Skip to main content
Formal Specification

Grandmother's Yarn Ball — Formal Specification

Estonia · dexterity-small · Confidence: medium

← Back to game page

Implementation Suitability

Suitability: engine-readyFidelity: probabilistic-proxyGap-fills: 4
# Vanaema lõngakera — AI-optimized specification ## General idea A topological cooperation and spatial puzzle game. A circle of agents creates an entangled graph of linked arms by stepping over and ducking under connections without breaking hand grips. A designated "Grandmother" agent must diagnose the permutation and direct a sequence of reverse moves to untangle the cycle back to an unknotted circle. ## Formal model - **Abstraction taken:** The continuous human knot is formalized as an unknotted 1D cyclic graph embedded in ℝ3 with discrete topological crossing operations (Reidemeister-style moves: step-over, duck-under). - **Player/agent count:** 1 solver agent (G) and N circle nodes (P1,…,PN). - **Turn model:** Two-phase game: Phase 1 (knot generation by circle agents), Phase 2 (sequential verbal directives from G). ## Formal rules R1. Setup: N agents form a simple planar cycle C=(P1,P2,…,PN,P1) with fixed edge connectivity (Pi,Pi+1). Agent G is placed in `ISOLATED_OBSERVATION` state (cannot observe Phase 1 actions). R2. **Entanglement Phase:** Circle agents execute a sequence of K valid topological moves (K∈[3,10]) [gap-fill]. - A legal move consists of an agent Pi passing under or over the link (Pj,Pj+1) where j∉{i−1,i}. - Edge connectivity must remain unbroken (grip(Pi,Pi+1)=TRUE). R3. Knot Freezing: After K moves, the tangle configuration state Stangle is locked. Agent G transitions to `ACTIVE_SOLVER`. R4. **Untangling Phase:** G issues sequential action instructions: `instruct(agent_id, move_type, target_edge)`. R5. Move Execution: The designated agent executes the inverse topological move if physically and topologically valid. If a move is invalid or breaks a grip, a fault counter increments [gap-fill]. R6. Win/Loss Conditions: - If the state returns to an open planar circle (0 crossings, planar embedding restored) within M=20 moves [gap-fill], G wins. - If 3 faults occur or G exceeds M moves, the Knot wins. ## Edge cases - Tie at the end condition: Not applicable; binary win/loss between G and the Knot. - A player with no legal move: In the knot phase, if no agent can step without breaking grip, the knot is declared complete. - Repetition / cycles: Governed by the strict move limit M=20 [gap-fill]; infinite looping by G results in a Knot victory. - Simultaneous claims on the same resource or position: In Phase 2, G issues strictly sequential, single-agent commands. - Player elimination mid-round; turn order repair: No elimination; if an agent disconnects, the cycle is repaired by bridging neighbors (Pi−1,Pi+1). - Resource exhaustion: Move budget M and fault limit are finite counters. - Termination guarantee: Monotonically guaranteed by the move counter bound M. ## State representation - Persistent state: Cyclic adjacency graph, ordered list of 3D link crossings, move budget m, fault counter. - Per-turn ephemeral state: Active command from G, validation status of target move. - Observable information: In Phase 1, circle agents observe local crossings. In Phase 2, G has full visual observation of the entangled 3D graph state. ## Action space - `step_over(edge_id)` / `duck_under(edge_id)`: Executed by circle agents in Phase 1. - `instruct(agent_id, action, edge_id)`: Executed by G in Phase 2. ## Playability notes - Single-player-vs-AI: Natural as a single-player combinatorial puzzle where the player acts as G untangling an algorithmically scrambled cycle. - Group games: Models cooperative multi-agent motion planning with topological constraints. - Physical-skill games: Physical flexibility is abstracted into discrete graph transformations. - What is lost: The physical comedy of awkward human contortions and physical contact. ## Implementation hints Treat as a knot theory / planar graph untangling problem. Milestone 1: Represent the linked circle as a planar cycle with crossing signed integers. Milestone 2: Build a greedy solver agent that searches for crossing-reduction moves.

Download as JSON

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

Download JSON
Cite this page

Grandmother's Yarn Ball (Estonia). (n.d.). Games of Countries. https://gamesofcountries.com/games/ee-estonia/vanaema-longakera

@misc{gamesofcountries-ee-estonia-vanaema-longakera,
  title = {Grandmother's Yarn Ball (Estonia)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/ee-estonia/vanaema-longakera}
}