Every game we build starts life as a scene full of grey cubes. No art, no sound, no UI — just the core mechanic and a build we can put in someone’s hands by Friday.
Why so ugly?
Polish is a lie detector’s worst enemy. Beautiful art makes a mediocre mechanic feel better in a playtest, and that false positive costs months. Strip the game to cubes and there’s nowhere for the boredom to hide: if testers replay the loop without being asked, we have something. If they check their phone, we don’t — and it only cost us a week to find out.
The prototype’s job isn’t to look like the game. Its job is to answer one question: would anyone play this twice?
The loop we run
- One mechanic, one week. Build the smallest playable version of the idea.
- Hands-on Friday. Everyone in the studio plays it — plus at least one person who’s never seen it.
- Keep, twist, or kill. Fun? Production candidate. Almost? One more week with one change. Neither? It goes in the graveyard, and the graveyard is allowed to be big.
Most ideas die in week one, and that’s the system working. The ones that survive have already proven the only thing that matters — the fun is in the mechanic, not in the promise.