Formal Specification
Blind He-Goat (Blind Man's Buff; traditional Norwegian blindfolded tactile tag game) — Formal Specification
Norway · chase-tag · Confidence: high
← Back to game pageImplementation Suitability
Suitability: adaptation-requiredFidelity: discretizedGap-fills: 8
# Blindebukk — AI-optimized specification
## General idea
Blindebukk ("Blind He-Goat") is an asymmetric hidden-information chasing and social deduction game where one blindfolded player navigates by sound and tactile contact to capture and correctly identify opponents. The structural distinctiveness lies in its two-stage resolution: physical containment alone is insufficient; the seeker must deduce the identity of the captured agent via noisy physical/auditory feature signals.
## Formal model
- **Abstraction taken:** The blindfold is modeled as zero visual range with noisy directional acoustic and tactile sensors. Physical touching is modeled as entering an adjacent grid cell, and feeling facial/clothing features is modeled as receiving a masked feature vector subject to sensory observation noise.
- **Player/agent count:** 4 to 20 agents (1 Goat, Dodgers). Standard model: 5 agents.
- **Turn model:** Simultaneous discrete movement ticks (), followed by an atomic interrogation/deduction step upon capture.
## Formal rules
R1. **Initial Setup:** The game space is a bounded 2D grid of size . At , the Goat starts at the center with orientation randomized [gap-fill: modeling the 3 initial spins]. Dodgers are initialized at distinct random coordinates units from the Goat.
R2. **Sensory Emissions:** Each tick, any Dodger moving at speed generates an acoustic emission of magnitude . The Goat observes a noisy bearing angle for the loudest nearby emission [gap-fill].
R3. **Simultaneous Movement:** Each tick, the Goat moves 1 unit in any orthogonal direction. Dodgers may move up to 1 unit or choose to `stay_silent` (0 units, producing zero acoustic emission).
R4. **Capture Condition:** If the Goat enters the cell occupied by a Dodger , Dodger is `HELD` and cannot move.
R5. **Identification Step:** Upon capture, the Goat receives a noisy observation vector of Dodger 's static profile and outputs a guessed agent ID .
R6. **Role Transition:** If (correct guess), Dodger becomes the new Goat for the next round, and the round terminates. If (incorrect guess), Dodger is released to an adjacent cell, the Goat suffers a temporary 2-tick stun penalty [gap-fill], and the chase resumes.
R7. [gap-fill] **Round Timeout:** If ticks elapse without a successful capture-and-identification, the round resets with the Goat retaining the role.
## Edge cases
- **Tie at the end condition:** Not applicable; identification resolution is sequential per captured individual. If two Dodgers occupy the cell the Goat enters, the Goat captures the one with lower ID [gap-fill].
- **A player with no legal move:** Boundary collisions block movement; Dodgers trapped in corners must either attempt to slip past or remain stationary.
- **Repetition / cycles:** Bounded by round limit [gap-fill] and bounded match rounds .
- **Simultaneous claims on the same resource or position:** Multiple Dodgers may share a cell; capture is evaluated against the lowest-index occupant first [gap-fill].
- **Player elimination mid-round; turn order repair:** No elimination; roles transfer upon successful identification.
- **Resource exhaustion:** No consumable resources.
- **Termination guarantee:** Each round strictly terminates within at most ticks. The match terminates after rounds.
## State representation
- **Persistent state:** `goat_id` , `round_scores` (array of rounds as Goat per agent), `profiles` (matrix of static feature vectors ).
- **Per-round state:** `tick` , `phase` .
- **Per-agent state:** `pos` , `held` .
- **Observable information:** Dodgers observe full grid positions. The Goat observes only own position, boundary collisions, and noisy directional sound bearings .
## Action space
- `move(dx, dy)`: Step vector .
- `stay_silent()`: Remain stationary with zero acoustic signature.
- `guess_identity(target_id)`: Goat action upon triggering `IDENTIFICATION`.
## Playability notes
- **Single-player-vs-AI:** Natural as an asymmetric stealth game: player as Goat using sound cues and profile deduction against heuristic bots, or player as sneaking Dodger evading an AI Goat.
- **Group games (n>4):** Reduces to a 2-agent duel by having 1 Goat pursue 1 Dodger on a smaller grid () with the identification phase replaced by deducing a randomized disguise profile [gap-fill].
- **Physical-skill games:** Tactile identification is modeled via statistical classification over Gaussian-perturbed attribute vectors (height, hair texture, clothing).
- **What is lost in the adaptation:** Tactile warmth, physical laughter, parlor camaraderie, and vocal mimicry.
## Implementation hints
- Partially Observable Markov Decision Process (POMDP) for the Goat agent.
- Suggested first milestone: Implement the 2D grid stealth movement with acoustic emissions, then layer the feature observation matrix and Bayesian identity guesser.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Blind He-Goat (Blind Man's Buff; traditional Norwegian blindfolded tactile tag game) (Norway). (n.d.). Games of Countries. https://gamesofcountries.com/games/no-norway/blindebukk
@misc{gamesofcountries-no-norway-blindebukk,
title = {Blind He-Goat (Blind Man's Buff; traditional Norwegian blindfolded tactile tag game) (Norway)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/no-norway/blindebukk}
}