Gamification for Mental Wellbeing: Lessons From Building GameHarbor

Gamification helps a wellbeing app only when it makes calming activities easier to start, not when it pressures people to come back. The evidence says points, badges and streaks are not the active ingredient in mental health apps, so responsible teams design for autonomy, short sessions with a natural end, no guilt mechanics and an AI companion with firm boundaries. This article explains that evidence and the design rules we set for GameHarbor, our calm, harbour-themed arcade of 62 short games. It is about method, not outcomes: GameHarbor is a wellbeing app, not therapy, and we make no claim that it treats or diagnoses anything.

What is gamification, and how is it different from a serious game?

Gamification adds game elements to something that is not a game; a serious game is a whole game built for a purpose beyond fun.

Gamification means points, levels, badges, progress bars and streaks inside a product that is not itself a game. A serious game is a complete game designed for a purpose such as learning a skill or practising a therapeutic technique. A calm game, the category GameHarbor sits in, is a short, low-pressure game meant to be relaxing rather than competitive.

The distinction matters because the evidence differs. A systematic review by Cheng and colleagues in JMIR Mental Health (2019) catalogued 50 gamified mental health and wellbeing apps and technologies. Levels or progress feedback appeared in 80 percent of them, followed by points and rewards. The authors warned that many researchers treat gamification as a "black box" without considering how it works.

What does the research say about games and wellbeing?

The research supports good therapeutic content and well-designed play, but not game mechanics as a shortcut to better outcomes.

• Gamification alone is not the active ingredient. A 2021 meta-analysis by Six and colleagues in JMIR Mental Health pooled 38 studies with 8,110 participants. Mental health apps produced a small to moderate reduction in depressive symptoms, but gamification was not a significant predictor of symptom change or of adherence.
• Purpose-built games can be tested properly. SPARX, a 3D fantasy game built on cognitive behavioural therapy techniques, was compared with usual care for 187 adolescents seeking help for depression in a 2012 BMJ randomised trial and performed as well as usual care on the main symptom measure. That is a clinical intervention with clinical testing, a different category from a wellbeing app.
• Motivation matters more than minutes played. An Oxford Internet Institute study of about 39,000 players, published in Royal Society Open Science in 2022, found little evidence that time spent playing affects wellbeing, and found that motivations for playing do play a role.
• Games meet basic psychological needs. Research by Ryan, Rigby and Przybylski in Motivation and Emotion (2006), based on self-determination theory, linked game enjoyment to feelings of autonomy, competence and relatedness more than to simple fun.

Our reading: points and streaks are not what makes a wellbeing experience work. Autonomy, gentle competence and a calm, welcoming space are better design targets.

What design rules did we set for GameHarbor?

We wrote four rules before building any game: short sessions, no streak guilt, no dark patterns and choice over prescription. They are design principles we hold ourselves to, not measured results.

1. Short sessions with a natural end

GameHarbor's 62 games are short by design, and the intent is that a session finishes rather than rolling into the next one by default. Infinite feeds and autoplay are engagement tools; a calm product should be comfortable letting people leave.

2. No streak guilt

Streaks exploit loss aversion: people keep going to avoid losing what they built, not because the activity helps. Victor Ayomipo's February 2026 Smashing Magazine analysis describes the tipping point as the moment a streak makes users anxious or guilty and becomes something they feel they cannot afford to lose. In a wellbeing product that contradiction is unacceptable, so our principle is to welcome people back without punishing absence.

3. No dark patterns

Dark patterns, also called deceptive patterns, are interface designs that steer people into choices they did not intend. A 2025 study of Calm, Headspace and Insight Timer by Özkürkçü and Doğanay, summarised on Deceptive Design, found manipulative design shaped how participants behaved in and experienced those apps. The rules we follow:

• No countdown timers, fake scarcity or "your progress will be lost" warnings.
• No leaderboards that compare a stressed person with strangers.
• No notifications written to trigger anxiety or guilt.
• Settings and exits as easy to find as the play button.

4. Choice over prescription

Players pick what to play and when, and breathing exercises are offered rather than forced. This follows the self-determination research above: autonomy supports wellbeing in a way external rewards do not.

How does engagement-first design differ from wellbeing-first design?

Engagement-first design maximises return visits and time spent; wellbeing-first design maximises how people feel when they leave. The same feature can be built either way.

How should an AI companion behave in a wellbeing app?

An AI companion must be honest that it is AI, step aside for human help in a crisis, and avoid fostering dependence.

GameHarbor includes Ollie, an AI companion (she/her) who can chat by text or voice and guide breathing exercises. A conversational companion is the most powerful and the most sensitive part of any wellbeing product. On 13 November 2025 the American Psychological Association issued a health advisory on generative AI chatbots and wellness apps: it says these tools should not substitute for care from a qualified professional, that their ability to guide someone in crisis is "limited and unpredictable", and that unhealthy dependence should be prevented.

The boundaries we design for:

1. Honest identity. Ollie is presented as an AI companion, never as a therapist, counsellor or clinician.
2. Crisis signposting. If a conversation suggests risk of harm, the companion should stop being playful and point to human help. In India, the government's Tele MANAS helpline is free and available 24x7 on 14416 or 1800-89-14416, in more than 20 languages. A global app needs region-aware signposting.
3. Healthy reliance. The companion encourages real-world support, rest and connection rather than positioning itself as the only place to turn.
4. Calm notifications. Any check-in should be occasional, gentle and easy to switch off.
5. Reporting and moderation. Google Play requires apps that generate content with AI to include in-app reporting of offensive AI output without leaving the app. Treat it as a safety feature, not a compliance box.

How do you test crisis flows?

Write a set of test conversations covering direct and indirect expressions of distress, in English and the regional languages you support, and run them after every model or prompt change. Check that the companion stops playful tone, gives the right helpline for the user's region and never argues or diagnoses. Log failures and fix them before release.

What should wellbeing apps avoid claiming?

Describe what the app does, such as short calming games and breathing guidance, and never what it cures, unless you have clinical evidence and the right clearances.

Google Play's health apps declaration separates "Stress Management, Relaxation, Mental Acuity", a health and fitness category, from "Mental and Behavioral Health", a medical category for counselling and treatment tools. Pick the declaration that honestly matches your features. Our app store launch checklist covers the declarations step by step.

Privacy deserves the same care. Wellbeing conversations are sensitive, so collect the minimum, explain it plainly and make deletion easy. India's DPDP Rules, notified in November 2025, bring most obligations into force 18 months after publication, as Storyboard18 reported; see our DPDP compliance guide.

Common mistakes to avoid

• Store copy that says "reduces anxiety" or "treats stress" without evidence.
• Push notifications that guilt people for missing a day.
• A chatbot persona that implies professional training.
• Collecting mood data "for later" without a clear purpose or consent.

How we approach wellbeing products at Old Port

We treat wellbeing as a design constraint from day one rather than a feature to bolt on. In practice that means writing the no-guilt and no-dark-pattern rules into the product brief, reviewing every notification and reward against the comparison table above, scripting crisis test conversations before the AI companion ships, and keeping store copy descriptive. We also prefer publishing our thinking openly so others can challenge it. Our UAE colleagues have written about building wellbeing apps responsibly from the same projects.

Where can you read more about GameHarbor?

We have published a research preprint about GameHarbor on Zenodo. A preprint has not been peer reviewed, and we present it as a description of the work rather than as evidence of clinical benefit. You can try GameHarbor on the web today, with Android and iOS builds, and read more on our UAE studio's GameHarbor page. Our sister project MindHarbor takes a slower, open-world approach, described in our article on calm game design.

If you are designing a wellbeing, education or habit product and want play without manipulation, our app development, UI/UX design and AI solutions teams can help, or get in touch to talk it through.