Every World Cup we repeated the same ritual: a shared spreadsheet, predictions by hand and, without fail, someone arguing about how the points were counted. It sort of worked. So I decided to build it properly.
The goal
A simple web app where each friend enters their match predictions and the leaderboard updates itself. No spreadsheets, no arguments about scoring: the rules are applied by the code.
Why Supabase
For a project among friends I didn't want to stand up a whole backend. Supabase gave me exactly what I needed:
- Managed Postgres for users, matches and predictions.
- Ready-to-use auth, no need to build login from scratch.
- An instant API over the tables.
The front is plain JavaScript (no heavy framework, none needed) and it's all deployed on Vercel.
What's interesting
- Scoring logic lives in one place. Exact result, correct winner… each case scores differently, and since it's in code, nobody argues.
- Pick the right tool for the size of the problem. Not everything needs microservices; sometimes Supabase + plain JS is the fastest, most maintainable answer.
- Projects with your friends are the best testing ground: they actually use it and tell you what breaks instantly.
For the technical summary (stack, decisions, status), see the project page.