Skip to main content
Formal Specification

Faroese Traditional Lifting Stones — Formal Specification

Faroe Islands · strength-endurance · Confidence: medium

← Back to game page

Implementation Suitability

Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 5
# Hevdasteinar — AI-optimized specification ## General idea A traditional Faroese stone-lifting strength contest where individual competitors attempt to hoist unworked, rounded boulders of graduated weights onto pedestals or chest-height ledges. The game tests strategic risk assessment (selecting stone tiers and lifting heights) coupled with physical stamina management, resolving each attempt via a probabilistic strength-vs-difficulty model. ## Formal model - **Abstraction taken:** The continuous biomechanical effort of straddling, lapping, and hoisting irregular stones is replaced by discrete attempt rounds where an agent selects a target stone tier and lift height, evaluated via a stochastic success function P(success)=f(strength,stamina,difficulty). [gap-fill] This stamina-decay and probability-sampling model replaces physical fatigue. - **Player/agent count:** 1 to n agents taking sequential turns in a structured trial. - **Turn model:** Strictly sequential rounds; each agent performs up to 3 lifting attempts per round. ## Formal rules R1. Setup: A set of M stones is available, indexed by mass and difficulty Dj∈{1,…,M} (e.g. Novice Stone 60 kg, Half-Strong Stone 100 kg, Fully-Strong Stone 140 kg). Each agent starts with a maximum stamina pool Smax⁡ and base strength attribute σ∈[0,100]. R2. Lift tiers and scoring: A lift attempt targets one of three recognized traditional tiers: - Tier 1: *Amli* (shin/knee level) — 1 base point. - Tier 2: *Hálfsterkur* (waist level) — 3 base points. - Tier 3: *Fullsterkur* (chest level or onto plinth) — 6 base points. R3. On an agent's turn, the agent declares a stone j and a target tier k. The probability of a successful lift is computed: P(success)=max⁡(0.05,min⁡(0.95,σ+Scurrent−Dj×k100)) [gap-fill] This formula formalizes the strength-to-stone ratio not quantified in ethnographic sources. R4. Resolution: If the sampled roll ≤P(success), the lift is successful; the agent earns the tier points, and current stamina decreases by Dj×0.5. If failed, zero points are awarded and stamina decreases by Dj×0.8 due to exertion. R5. [gap-fill] Round limit: Each match consists of exactly 3 rounds per competitor. An agent may pass an attempt to conserve stamina. R6. [gap-fill] Tie-break: If two or more agents have equal point totals after 3 rounds, the winner is the agent who completed the highest single stone tier in the fewest total attempts; if still tied, stamina remaining decides. R7. Win condition: The competitor with the highest cumulative point score at the conclusion of all rounds wins the trial. ## Edge cases - Tie at the end condition: Resolved deterministically by R6 (highest single lift difficulty, then fewest failed attempts, then remaining stamina). - A player with no legal move: Not applicable — an agent always has the option to attempt any available stone tier or deliberately pass to rest. - Repetition / cycles: Bounded by R5 (strictly 3 rounds with max 3 attempts per round). - Simultaneous claims on the same resource or position: Not applicable — turns are strictly sequential; one agent lifts at a time. - Player elimination mid-round; turn order repair: If an agent's stamina reaches 0, that agent can no longer make legal lifts and is skipped in subsequent rotations; remaining agents continue in original order. - Resource exhaustion: Stones are permanent and not consumed. Stamina is depleted per attempt. - Termination guarantee: Provably terminates in at most 3×n turns since round count is fixed. ## State representation - Persistent state: Stones list [(D1,name1),…,(DM,nameM)], per-agent score, per-agent current stamina Scurrent, attempt history log per agent. - Per-turn ephemeral state: Declared stone index, declared target tier, sampled stochastic roll. - Observable information: Fully observable — all stone weights, agent scores, and remaining stamina are public information. ## Action space - `attempt_lift(stone_id, tier_level)`: Attempt to hoist specified stone to `amli`, `halfsterkur`, or `fullsterkur`. Precondition: agent has Scurrent>0. - `pass_turn()`: Forgo the attempt to recover a fixed amount of stamina [gap-fill] (+10 stamina points). ## Playability notes - Single-player-vs-AI: Natural — single-player mode functions as a solo high-score puzzle optimizing stamina expenditure across stone tiers; vs-AI mode pits the player against AI agents with varying risk profiles. - Group games (n>4): Scales linearly — turn rotation easily accommodates any number of competitors without structural breakdown. - Physical-skill games: Physical grip and back levering are abstracted into the scalar strength attribute σ and stochastic check. - What is lost: The raw sensory friction of unhewn granite against skin, balance stabilization, and crowd cheering are completely abstracted away. ## Implementation hints Low complexity. The primary balancing task is tuning the difficulty constants Dj and stamina costs so that attempting *Fullsterkur* early carries high risk of exhaustion versus steady accumulation on *Hálfsterkur*. First milestone: implement a 2-agent, 3-round text match with 3 stone tiers.

Download as JSON

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

Download JSON
Cite this page

Faroese Traditional Lifting Stones (Faroe Islands). (n.d.). Games of Countries. https://gamesofcountries.com/games/fo-faroe-islands/hevdasteinar

@misc{gamesofcountries-fo-faroe-islands-hevdasteinar,
  title = {Faroese Traditional Lifting Stones (Faroe Islands)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/fo-faroe-islands/hevdasteinar}
}