Skip to main content
Formal Specification

Highland Log Toss — Formal Specification

Kosovo · dexterity-throw · Confidence: high

← Back to game page

Implementation Suitability

Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 4
# Hedhja e llastarit — AI-optimized specification ## General idea A turn-based throwing game where agents compete across successive rounds to achieve the maximum linear distance with a heavy tapered wooden log (*llastar*). Agents must select balance stability, sprint momentum, and upward launch thrust, managing a dual-failure risk: stepping over the scratch line (foot fault) or failing to generate enough rotational torque to flip the log end-over-end (flip fault). ## Formal model - **Abstraction taken:** Three-dimensional aerodynamic flight, angular momentum, and timber ground rebound are abstracted into an agent decision tuple: sprint speed v∈[0.5,1.0], upward thrust angle ϕ∈[60∘,85∘], and rotational torque effort τ∈[0.5,1.0]. - **Player counts:** 2 to N agents (benchmark: 2 agents). - **Turn model:** Sequential throw attempts across 3 rounds. ## Formal rules R1. **Initial Setup:** Each agent i has strength rating Pi∈[1,10] and balance rating Bi∈[1,10]. Tournament length R=3 rounds. R2. **Turn Action:** On each round r∈{1,…,R}, active agent i submits an action tuple (v,ϕ,τ). R3. **Fault Check 1 — Baseline Scratch:** [gap-fill] - Probability of foot fault during approach: Pscratch=max⁡(0.05,0.40×v2×(1.0−Bi/15)) - If ξ1<Pscratch, throw is a FOUL (di,r=0). R4. **Fault Check 2 — End-over-End Flip:** [gap-fill] - The log requires sufficient vertical angle and torque to achieve a forward flip. Flip success probability: Pflip=min⁡(0.95,τ×sin⁡(ϕ)×(Pi/10)) - If ξ2>Pflip, log fails to turn over (di,r=0). R5. **Distance Resolution:** If both fault checks pass: - [gap-fill] Thrown distance: di,r=max⁡(0,5.0+8.0×v×cos⁡(ϕ−70∘)+0.5×Pi+𝒩(0,0.5)) R6. **Ranking & Victory:** Each agent's best throw Dmax(i)=maxr≤R⁡di,r determines standings. The agent with highest Dmax wins. R7. **Tie-Breaker:** [gap-fill] Ties broken by the second-best valid throw; if still tied, a single sudden-death toss is executed. ## Edge cases - **Tie at the end condition:** Handled explicitly by R7 (secondary throw distance comparison, followed by single-toss sudden death). - **A player with no legal move:** Not applicable; continuous parameter action space is always available within valid intervals. - **Repetition / cycles:** Tournament is strictly bounded to R=3 rounds (maximum 3N turns). - **Simultaneous claims on the same resource or position:** Not applicable; throws are executed sequentially. - **Player elimination mid-round / turn order repair:** If an agent disconnects, their attempt is recorded as a foul (d=0) and rotation advances. - **Resource exhaustion:** Not applicable; the competition log is non-depleting. - **Termination guarantee:** The game counter strictly increments from round 1 to 3, guaranteeing termination in exactly 3N steps. ## State representation - **Persistent state:** - `active_agent_idx`: Integer ∈[0,N−1] - `current_round`: Integer ∈[1,3] - `attempt_matrix`: N×3 array of float distances / foul markers - `best_distances`: Array of N floats - **Observable information:** Complete public visibility of scores and past throws. ## Action space - `toss_log(speed, angle, torque)`: - `speed`: Float ∈[0.50,1.00] (sprint approach intensity). - `angle`: Float ∈[60.0,85.0] (launch angle in degrees). - `torque`: Float ∈[0.50,1.00] (rotational wrist/arm flick). ## Playability notes - **Single-player-vs-AI:** Natural. Single human player tunes parameter combinations against AI competitors with varied risk tolerances. - **Group games (n>4):** Scales trivially to any N participants without changing game mechanics. - **Physical-skill games:** Modeled as a dual-constraint optimization problem (preventing scratch fouls on entry and rotation failures on landing). - **What is lost in the adaptation:** The physical strain of balancing a heavy vertical log against the shoulder and the acoustic crash of timber on turf. ## Implementation hints - Bivariate failure envelope (Pscratch vs Pflip) makes strategic optimization non-trivial: high speed increases distance but spikes scratch risk, while low angle increases distance but spikes flip failure. - First milestone: Implement distance function and calibrate failure probabilities to match real-world high-pasture distances (7–15 meters).

Download as JSON

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

Download JSON
Cite this page

Highland Log Toss (Kosovo). (n.d.). Games of Countries. https://gamesofcountries.com/games/xk-kosovo/hedhja-e-llastarit

@misc{gamesofcountries-xk-kosovo-hedhja-e-llastarit,
  title = {Highland Log Toss (Kosovo)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/xk-kosovo/hedhja-e-llastarit}
}