Quick winning principles
- Clue words must be orthogonal—no shared roots across your three clues per round.
- Never reuse a clue concept that opponents saw you use in round 1.
- Interceptor teams log your clue-to-position mapping after first intercept.
- Misdirect with clue 3: make it look like a repeat pattern while clue 1-2 stay clean.
Ideal strategy
- 1
Entropy independence: each clue should partition the keyword space without overlapping semantic fields. If keywords share a domain, use number/texture/color axes instead of topic.
- 2
Interceptor timing: intercept on round 2 when you have 2 data points per team—round 3 intercept is often too late with only one guess left.
- 3
Black cell exploitation: when you must clue the black position, choose a word that could plausibly map to two different opponent keywords—forces wrong guesses.
- 4
Opponent modeling: track which teammate gives the loosest clues; opponents will target that player's patterns first.
- 5
Endgame lock: at 1 misdecode from loss, switch to ultra-literal clues even if intercept risk rises—one black is better than two.