An experiment in futarchy. Inspired by Twitch Plays Pokemon (and Wordle, ofc)

Example Question

If we guess “IRATE” in round 2, will we finish this Wordle in 4 rounds?

• Alternatives

Description

Manifold Plays Wordle: An Experiment in Futarchy

Rules of the game:
1. Manifold generates 3 random words for a Wordle round
2. Manifold sets up a conditional market for each word
3. Traders have 1 hour to place their bets
4. The highest % word is chosen for the round; other markets resolve to N/A
5. If the Wordle is not finished, return to step 1
6. If the Wordle is finished, resolve each of the selected markets

Note: We play on <https://hellowordl.net/> so there isn’t any insider trading by people who have already done today’s Wordle. Candidate words are chosen from all possible words that satisfy Wordle's "Hard Mode" constraints.

Inspired by Twitch Plays Pokemon (and Wordle, ofc)

Round 1 candidates: STAGE, BONER, IGLOO
Result: STAGE ⬜⬜🟨⬜⬜
Round 2 candidates: HARPY, ADORN, VALOR
Result: ADORN 🟨⬜🟨⬜⬜
Round 3 candidates: MACHO, POLKA, VOILA
Result: MACHO ⬜🟨🟩⬜🟨
Round 4 candidates: COCOA, VOCAL, LOCAL

• First game: https://hellowordl.net/?challenge=Y29jb2E

Wordle 237 6/6

⬜🟨⬜🟨⬜
🟨⬜⬜⬜⬜
⬜🟨🟨🟨⬜
⬜🟨🟨🟨⬜
🟨🟨🟩🟨🟨
🟩🟩🟩🟩🟩

Rules of the game

1. Choose 5 different words each round

   A. Randomly from among all valid Wordle guesses

   B. Randomly from among all valid Wordle words

   C. Filtered for valid Hard Mode guesses

   D. Allowing users to submit guesses

2. Create a Manifold market for each of the words

   • Scalar markets with values between 1 (YES) and 7 (NO)
     • 1 = first guess, 6 = final guess, 7 = fail to finish
     • Equivalent linear mapping:
       • scalar = 7 - (6/100 * mkt%)
   • Or could invert where 7 (YES) and 1 (NO)?

3. Give everyone 10 minutes to trade

   • Close all markets at the end of 10 minutes

4. Select one of the markets to guess; immediately resolve the others to N/A

   A. Always the highest valued market

   • Pros: simple. Cons: market manipulation?

   B. Choose randomly among markets weighted by likelihood of success

   1. Weigh each market by 7 - scalar or 100 - mkt%
   2. More complex weighting scheme?

5. If the Wordle is not finished, return to step 1

6. If the Wordle is finished, resolve each of the selected markets

And that’s it! 🎉

TODOs

• [x] Find a Wordle clone that lets us reset to a different word
  • Play multiple times each day
  • Prevent insider trading for people who already did the Wordle
  • Decided: https://hellowordl.net/
    • Pros: Open source, varied letter length, can set seed
    • Cons: different word list
  • https://wordle.hannahmariepark.com/
    • Pros: Open source, pretty. Cons: Once per day
• [x] Find or build valid hard-mode guesses
  • Given list of