Ethics in educational game design: what to avoid

It is half past eleven at night and a fourth-semester student opens the review app her class has adopted. She had already studied enough for the day, but a notification warns that her 41-day streak will be lost at midnight. On the screen, a countdown timer, a blinking trophy and a closed chest promising a surprise prize if she gets ten more questions right. She gets eight right, misses two on purpose to finish quickly, opens the chest, earns a new profile icon and goes to sleep later than she wanted. The next day, she does not remember any of the questions.

Nothing in the scene is illegal, and each element, on its own, seems harmless. Together, they show a problem that medical education has only recently begun to discuss: the same design features that make a game engaging can be used to hold attention without teaching, or even against the interest of the person playing. This text explains what these patterns are, what studies from other fields say about them, how to recognize them in a teaching game and what an educational game that respects learners usually does.

Key points

• Dark patterns are interface choices that lead the user to decisions they would not make if fully informed, such as spending more time or money than intended (Newall, 2025).
• Unpredictable and frequent rewards are one of the mechanisms common to slot machines and to several digital products with potential for compulsive use (Clark and Zack, 2023).
• Paid reward boxes in games have been associated with gambling problems in studies with adults and adolescents (Zendle and Cairns, 2019; Zendle et al., 2019).
• Expected external rewards can reduce intrinsic motivation for a task, while positive, informative feedback tends to increase it (Deci et al., 1999).
• In medical education, gamification can improve learning when game elements encourage study behaviors, but few studies explain why or measure negative effects (van Gaalen et al., 2021; Aster et al., 2024).
• An ethical teaching game has a clear learning goal, feedback that explains, a pace the player controls, transparency about data and rules and no pressure to spend.

Why talk about ethics in the design of teaching games

Those who create an educational game usually start from a good intention: making study more pleasant and more frequent. The problem is that the tools available for this come, in large part, from the games and apps industry, where success is measured in screen time, daily return and revenue. Points, streaks, rankings, notifications and surprise prizes were refined to increase engagement, not learning. When they are transplanted into education without reflection, they bring their original goal along.

An analysis of gamification in health and fitness tracking apps concluded that these features raise ethical questions that, if ignored, are morally problematic and also harm users' health and engagement itself (Arora and Razavian, 2021). The same logic applies to teaching: a game that hooks without teaching fails twice, as an educational product and as a relationship of trust with the person studying.

Three starting questions

• Does this feature exist to help someone learn or to keep the person in the game?
• If the player understood exactly how the feature works, would they still accept it?
• Who benefits when the player spends more time here: the player or the person who made the game?

Medical students as an audience

Medical students are a particularly sensitive audience to these mechanisms. The study load is high, competition for residency is intense, and sleep and rest are usually the first things to give. A game that exploits the fear of losing a streak or dropping in the ranking acts on people who are already under pressure. The text about burnout in medical school describes the signs that this pressure has gone past the limit.

Engagement is not learning

A point that runs through this whole text is the difference between engagement and learning. Engagement is how much a person uses the game: frequency, session length, coming back the next day. Learning is what they can do afterward, without the game, faced with a new problem. The two can go together, and ideally they should. But it is possible to increase engagement with features that teach nothing, and it is possible to teach a lot in a short session with no frills. When an educational game reports only usage numbers, the part that matters most to anyone who studies it is missing: what changed in what the person knows and does.

What dark patterns are

The term dark patterns was created to describe interfaces that deceive or pressure the user. A taxonomy published in 2025, based on online betting platforms, separates three ideas that often appear together (Newall, 2025).

Harmful friction

It is the difficulty deliberately placed in the way of a choice that goes against the platform's interest. The classic example is canceling a subscription in several steps, while signing up takes one click. In betting, it appears in the withdrawal process (Newall, 2025).

Deceptive interface

It includes friction and other features, such as high suggested amounts, hidden decline buttons and fake countdowns. A scoping review on online betting mapped patterns such as urgency reminders, limit settings that are hard to find and ineffective play-time warnings, and noted that the effects on behavior are still poorly measured because the data stays with the companies (McGarrigle et al., 2026).

Dark nudges

They are ways of presenting the choice that lead the user in a direction without them noticing, such as showing a risky activity as harmless fun (Newall, 2025). A study on apps funded by the alcohol industry found this type of nudge, as well as incomplete information about risks, in tools that presented themselves as help to drink less (Roy-Highley et al., 2024).

Patterns that appear in educational games

Not every game feature is a dark pattern. The same element can serve learning or work against it, depending on how it is used. The table summarizes the most common ones.

Feature	Use that helps learning	Use that becomes a trap	What the studies say
Surprise reward	Rarely necessary	Chests and draws that keep the player trying	Variable and frequent reward is a common prerequisite of compulsive behaviors (Clark and Zack, 2023)
Paid boxes	None	Buying item draws with money	Spending associated with gambling problems (Zendle and Cairns, 2019; Etchells et al., 2022)
Points and prizes	Marking real progress	Rewarding only time spent	Expected rewards reduced intrinsic motivation in experiments (Deci et al., 1999)
Leaderboard	Optional, among peers who chose to compete	Public, mandatory, always visible	Some students found the live scoreboard stressful (Sharmin et al., 2025)
Time pressure	Training quick decisions when that is the goal	Countdowns only to rush	Time pressure is one of the typologies of manipulative design in children's apps (Radesky et al., 2022)
Character that nags	None needed	Sad mascot when the player leaves	Pressure through a relationship with a character found in children's apps (Radesky et al., 2022)
Feedback	Explaining why the answer is right or wrong	Only sound and animation, with no explanation	Feedback is one of the most used elements in serious games (Aster et al., 2024)

Unpredictable reward

A review on the addictive potential of rewards showed that the variability and frequency of rewards, central features of slot machines, also appear in games, online shopping, social media and other digital products linked to compulsive behaviors (Clark and Zack, 2023). In an educational game, the reward that helps is the predictable one tied to effort: you got it right because you used the right information, and the game shows that.

Reward boxes

Loot boxes are item draws inside the game, often paid. In a sample of more than a thousand players who did not know the aim of the study, spending on these boxes was associated with the severity of gambling problems (Zendle and Cairns, 2019). Among adolescents aged 16 to 18, the association was of moderate to large magnitude (Zendle et al., 2019). A Bayesian analysis with more than two thousand players confirmed the association with gambling problems, with no relation to mental well-being or psychological distress (Etchells et al., 2022). In a study on the gateway hypothesis, most participants who reported this effect had bought their first box before age 18 (Spicer et al., 2022). Cross-sectional studies do not prove cause, but the direction of care is clear: there is no place for paid draws in a teaching game.

Time pressure and characters

In a study of apps used by children aged three to five, researchers found four types of manipulative design to prolong use or encourage purchases: pressure through relationships with characters, time pressure, navigation restrictions and attractive lures (Radesky et al., 2022). Apps used by children from families with less schooling had more of these features (Radesky et al., 2022). The audience is different, but the tactics are the same ones that appear in study apps for adults.

Notifications and streaks

Notifications are useful for reminding someone of a routine they chose themselves. They become pressure when they use loss language, arrive during rest hours or multiply when the user disappears for a day. The review on online gambling describes urgency messages as one of the recurring patterns of deceptive design (McGarrigle et al., 2026). In study games, the equivalent is the warning that "everything will be lost" at midnight. A good practice is to let the player choose whether they want reminders, at what time and how often, and never use the loss of progress as a threat.

Reward, motivation and learning

There is a subtler question than compulsive use: even well-intentioned rewards can change the reason a person studies.

When the prize replaces the interest

A classic meta-analysis of 128 experiments showed that tangible and expected rewards, given for taking part, completing or reaching a performance level, reduced intrinsic motivation for the task, measured by the time people devoted to it by choice (Deci et al., 1999). Positive feedback, on the other hand, increased motivation and interest (Deci et al., 1999). For those who design teaching games, the implication is direct: trophies for mere use tend to teach that studying is worth it for the trophy, while feedback that shows real progress tends to strengthen interest in the content.

What gamification in health shows

A systematic review on gamification in the health professions found that it can improve learning outcomes, especially when game elements encourage study behaviors and attitudes; few studies explained the mechanisms, and most had no well-defined control groups (van Gaalen et al., 2021). Another review, on design elements in serious games in the field, found points, narrative and feedback as the most common elements and no study that assessed a specific element in isolation (Aster et al., 2024). In the Digital Health Education Collaboration review, adverse events were among the planned outcomes, but evidence on them was scarce (Gentry et al., 2019).

In other words, we know games can help and we know little about when they get in the way. That is one more reason to design with caution. The text on gamification in medical education sums up the benefits side, and the text on consistency, accuracy and speed discusses which metrics make sense to track.

Competition and ranking

Rankings motivate some people and make others uncomfortable. In a study with a real-time quiz platform in a dental hygiene class, most respondents said competition motivated them, and about a quarter found the live scoreboard stressful (Sharmin et al., 2025). A simple solution is to make the ranking optional or replace it with a comparison against one's own previous performance.

How to recognize a game that respects the learner

You do not need to know the theory to assess a game. Just observe what it does with your time, your attention and your mistakes.

Signs of respect

• The learning objective is clear, and the mechanics lead to it.
• The feedback explains why the answer was right or wrong, and does not just celebrate.
• The game has a natural end to the session, instead of always offering one more round.
• The rules and limits of use are written somewhere easy to find.
• Pausing or stopping carries no disproportionate penalty.
• The data policy explains what is collected and why.

Warning signs

Countdown timers that appear with no pedagogical reason, notifications that exploit the fear of missing out, random prizes, purchases inside draws, mascots that demand attendance and mandatory public rankings are signs that the game was designed to retain, not to teach. A single sign does not condemn the game, but several together deserve suspicion.

The next-day test

A practical way to assess is to ask yourself the next day: what did I learn yesterday in the game? If the answer is "I kept my streak" and not "I understood why that finding changes the diagnosis", the game is winning your attention without giving learning back.

Questions to ask whoever created the game

If the game was recommended by the medical school or by a professor, it is worth asking the responsible team how it was designed. Which learning objectives guided the choices? How are student data used, and who has access to them? Was the game tested with students before it was adopted, and did anyone measure unwanted effects, such as stress or sleep loss? Is there a channel for reporting problems? Teams that thought about these questions usually answer clearly. Vague answers do not prove bad intent, but they indicate that ethical questions were not at the center of the design.

Principles for those who create teaching games

Teachers, student leagues and product teams who create games for students can use the principles below as a starting point. They come from the literature cited and from the practice of those who design educational material.

Start with the objective

Every element of the game should have a pedagogical reason. If you cannot explain how the element helps learning, it is probably there to retain. A conceptual framework for choosing game elements based on theory was proposed precisely because most games in the field do not justify their own choices (Aster et al., 2024).

Prefer feedback to prizes

Show the player what they did, what was missing and how to improve. The meta-analysis on rewards suggests that this kind of return strengthens motivation, while prizes expected for use can weaken it (Deci et al., 1999). The text on feedback in clinical cases discusses what kind of explanation actually helps.

Give the player control

Allow pausing, leaving and coming back without losing learning progress. If there is a daily streak, consider allowing recovery or making the streak informative rather than punitive.

Be transparent

Explain the rules, the limits of use and what is done with the data. The ethical analysis of gamified health apps places on the designer the responsibility to anticipate adverse effects (Arora and Razavian, 2021).

Take care of accessibility

Fast animations, mandatory sounds and countdowns exclude people with visual, hearing or processing disabilities. The text on digital accessibility in medical software offers a guide.

Protect the data of those who study

Teaching games record correct answers, errors, times and response times. These data can help personalize study and can also expose the student, for example if they are shared with the institution without notice or used to assess performance without consent. Collect only what is necessary, say what each piece of data is for, let the person see and delete what is theirs and never turn someone's history of errors into an instrument of public exposure. Data protection laws, such as the LGPD in Brazil and the GDPR in the European Union, establish rights of the data subject that also apply to educational platforms.

Measure what matters

Engagement is easy to measure and misleading as a goal. Also measure retention, transfer to new cases and, if possible, unwanted effects such as stress and sleep loss.

How NEXO thinks about the pace of play

NEXO has a pace set by its own format: one clinical case per day, fictional, with a real clinical presentation, solved in up to six clues, with feedback by proximity in ICD-10 at five levels. The round has a beginning, a middle and an end, and the feedback shows how far each guess was from the diagnosis. The rules of use are described on the NEXO economy page, and what is done with the data is on the privacy page. To understand the whole round, see what NEXO is.

Ethics in game design is also an open research topic. Researchers who want to study the effects of diagnostic games, including unwanted effects, can propose studies through NEXO Researcher. Coordinators who want to use the cases with their class can learn about NEXO Closer, physicians who review cases take part through NEXO Curator, and those who create health content on social media can take part through NEXO Creators.

Frequently asked questions

Is gamification always unethical?

No. Game elements can improve learning when they encourage study behaviors (van Gaalen et al., 2021). The problem lies in using features to keep the user engaged against their own interest.

Is a day streak a dark pattern?

It depends on the design. A streak that reports regularity helps build a habit; a streak that punishes a pause with total loss and urgency notifications puts on pressure. Notice how you feel when you cannot play on a given day.

Does a ranking help or get in the way?

Both. Some students are motivated by competition and some get stressed by the live scoreboard (Sharmin et al., 2025). Optional rankings or comparisons with your own previous performance reduce the risk.

Why are reward boxes a problem?

Because studies with adults and adolescents linked spending on them to gambling problems (Zendle and Cairns, 2019; Zendle et al., 2019). In a teaching game, they have no pedagogical function.

How can teachers choose more ethical games?

Play before recommending, watch for the warning signs described above, read the rules and the data policy and ask the class, after a few weeks, what they learned and how they felt.

What should I do if I feel I cannot stop playing?

Turn off notifications, set a fixed time and talk to someone you trust. If your use is affecting sleep, study or relationships, seek the help you need at your college support service or in the health system.

Checklist for evaluating a teaching game

• The learning goal is clear.
• The feedback explains, and does not just celebrate.
• The session has a natural end.
• There is no paid draw and no random prize.
• Countdowns appear only when training speed is the goal.
• The ranking is optional.
• Rules, usage limits and data policy are easy to find.
• The next day, I can say what I learned.

Want a game with a beginning, a middle and an end every day? Play the case of the day on NEXO.

References

1. Arora C, Razavian M. Ethics of Gamification in Health and Fitness-Tracking. Int J Environ Res Public Health. 2021. https://pubmed.ncbi.nlm.nih.gov/34769570/
2. Radesky J, Hiniker A, McLaren C, et al. Prevalence and Characteristics of Manipulative Design in Mobile Applications Used by Children. JAMA Netw Open. 2022. https://pubmed.ncbi.nlm.nih.gov/35713902/
3. McGarrigle J, Smith J, Griffiths J, et al. Dark patterns in online gambling: A scoping review and classification of deceptive design practices. J Behav Addict. 2026. https://pubmed.ncbi.nlm.nih.gov/41493466/
4. Newall P. Sludge, dark patterns and dark nudges: A taxonomy of online gambling platforms' deceptive design features. Addiction. 2025. https://pubmed.ncbi.nlm.nih.gov/40302118/
5. Clark L, Zack M. Engineered highs: Reward variability and frequency as potential prerequisites of behavioural addiction. Addict Behav. 2023. https://pubmed.ncbi.nlm.nih.gov/36701907/
6. Zendle D, Cairns P. Loot boxes are again linked to problem gambling: Results of a replication study. PLoS One. 2019. https://pubmed.ncbi.nlm.nih.gov/30845155/
7. Zendle D, Meyer R, Over H. Adolescents and loot boxes: links with problem gambling and motivations for purchase. R Soc Open Sci. 2019. https://pubmed.ncbi.nlm.nih.gov/31312481/
8. Etchells PJ, Morgan AL, Quintana DS. Loot box spending is associated with problem gambling but not mental wellbeing. R Soc Open Sci. 2022. https://pubmed.ncbi.nlm.nih.gov/35991334/
9. Spicer SG, Fullwood C, Close J, et al. Loot boxes and problem gambling: Investigating the "gateway hypothesis". Addict Behav. 2022. https://pubmed.ncbi.nlm.nih.gov/35397261/
10. Deci EL, Koestner R, Ryan RM. A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychol Bull. 1999. https://pubmed.ncbi.nlm.nih.gov/10589297/
11. van Gaalen AEJ, Brouwer J, Schönrock-Adema J, et al. Gamification of health professions education: a systematic review. Adv Health Sci Educ Theory Pract. 2021. https://pubmed.ncbi.nlm.nih.gov/33128662/
12. Aster A, Laupichler MC, Zimmer S, et al. Game design elements of serious games in the education of medical and healthcare professions: a mixed-methods systematic review of underlying theories and teaching effectiveness. Adv Health Sci Educ Theory Pract. 2024. https://pubmed.ncbi.nlm.nih.gov/38563873/
13. Sharmin N, Wazir A, Chow AK. Kahoot! in a dental hygiene class: Motivating or stressful for students? Can J Dent Hyg. 2025. https://pubmed.ncbi.nlm.nih.gov/40160410/
14. Roy-Highley E, Körner K, Mulrenan C, et al. Dark patterns, dark nudges, sludge and misinformation: alcohol industry apps and digital tools. Health Promot Int. 2024. https://pubmed.ncbi.nlm.nih.gov/39377424/
15. Gentry SV, Gauthier A, L'Estrade Ehrstrom B, et al. Serious Gaming and Gamification Education in Health Professions: Systematic Review. J Med Internet Res. 2019. https://pubmed.ncbi.nlm.nih.gov/30920375/