Skip to content
PUZZLEHEADSTUDIO
← All devlogs
STUDIOPROCESS

From prototype to playable puzzle

6 min readBy The Puzzlehead Team
NOTEBOOK · PROTOTYPE LOG P.34

Most mechanics are interesting for eleven minutes. The job of a prototype is to find out whether this one is interesting for eleven hours.

Every Puzzlehead prototype gets the same three gates before anyone is allowed to draw concept art for it. They are deliberately cheap, and deliberately brutal.

GATE ONE — THE UNEXPLAINED BUILD

The mechanic ships to a playtester with no instructions of any kind. If they cannot form a working theory within two minutes, the mechanic is not legible enough to carry a game. Not wrong — just not load-bearing.

NOTEBOOK · PROTOTYPE LOG P.34

GATE TWO — THE SECOND ROOM

We build one more space using the same mechanic and nothing new. If the second room is boring, the mechanic has one idea rather than a system. Blood Trials' core loop cleared this gate on the first attempt; four other prototypes did not and are still on the shelf.

A mechanic that only works in the room it was invented for is a set piece. We are not in the set piece business.

GATE THREE — THE HOSTILE PLAYER

Finally we hand it to someone whose explicit job is to break it. Not to find bugs — to find the boring optimal strategy. If a dominant strategy exists and it is dull, the mechanic goes back to the bench.

Roughly one prototype in six survives all three. That ratio has not moved in two years, and we have stopped trying to improve it.

  • #process
  • #prototyping