Inspiration
Most games test how good you are.
Very few test how well you understand yourself.
In real life, performance and confidence don’t always match. We overestimate. We underestimate. And that gap shapes decisions far beyond games.
We wanted to build something simple but revealing — a daily challenge that measures not just memory, but self-calibration.
Not “Are you smart?” But “Do you know how you’ll perform?”
That question became Wrong Turn.
What it does
Maps: Wrong Turn is a daily global spatial memory challenge.
Every player receives the same 30x30 grid route.
Before starting, you answer one question:
How confident are you? (0–100%)
You then get 5 seconds to memorize the path — including turns and symbol markers.
After it disappears, you reconstruct it step by step.
When you finish, you see:
Where you made your first wrong turn
A visual comparison of your route vs the correct one
Your Confidence Error
We calculate:
$$Confidence Error = Predicted Confidence – Actual Performance$$
Players are ranked not only by accuracy — but by how closely their confidence matched reality.
The headline leaderboard is:
Who Knows Their Limits
It reframes competition from raw performance to calibrated performance.
How we built it
We focused on fairness, clarity, and scalability.
Frontend
React
Tailwind CSS
Devvit UI components
Backend
Reddit Devvit platform
Redis for:
Daily deterministic map storage
Player attempts
Sorted-set leaderboards
Calibration ranking
Core systems
Scheduled global daily reset
Deterministic map generation (same for everyone)
Validation engine for user-generated maps (5–15 turns, boundary-safe)
Community filtering (Trending, Hardest, Calibrated, New)
The architecture supports both daily global competition and a growing UGC ecosystem.
Challenges we ran into
Designing a calibration metric that feels intuitive and fair
Preventing trivial or exploitative user-generated maps
Balancing difficulty across a global player base
Ensuring confidence prediction feels meaningful — not cosmetic
Delivering feedback that’s immediate and clear without overwhelming the player
The hardest part wasn’t building the grid logic.
It was making self-awareness measurable.
Accomplishments that we're proud of
Turning metacognition into a core competitive mechanic
Launching a calibration-first leaderboard
Building a fully validated community map system
Creating transparent global stats (confidence gaps, common mistakes)
Designing a daily shared cognitive benchmark
We’re especially proud that performance isn’t the only thing that matters — understanding your performance does too.
What we learned
Many players are consistently overconfident — and are surprised by it.
Calibration creates deeper engagement than raw success rate.
Shared daily challenges drive repeat participation.
Simplicity in design increases cognitive focus.
A small rule change (predict first, then perform) completely changes player psychology.
We learned that awareness adds a second layer to competition.
What's next for Maps: Wrong Turn
ELO-style confidence rating
Adaptive difficulty based on calibration history
Personal confidence trend tracking
Seasonal calibration leagues
AI-generated maps tuned to individual blind spots
Expanded community analytics dashboard
Long term, we see Wrong Turn evolving into a daily cognitive benchmark — one that measures not just performance, but judgment.
Log in or sign up for Devpost to join the conversation.