Skip to main content
Formal Specification

Buzz (Greek Slap and Guess Game) — Formal Specification

Greece · guessing-hidden · Confidence: high

← Back to game page

Implementation Suitability

Suitability: engine-readyFidelity: directGap-fills: 1
# Biz — AI-optimized specification ## General idea A discrete, hidden-action deception and deduction game played between one blinded guesser ("Mana") and a circle of candidate attackers. One circle agent secretly executes a strike action; subsequently, all circle agents emit noisy behavioral signals (bluffing features). The guesser must decode the signals and select the true striker to transfer the central role. ## Formal model - **Abstraction taken:** Physical facial expressions, micro-movements, and posture ticks are abstracted into a vector of discrete scalar signals: `[auditory_noise, tell_intensity, baseline_composure]`. - **Player/agent count:** 1 Guesser (G) and K Circle Agents (C1..CK, K∈[3..7], default K=4). - **Turn model:** Round-based. Each round comprises (1) Secret striker selection, (2) Signal emission / bluff generation, (3) Guesser deduction. ## Formal rules R1. Setup: Central agent G is blind (`eyes_open = False`). Circle agents C1..CK form the candidate set. R2. Strike Selection: - Circle agents select exactly one striker i∗∈{1..K} via a decentralized bidding or random coordinator. R3. Signal Generation Phase: - Each circle agent i emits an observable signal feature vector: Si={Sample(𝒩(μguilt,σ2))if i=i∗Sample(𝒩(μinnocent,σ2))if i≠i∗ where agents with higher `bluff_skill` shift μguilt closer to μinnocent. R4. Guessing Phase: - Guesser G opens eyes (`eyes_open = True`) and observes the vector of signals {S1,S2,…,SK}. - Guesser G selects one candidate index iˆ∈{1..K}. R5. Resolution: - If iˆ=i∗ (Correct): Round ends; agent Ci∗ becomes the new Guesser G. - If iˆ≠i∗ (Incorrect): Guesser G incurs a failure count (`consecutive_fails += 1`) and remains Guesser for the next round. R6. [gap-fill] Mercy Rule / Elimination Cap: If `consecutive_fails == 5`, the role of Guesser automatically transfers to the lowest-scoring circle agent to prevent infinite entrapment. ## Edge cases - Tie at the end condition: Not applicable — each round produces a binary success/failure evaluation for the active guesser. - A player with no legal move: Guesser must always select a single index iˆ∈{1..K}. - Repetition / cycles: Bounded by R6 (maximum 5 failed rounds per guesser). - Simultaneous claims on the same resource or position: Exactly one striker i∗ is selected per round by construction. - Player elimination mid-round; turn order repair: Not applicable (no eliminations; role exchange only). - Resource exhaustion: Not applicable. - Termination guarantee: Enforced by the bounded retry cap (R6) and non-zero probability of correct identification (P≥1/K). ## State representation - Persistent state: - `guesser_id`: Index of current guesser. - `consecutive_fails`: Integer counter. - `round_history`: List of `(striker, guess, success)` tuples. - Per-round ephemeral state: - `secret_striker`: Hidden integer i∗. - `emitted_signals`: Array of K floats [s1,…,sK]. - Observable information: Guesser observes only `emitted_signals` and historical guesses. Secret striker identity i∗ is private to the circle agents until resolution. ## Action space - `select_striker()`: (Circle action) Choose i∗∈{1..K}. - `emit_bluff(effort)`: (Circle action) Modulate signal emission. - `accuse(candidate_idx)`: (Guesser action) Output iˆ∈{1..K}. ## Playability notes - Single-player-vs-AI: Highly effective Bayesian deduction challenge for a human player against AI bots generating simulated behavioral signals. - Group games (n>4): Naturally scales from K=3 to K=10 circle agents; larger circles reduce baseline random guessing accuracy from 1/3 to 1/10. - Physical-skill games: Physical slapping and suppression of giggles replaced by signal detection theory. - What is lost: Organic social hilarity of watching friends try not to laugh. ## Implementation hints - Model Guesser as a Bayesian classifier updating belief priors over candidate guilt. - Circle agents optimize their bluffing parameters using basic game-theoretic mixed strategies.

Download as JSON

Download the full game corpus entry for this game as structured JSON.

Download JSON
Cite this page

Buzz (Greek Slap and Guess Game) (Greece). (n.d.). Games of Countries. https://gamesofcountries.com/games/gr-greece/biz

@misc{gamesofcountries-gr-greece-biz,
  title = {Buzz (Greek Slap and Guess Game) (Greece)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/gr-greece/biz}
}