The four-hour game

Rules

Four hours, from idea to deploy. When the time is up it ships as it is. The moment polishing turns it into two days, this stops being the same exercise.

No external assets. Art and sound are generated in code. Start browsing asset packs and two of the four hours disappear there. The constraint is what makes it fast.

A round has to be short. Over in thirty seconds, and immediately restartable. Short rounds are also faster to verify while building.

If it does not get finished, that gets written down. A failed round is still a round, and knowing what was misjudged makes the next one faster.

What should I build?

Suggest the next theme. A game you want to see, a feature that should exist, something that annoys you about what is already here — one line is enough. No sign-in, no email.

Suggestions are not published. This is a place to listen, not a board that displays other people's writing, so the page only shows how many have come in. Anything I act on gets described in that round's post in my own words. I do not collect contact details, so I cannot reply individually — if you need an answer, use contact instead.

Rounds

Round 1 · Stop It — press when the needle crosses the green arc. Every hit narrows it and speeds the needle up.

The other games are in the app list. Those were built without a clock.

Why four hours

Not a day, and not an hour. An hour only gets an idea onto the screen, which is a demo rather than a game. Give it a day and polishing starts, and polishing has no end — the next round never arrives.

Four hours turned out to be exactly enough to do one thing properly and give up the rest. Without deciding what the core is first, nothing finishes in time. So this format measures judgement more than it measures speed.

How it gets built

The tools change per round. Something browser-based ends up as a single canvas; something 3D gets an engine. But because deploying is inside the four hours, anything with a slow build filters itself out.

The write-up is not just the result. It records what was cut and whether that call was right, including the cuts that turned out to be mistakes — that is the only thing that carries over to the next round.