I Studied 500+ Gamified Apps. Here’s What Actually Works. (video)
A pattern study by Tim Gabe (designer, ex-Spotify, founder of Zipsnap) surveying hundreds of gamification implementations across mobile products. The framing: the three mechanics every app reaches for first — points, badges, leaderboards — are the three most documented failures in product history. The video walks seven research-backed patterns, contrasting failed implementations (LinkedIn, Foursquare, Habitica, Snapchat) with successful ones (Strava, Apple Watch, Peloton, Duolingo).
The thesis in one sentence
PBL is the scoreboard of a game, not the game itself. Most apps build the scoreboard and forget to build the game. — Yu-kai Chou (quoted)
The seven patterns below are alternatives to scoreboard-building.
Pattern 1 — The PBL fallacy
Points, Badges, Leaderboards are the default first move. They’re also the documented failure mode. See PBL fallacy for the standalone concept.
- LinkedIn (2024): quietly retired Community Top Voice gold badges. Internal read: badge-motivated users produced quantity over quality, chasing the badge instead of sharing expertise.
- Foursquare (2014): scrapped mayorships and badges after data showed gamification drove check-ins but not the discovery behavior the business actually needed.
- Google News: killed its badge system for the same reason.
Pattern 2 — Hyper-local micro-competitions (the Strava counterexample)
Strava has 180 million registered users and an engagement number Tim flags as more like influence than retention: users average ~1 hour of real-world activity for every 2 minutes spent in the app.
The mechanism doing the work is segments — any user-defined stretch of road or trail. Run it, your time gets logged on a leaderboard sorted by age and gender cohort. This sounds like the global leaderboard the PBL fallacy condemns, but it isn’t:
Strava didn’t build one global leaderboard, they built thousands of hyper-local micro-competitions.
The hill on your morning route. The segment outside your office. Winnability is the strongest predictor of competitive motivation (cited: 2022 Science Direct study). A leaderboard you can never crack demotivates. One where the local pack is realistic to chase activates competition.
The social side compounds:
- Kudos (the small social acknowledgement when someone completes a run) increase future run frequency.
- Strava clubs grew 59% in 2024.
- The platform handed out 14 billion kudos in 2025.
Takeaway: engineer the size of the competition rather than inflating empty metrics.
Pattern 3 — The S-curve problem (feature richness)
A 2025 peer-reviewed study in Frontiers in Psychology found that gamification feature richness follows an S-shaped curve. Adding features helps engagement up to a point. After that point, more features actively reverse engagement.
The empirical proof: Habitica, probably the most aggressively gamified productivity app ever built — daily tasks become quests, habits become character stats, missing a task damages your HP. The peer-reviewed study on Habitica found that 100% of participants experienced counterproductive effects. Users got so absorbed in managing the game layer that the actual productivity behavior got buried.
If you’re stacking streaks plus points plus badges plus challenges plus leaderboards, the data says you’re past the peak of the curve, not climbing it.
See Gamification S-curve for the standalone concept.
Pattern 4 — The streak trap
Research from The Decision Lab (cited): streaks gradually shift from motivational to obligational the longer they run. Users move from “I want to do this” to “I can’t miss today.”
The canonical case is Snapchat streaks. A 2023 Belgian study of ~2,500 adolescents found streak frequency correlates with FOMO, problematic smartphone use, and reduced social media self-control. In 2024 the Nevada Attorney General filed litigation against Snapchat. The EU Digital Fairness Act, heading toward a legislative proposal in late 2026, is specifically aiming at addictive streak mechanics.
The video flags Duolingo as the contrast — they’ve run hundreds of streak experiments and ship one that retains, but notice how:
- Users choose their goal level.
- Streak freeze can be bought.
- The mechanic is wrapped in agency.
A streak you can’t pause, can’t influence, and can’t escape is the design researchers are currently flagging.
See Streak for the standalone concept.
Pattern 5 — Variable reward magnitude (anticipation, not loss aversion)
Streaks run on fear of losing what you built. Variable rewards run on the pull towards what’s next. Same surface behavior, opposite emotional engine. One burns out while the other keeps recharging itself.
The video’s worked example is Gameblazers, a fantasy card game (designed by Tim’s studio Sipsap/Zipsnap). The pack-opening flow has three engineered stages:
| Stage | What happens | Emotional engine |
|---|---|---|
| Anticipation | Tap a pack, no idea what’s inside | Pull toward unknown reward |
| Reveal | Cards flip one at a time | Each card resets the anticipation cycle — one dopamine event becomes five |
| Celebration | Rare card hits, screen reacts, glows, haptics | Closure + dopamine spike → tap the next pack |
This pattern is the design corollary of Variable ratio reinforcement (Skinner) and Reward prediction error (Schultz). The video doesn’t name those — it reframes them as “pull towards what’s next” vs the loss-framed streak engine.
Tim’s studio also designed for Freecash (a rewards platform that has paid out €300M+).
Pattern 6 — Completion drive (the Apple Watch insight)
Tim singles this out as probably the most important point in the video.
Apple Watch activity rings — three rings (Move, Exercise, Stand) — drove a 49.5% behavior change in 160,000 people. The mechanism is one principle from cognitive psychology: the Gestalt principle of closure.
The brain is hardwired to perceive incomplete patterns as demanding completion. A 90% filled circle creates an open loop, and the brain wants to close it.
Downstream outcome: users who regularly close rings are 48% less likely to experience poor sleep quality. Real-world positive output, not engagement theater.
Apple’s rings work because they make you want to finish what you started. Anticipation toward closure. That’s the engine.
See Completion drive for the standalone concept.
Pattern 7 — Competence vs badge theater
A 2024 Springer Nature meta-analysis on gamification found something Tim calls brutal:
Gamification reliably improves a user’s perception of autonomy and relatedness, but has minimal impact on competence — the one psychological need most tied to long-term intrinsic motivation.
This maps onto the three psychological needs of Self-Determination Theory (Deci & Ryan). Most apps engineer recognition and forget to engineer mastery.
The clean counterexample is Peloton. Members who use the social and output features work out 15% more frequently. But the mechanism isn’t competition — it’s competence feedback:
- Output measured in real time.
- Personal records auto-flagged.
- A “100 ride badge” means something because it represents 100 actual rides — evidence of skill development, not opening the app a lot.
Other examples in the same family:
- Chess.com — ELO ratings.
- Garmin — training readiness and body battery scores.
Build mechanics that signal you got better at the actual thing, not mechanics that signal you opened the app a lot.
Cross-pattern summary
The seven patterns split cleanly into fails / works pairs:
| Failure mode | Working alternative |
|---|---|
| Global leaderboards (Pattern 1) | Hyper-local micro-competitions (Pattern 2) |
| Feature stacking (Pattern 3) | Restraint past the S-curve peak |
| Streak coercion (Pattern 4) | Streak with agency, or no streak |
| Loss-framed retention | Variable reward magnitude (Pattern 5) |
| Recognition (badge theater) | Completion drive (Pattern 6) |
| Recognition (badge theater) | Competence feedback (Pattern 7) |
Notable numbers
| Claim | Number | Source cited |
|---|---|---|
| Apple Watch ring-closure behavior change | 49.5% | (160k-person study; source not named in video) |
| Ring-closers vs sleep quality | 48% less likely to have poor sleep | (not cited in video) |
| Strava users worldwide | 180M | (not cited) |
| Strava real-world : in-app ratio | ~1 hour : 2 min | (not cited) |
| Strava club growth, 2024 | +59% | (not cited) |
| Strava kudos given, 2025 | 14B | (not cited) |
| Habitica counterproductive effect rate | 100% of study participants | (peer-reviewed; source not named) |
| Belgian Snapchat-streak adolescent study | ~2,500 participants | 2023 |
| Peloton output/social engagement uplift | +15% frequency | (not cited) |
| Freecash payouts | €300M+ | (Tim’s own client) |
Tim Gabe’s commercial frame
Tim runs Zipsnap / Sipsap (the video uses both spellings — likely the captions; the studio appears to be one entity). The video repeats a CTA three times for “free design strategy calls each month” at zipsnap.com. Disclosed clients include Gameblazers and Freecash. Tim worked at Spotify previously.
Notable references
- Yu-kai Chou — gamification author and consultant; the “scoreboard of a game” quote
- LinkedIn, Foursquare, Google News — PBL fallacy case studies
- Strava, Apple Watch (Activity), Peloton, Chess.com, Garmin, Duolingo, Gameblazers — working examples
- Habitica, Snapchat — failure-mode case studies
Concepts introduced
PBL fallacy · Streak · Completion drive · Self-Determination Theory · Gamification S-curve
Reinforces existing concepts
- Variable ratio reinforcement — anticipation framing, the three-stage pack-opening flow
- Reward prediction error — the dopamine engine under Pattern 5
- Fear of missing out — Snapchat streaks and the regulatory landscape
- Loss aversion — streaks as the loss-framed retention engine, contrasted with anticipation-framed variable rewards
- Habit-vs-surprise dilemma — Pattern 5 sits squarely on the “surprise” leg; streaks sit squarely on the (degraded) habit leg
Open questions
- The Apple Watch 49.5% / 48% sleep numbers are stated but not sourced. The 160k-person study is plausibly real but unverified here.
- The “500+ gamified apps” sample is the framing; no methodology is given (which apps, how scored, what “gamified” means).
- The “1 hour : 2 minutes” Strava engagement-vs-influence stat is striking but not citation-attributed.
- Some claims (Strava clubs +59%, 14B kudos in 2025) are forward-dated, suggesting Tim is citing recent platform disclosures the video doesn’t link.
Sources
raw/gamification-patterns.txt— YouTube auto-captions cleaned to text (yt-dlp, 2026-05-23)- Original: https://www.youtube.com/watch?v=LXX_qOA5D8E