Too Fun to Put Down: How We Decide What to Build

Every Dirtcube project starts as a playable prototype. If the team can't stop playtesting it, we greenlight it. If they can, we kill it. Here's the method behind the madness.

Ideas are cheap. Everyone has them. The hard part is knowing which ones deserve a year of your life — and the honest truth is you usually can't tell from a pitch deck. You can only tell from playing.

The bar is embarrassingly simple

Every project begins as a rough, ugly, playable prototype — built in weeks, not months. Then we watch what happens. Does the team keep 'just testing it one more time'? Do people fight over whose turn it is? If a half-broken prototype is already too fun to put down, that's the signal. That's a greenlight. If we have to talk ourselves into it, that's a no.

We make the things that become them — not the things that chase them.

This sounds obvious, but most of the industry does the opposite: it greenlights on a market thesis and hopes the fun shows up later. We've watched that fail too many times. Fun is the hardest thing to add in post. So we make it the entry requirement, not the finishing touch.

Prototype cheap, kill fast

The point of a two-week prototype isn't to build the game — it's to buy information as cheaply as possible. A prototype that dies in a fortnight cost us a fortnight. A production that dies in year two cost us a team and a runway. Killing quickly isn't failure; it's the whole discipline. It's what protects the ideas that deserve the years.

Own the stack

Once something is greenlit, we control the technology so it never controls us — from custom editor tools that halve iteration time to our own Specter backend running under every title. Owning the stack is what lets a small team move like a big one. When your backend is a product you already built, standing up a new game's economy is a configuration, not a project.

And launch day is milestone one, not the finish line. In a live-service world, a game worth building is a game worth supporting for years — patching, balancing, and listening to the people actually playing it. The greenlight is just the beginning of the conversation.