I remembered some game played in my philosophy class that was supposed to be about science, so I took it and turned it into a more Bayesian form.
- Players sit on the opposite side of the dealer in a row with a deck of cards, paper, and pencils.
- The dealer chooses a secret pattern and looks through the deck to find a card that matches it.
- Shuffle deck. Each player gets seven cards.
- Everyone grabs a blank piece of paper/marks a new round, then writes all hypotheses for the pattern and percent chance of correctness for each hypothesis.
- If anyone says ‘I know the pattern’ then they state what they think it is. If they are wrong, they don’t get to participate in future rounds, but if they’re right the game ends. (and they’ll get extra points?)
- Starting on the far player’s left and working rightwards, the first person sets down a card. The dealer states if this does or does not match the pattern. If it matches, then the new card is placed next to the last one. If not, it is discarded. The player draws a new card to a full seven.
- When the game is over, you add up the percentages for the correct hypothesis from each round and multiply by 100. That’s how many points you get.
- Total points (including potential bonus for correct guesser) to see who won.