Solo Testing Your Prototype

To check whether the game *functions* — every rule executable, turns have meaningful choices, and the game reaches an ending — not whether it's fun.

Human playtests cost social capital; friends will only sit down with a half-finished game so many times. Solo testing costs only your own afternoon.

Playing each side with a fixed, written-down simple policy, never letting a side use information it shouldn't have — stress-testing whether rules resolve cleanly.

Greedy maximizer, turtle (defensive), spoiler (denies opponent), and randomizer (rolls to pick among legal moves).

If a random player is competitive with a greedy player, your choices probably don't matter — a serious design problem.

After each move, asking aloud 'If I were trying to break this game, what would I do now?' to catch degenerate strategies early.

A turn with zero/one legal option or where all options are obviously equivalent. If more than ~15% of turns are dead, the decision space is too thin.

The gap between the turn the winner was effectively decided and the turn the game actually ended; it should be near zero.

Undefined states, dead turns, dominant strategies, length problems, and bookkeeping friction.

No interaction, a dominant card, and first-player advantage — change only ONE thing, then re-run the log.