.

The design may also support repetition driven by commercial urgency and social comparison. The review of features identifies time-limited discounts, rarity rewards, and progress cues as common components, but acknowledges that analyzing only the first hour of play does not capture the full experience (Zheng et al., 2025). The evidence thus supports a hypothesis involving reinforcement and attention, although it does not yet establish which combinations of elements cause the greatest harm or in which players.

Virtual Currencies, Willingness to Pay, and Risk-Taking Spending Mechanisms

Virtual currencies separate the act of paying from the act of opening the box, which can reduce the immediate perception of cost. In a randomized experiment with 753 participants from the United Kingdom, the willingness to pay for boxes presented as risky and ambiguous lotteries was higher when the transaction used virtual currency with a one-to-one conversion rate than when it was expressed in British pounds (Toccafondi et al., 2025). The study observed monetary illusion distortions regardless of whether the exchange rate was intuitive, but found no evidence that non-intuitive rates altered willingness to pay compared to intuitive rates of comparable magnitude (Toccafondi et al., 2025). These results support an effect of the monetary format, not a general explanation based on any complex conversion.

Spending is also linked to probability, rarity, and the possibility of recouping value. Some games incorporate transferable items, so that a virtual reward can acquire monetary value outside the platform (Xiao & Henderson, 2024). In this context, the loot box more clearly resembles a regulatable wager, although legal definitions continue to depend on whether there is actual payment, randomness, and transferable value. The direct purchase of non-random items allows us to distinguish the effect of digital spending from the specific effect of uncertainty: among young adults who did not gamble initially, spending on loot boxes and risk scores predicted subsequent initiation, whereas direct microtransactions did not when controlling for loot box spending (Palmer et al., 2025). The distinction between digital purchases and random rewards will therefore be necessary to assess economic and regulatory risks.

Relationships Between Loot Box Purchases, Problem Gambling, and Associated Harms

Associations with the severity of problem gambling and financial difficulties

The association between loot boxes and problem gambling emerges across various study designs, but its magnitude and scope depend on the population studied. In a stratified national sample of regular gamblers who also played video games, both the frequency of purchases and the amount of money spent on loot boxes were associated with the severity of problem gambling and with gambling-related harms, even after controlling for other forms of gambling (Kim et al., 2026). In an international survey, spending on loot boxes was also linked to problem gambling, while spending on certain physical card packs showed a weaker association (Xiao et al., 2025). These studies focus on people who were already gambling or on survey samples, so they do not allow us to conclude that loot boxes caused their problems.

The harms are not limited to conventional gambling. In a community sample from a British island, 5.9% reported having purchased boxes during the previous year; purchasing them was not associated with PGSI scores, but was associated with a higher likelihood of financial problems and poor mental well-being, with odds ratios of 3.39 and 2.99, respectively, after adjusting for sociodemographic variables and problem gambling (Wilson et al., 2024). This finding contrasts with an international study that found no association between spending on similar products and poor mental health (Fernández & Quagliato, 2025). The difference may reflect the fact that one community-based sample measures financial difficulties and well-being in the general population, while the other study focuses on gamblers and uses different mental health indicators.

Links to alcohol consumption during gambling sessions have also been observed. Among participants in Australia, Aotearoa, and the United States, drinking while gambling was weakly or moderately associated with higher spending and risky use of boxes, but problematic alcohol use did not show such an association (East et al., 2025). The finding suggests that the context of the gaming session may matter more than a general measure of consumption, although the cross-sectional design prevents establishing the direction of causality.

Indirect Pathways Between Problematic Video Game Use, Loot Boxes, and Real-Money Gambling

Mediation models describe loot boxes as a possible link between video game problems and monetary gambling. Among 542 gamers recruited from Spanish schools, internet gaming disorder was associated with problematic loot box use and online gambling disorder; the direct effect between the two disorders was not significant, whereas the indirect effect via problematic use of loot boxes was significant (González‐Cabrera et al., 2024). Full mediation was also observed in the subgroups of minors and young adults (González‐Cabrera et al., 2024). These results are consistent with an intermediate pathway, but the cross-sectional design and cross-sectional measurement do not prove a causal sequence.

A Romanian study of 703 frequent gamblers found a similar pattern: problematic engagement with slot machines mediated the association between problem gambling and problem online gambling (Huțul et al., 2024). Among Flemish adolescents, a two-wave longitudinal model showed that initial interaction with gambling-like elements indirectly predicted subsequent gambling behavior through attitudes and intentions (Grosemans et al., 2025). The second study broadens the analysis to include various game elements, not just gacha systems, and therefore does not allow the entire effect to be attributed to a specific mechanic.

The most direct prospective evidence comes from young adults in British Columbia. Among those who did not gamble at the outset, spending on loot boxes and the risk index predicted both the initiation of gambling and gambling expenditure six months later, whereas spending on direct microtransactions did not predict initiation (Palmer et al., 2025). In contrast, a Finnish population-based follow-up study of 949 adults between 2021 and 2024 found positive effects of microtransactions on gambling problems that disappeared when gambling participation was added to the model (Vuorinen et al., 2025). The discrepancy may stem from the fact that the Canadian study isolated random rewards and the Finnish study grouped microtransactions together, but both show that prior participation in gambling is a central alternative explanation.

Impulsivity, depression, anxiety, stress, and mental well-being as moderating factors

Psychological factors can increase the likelihood of risky spending or intensify its relationship with other behaviors. Among 1,416 adults, risky box purchases were associated with depression, anxiety, stress, and impulsivity, as well as problem gambling and problematic video game use (Villalba-García et al., 2025). Anxiety and impulsivity strengthened the association with symptoms of problem gambling, while impulsivity and depression moderated the association with symptoms of online problem gambling (Villalba-García et al., 2025). Stress was associated with risky purchasing but did not moderate these relationships (Villalba-García et al., 2025).

Among 1,358 Polish players who had opened loot boxes in the previous year, risky use was associated with frequency of use, spending, symptoms of gambling disorder, gambling problems, and perceived stress; it was also higher among women and decreased with age among female players (Lelonek-Kuleta et al., 2026). Motivations related to achievement and escapism were associated with risky use (Lelonek-Kuleta et al., 2026). These findings describe vulnerability profiles, not individual causes, and suggest longitudinal models that distinguish between psychological predisposition, commercial design, and spending consequences.

Adolescents, Commercial Exposure, and Vulnerability to Gambling Functions

Adolescent Motivations, Risk Attitudes, and Sociodemographic Differences

Adolescence combines early exposure, a search for belonging, and limited financial autonomy. A systematic review of studies on minors found a consistent positive association between the use of betting terminals and gambling tendencies, although only four studies met the inclusion criteria, and higher levels of engagement and risk were observed among adolescent males (Fernández & Quagliato, 2025). In interviews with Australian youth aged 18 to 24, the initial motivations were primarily social and linked to enhancing the gaming experience; continued participation was associated with social and psychological rewards, and to a lesser extent, economic rewards (Culbong et al., 2026). Motivation, therefore, is not limited to the expectation of monetary gain.

Longitudinal data from 561 Flemish adolescents show that interaction with gambling-like elements was associated with gambling attitudes, norms, intentions, and behaviors, and that the former indirectly predicted behavior one year later (Grosemans et al., 2025). The study is not limited to slot machines, but it suggests that the normalization of gambling practices may operate through cognitive changes before becoming established as spending.

Surreptitious advertising, in-game purchases, and the normalization of gambling

Commercial exposure is linked to in-game purchases and problem gambling. Among 2,595 seventh-grade students in Taiwan, 94% played online games, 38% made in-game purchases, and 9% played simulated gambling games; greater exposure to marketing, advertising influence, and in-game purchases, as well as less parental supervision, were associated with problem gaming and psychological distress (Chang et al., 2024). The cross-sectional design does not allow for separating the influence of advertising from content selection by young people who are already interested.

Advertising compliance is particularly limited. Among 2,358 game ads featuring loot boxes analyzed using the Meta repository, only 8.4% of ads from the United Kingdom and 7.6% of those from South Korea disclosed the presence of loot boxes; given the required visual prominence, actual compliance dropped to 2.4% and 4.2%, respectively (Xiao, Zendle, et al., 2025). Exposure can thus present gambling mechanics as a routine part of entertainment, without sufficient information on how they work.

Regulatory Protection, Legal Noncompliance, and Limitations of Youth Protection Policies

Existing measures reveal a gap between regulation and enforcement. A review of youth policies concludes that rules regarding loot boxes and monetary spending in Europe and Asia suffer from poor compliance and enforcement, while playtime restrictions can be easily circumvented (Xiao et al., 2026). On Steam, the presence of loot boxes with transferable content remained unchanged one year after their identification, despite their potential incompatibility with gambling laws (Xiao & Henderson, 2024). Age ratings also do not guarantee that young people and families will follow the recommendations (Xiao et al., 2026).

Protection requires regulating access, advertising, the transparency of odds, and spending, with special attention to minors. Current evidence does not allow us to determine which regulatory combination most effectively reduces harm, because many policies have not been evaluated using comparable behavioral outcomes. It does show that self-regulation without sufficient sanctions fails to address commercial exposure and access to potentially illegal products (Xiao et al., 2026; Xiao, Zendle, et al., 2025).

Critical Synthesis: What Can Be Said About the Transition from Loot Boxes to Gambling

Convergence among experimental studies, surveys, mediation models, and longitudinal analyses

The literature converges on three observations: loot boxes incorporate gambling features, their purchase is associated with gambling problems, and certain profiles show greater vulnerability. This convergence is strongest when methods are combined, because each design addresses a different aspect of the problem.

Canadian prospective evidence and the Flemish adolescent model support a possible transition, but they are not equivalent: the former studies young adults and distinguishes between direct purchase boxes, while the latter groups gambling elements and uses attitudes and intentions as mediating variables (Grosemans et al., 2025; Palmer et al., 2025). The Finnish results suggest that microtransactions may function as a marker of gambling participation rather than as an independent cause (Vuorinen et al., 2025). The reasonable conclusion is that there is a prospective association consistent with migration, but its magnitude and specificity remain unestablished.

The Debate on Causality, Migration to Gambling, and Alternative Explanations

The migration hypothesis proposes that exposure to or spending on gacha games precedes and facilitates monetary gambling. The temporal precedence observed in Canada and Belgium is consistent with this hypothesis (Grosemans et al., 2025; Palmer et al., 2025). However, pre-existing vulnerability may explain part of the relationship: those who already gamble are more heavily involved in microtransactions, and factors such as impulsivity, anxiety, stress, or depression are associated with risky use (Lelonek-Kuleta et al., 2026; Palmer et al., 2025; Villalba-García et al., 2025). There may also be a third explanation, according to which high overall engagement with video games simultaneously increases exposure to loot boxes and gambling.

Therefore, the “gateway” effect should be treated as a hypothesis, not as a proven fact. Statistical mediation does not replace the identification of individual changes over time, and the association between loot boxes and financial harm may reflect pre-existing economic vulnerability. Future research will need to compare trajectories with groups exposed to direct purchases, loot boxes without transferable value, and the absence of randomized monetization.

Research Gaps and Limitations of the Available Evidence

Lack of longitudinal designs, comparable measures, and representative samples

The evidence remains limited by cross-sectional surveys, incidental sampling, and heterogeneous definitions of purchase, risky use, and problem gambling. Although there are follow-ups at six months, one year, and several years, they remain few in number and involve distinct populations, ranging from schoolchildren to regular gamblers and young adults (Grosemans et al., 2025; Palmer et al., 2025; Vuorinen et al., 2025). The measures are also not comparable: some studies record money spent, while others record frequency, symptoms, PGSI scores, or self-reported purchasing. The review of minors included only four studies, which limits international generalizability (Fernández & Quagliato, 2025).

Need to assess cumulative harms, commercial design, and regulatory effectiveness

Research must track spending, debt, well-being, and gambling over sufficiently long periods and distinguish between exposure, purchases, and financial losses. It should also examine entire gaming sessions—not just the first hour—to capture time-limited offers, progressive rewards, and social dynamics (Zheng et al., 2025). Policy studies should measure actual compliance, behavioral changes, and distributional effects among minors and vulnerable groups. Without independent evaluations of penalties, transparency, and spending limits, it is impossible to know whether a regulation reduces harm or merely alters the product’s presentation.

Conclusion: Recognize the risk without confusing association with causation

Loot boxes combine chance, payment, variable rewards, and commercial cues that resemble gambling mechanisms. Experimental, psychophysiological, cross-sectional, and longitudinal studies consistently link them to willingness to pay, problem gambling, financial difficulties, and specific vulnerability profiles (Sanmartín et al., 2024; Toccafondi et al., 2025; Wilson et al., 2024). Some follow-up studies support a subsequent transition to monetary gambling, while others indicate that prior gambling participation explains much of the link (Palmer et al., 2025; Vuorinen et al., 2025). The regulatory response must specifically protect minors and enforce advertising and commercial compliance, without treating the issue of causality as settled. Recognizing the risk means intervening in design, access, information, and spending while longitudinal studies with comparable measures are conducted.

References

• Chang, F., Chuang, W.-C., Rutherford, R., Chen, W.-Y., Yang, C.-Y., Chiu, C., Chen, P., Miao, N., & Chuang, H. (2024). Associations between marketing exposure, in-game purchases, problem gaming, simulated gambling, and psychological distress among adolescents. Journal of Gambling Studies, 41(1), 187–201. https://doi.org/10.1007/s10899-024-10353-z
• Culbong, H., Masters, A. B., Noble‐Carr, D., & Suomi, A. (2026). Why do young people engage with gambling features in video games: rewards and motivations. Journal of Youth Studies, 1–18. https://doi.org/10.1080/13676261.2026.2690206
• East, L. C., Sauer, J. D., Lowe-Calverley, E., Hall, L. C., Martin, B. J. E., & Drummond, A. (2025). Associations between alcohol consumption and spending on gambling-like mechanisms in video games. Scientific Reports, 15(1), 32014. https://doi.org/10.1038/s41598-025-15301-4
• Fernández, P., & Quagliato, L. A. (2025). The association between loot boxes and gambling in children and adolescents. Trends in Psychiatry and Psychotherapy. https://doi.org/10.47626/2237-6089-2025-1217
• González‐Cabrera, J. M., Caba‐Machado, V., Díaz‐López, A., Jiménez‐Múrcia, S., Mestre‐Bach, G., & Machimbarrena, J. M. (2024). The mediating role of problematic use of loot boxes between internet gaming disorder and online gambling disorder: A cross-sectional analytical study. JMIR Serious Games, 12, e57304. https://doi.org/10.2196/57304
• Grosemans, E., De Cock, R., Bradt, L., & Soenens, B. (2025). From gambling-like elements in and around video games to monetary gambling: testing a pathway model among adolescents. International Gambling Studies, 25(3), 472–494. https://doi.org/10.1080/14459795.2025.2567896
• Huțul, A., Holman, A. C., & Huțul, T. (2024). Are today’s gamers tomorrow’s gamblers? The relationship between problem gaming and online problem gambling, and the indirect effect of the risk associated with purchasing loot boxes. Cyberpsychology, Behavior, and Social Networking, 28(2), 132–137. https://doi.org/10.1089/cyber.2024.0319
• Kim, H. S., Allami, Y., Christensen, D. R., Hodgins, D. C., McGrath, D. S., Nicoll, F. J., Shaw, C. A., Stevens, R. M. G., & Williams, R. J. (2026). Loot box purchases are associated with problem gambling severity and harms beyond traditional gambling activities. Addictive Behaviors, 182, 108783. https://doi.org/10.1016/j.addbeh.2026.108783
• Lelonek-Kuleta, B., Cudo, A., David, J., & Bartczuk, R. (2026). Sex differences, perceived stress, gaming motives, and gambling-like attitudes toward loot boxes predict risky loot box use among Polish gamers. Scientific Reports, 16(1). https://doi.org/10.1038/s41598-026-46637-0
• Palmer, L., Brooks, G. A., & Clark, L. (2025). A longitudinal replication study testing migration from video game loot boxes to gambling in British Columbia, Canada. BMC Psychology, 13(1), 459. https://doi.org/10.1186/s40359-025-02766-1
• Sanmartín, F. J., Velasco, J., Cuadrado, F., & Moriana, J. A. (2024). The thrill of chance: Psychophysiological responses to loot boxes and simulated slot machines. Computers in Human Behavior Reports, 16, 100545. https://doi.org/10.1016/j.chbr.2024.100545
• Toccafondi, N., Di Paolo, R., & Di Guida, S. (2025). Virtual currencies in online gaming increase the willingness to pay for loot boxes: an experimental analysis. Experimental Economics, 28(6), 1262–1280. https://doi.org/10.1017/eec.2026.10049
• Villalba-García, C., Griffiths, M. D., Demetrovics, Z., & Czakó, A. (2025). The relationship between loot box purchases, gambling, internet gaming, and mental health: Investigating the moderating effect of impulsivity, depression, anxiety, and stress. Computers in Human Behavior, 166, 108579. https://doi.org/10.1016/j.chb.2025.108579
• Vuorinen, I., Savolainen, I., & Oksanen, A. (2025). Pay to play or pay to lose? The impact of in-game spending in digital games on gambling problems. Journal of Gambling Studies, 42(2), 685–695. https://doi.org/10.1007/s10899-025-10453-4
• Wilson, C., Butler, N., Quigg, Z., & Sumnall, H. R. (2024). Associations between loot box purchasing and gambling behaviors, financial problems, and low mental well-being in a household sample from a British island: a cross-sectional study. Journal of Public Health, 34(5), 1187–1195. https://doi.org/10.1007/s10389-024-02355-3
• Xiao, L. Y., & Henderson, L. L. (2024). Illegal video game loot boxes with transferable content on Steam: a longitudinal study on their presence and non-compliance with and non-enforcement of gambling law. International Gambling Studies, 24(3), 498–524. https://doi.org/10.1080/14459795.2024.2390827
• Xiao, L. Y., Czakó, A., Deery, C., Heath, C., Király, O., King, D. L., & Demetrovics, Z. (2026). Global review of assessments of online video gaming policies affecting young people: Loot boxes, age ratings, gameplay time restrictions, and beyond. Comprehensive Psychiatry, 148, 152713. https://doi.org/10.1016/j.comppsych.2026.152713
• Xiao, L. Y., Deery, C., Petrovskaya, E., Park, S., & Newall, P. W. S. (2025). Widespread illegal video game advertising in the UK and South Korea: Many ads found not to disclose the presence of loot boxes using Meta’s ad repository. Journal of Behavioral Addictions, 14(2), 714–723. https://doi.org/10.1556/2006.2025.00057
• Xiao, L. Y., Zendle, D. I., Petrovskaya, E., Nielsen, R. K. L., & Newall, P. W. S. (2025). Spending on physical card packs and, especially, video game loot boxes is positively correlated with problem gambling but not linked to negative mental health: An international survey. Psychology of Addictive Behaviors, 40(1), 66–85. https://doi.org/10.1037/adb0001082
• Zheng, L., Xiao, L. Y., & Sharman, S. (2025). An overview of gambling-related psychological features of loot boxes [Preprint]. Center for Open Science. https://doi.org/10.31234/osf.io/zd7ab_v1