Hypothesis · Experiment · Revision
Scientific Method Games
Why Rulehunt plays like an experiment, not a quiz.
Most puzzle games ask you to recall or compute. A scientific method game asks something different: you are handed a system whose law is hidden, and the only way forward is to guess at the law, run an experiment, and let the result tell you whether your guess survives.
Rulehunt is built on exactly that loop. Every day a secret rule decides which playing card may follow the sequence. You never see the rule. You see only which cards it accepts and which it turns away.
The loop, step by step
- Observe. Three cards are dealt that already obey the hidden rule. That's your starting evidence.
- Hypothesise. Read those cards for a pattern — colour, rank, suit, alternation, a relation to the card before.
- Experiment. Play the card from your hand that best distinguishes between two competing hypotheses, not the card you are most confident about. A test that can only confirm you teaches nothing.
- Read the result. Accepted, and the card joins the sequence. Rejected, and that's a misfire — but a rejection often rules out more possibilities than an acceptance does.
- Revise. Narrow the rule and test again. Place eight cards before four misfires and the rule is revealed in full.
Negative results carry information
The instinct in most games is to avoid being wrong. In a scientific method game being wrong is how you make progress: a disconfirmed prediction eliminates a whole family of rules at once, which a lucky correct guess never does.
Rulehunt gives you four misfires precisely so that falsification is affordable. The budget is the lesson.
Where the idea comes from
The design descends from Eleusis, a card game invented by Robert Abbott in 1956 and popularised by Martin Gardner in his Scientific American column. Abbott's explicit aim was to make a game whose play imitated inductive science — the dealer holds a secret rule of nature and the players are the researchers.
Rulehunt keeps that aim and removes the setup: no dealer, no explaining rules to a table, one puzzle a day, about three minutes. Everyone gets the same hidden rule on the same day, so comparing notes afterwards is part of the point.
Useful in a classroom
Because the rule is a formal statement and the feedback is unambiguous, a single game makes a clean demonstration of hypothesis, prediction, test, and revision — and of the difference between a hypothesis that fits the data and one that has actually been tested against it.
It's free, needs no account, and runs in a browser, so a class can all play the same day's rule and compare reasoning afterwards.