Skip to main content
Formal Specification

Caber Toss (Tossing the Caber) — Formal Specification

United Kingdom · strength-endurance · Confidence: low

← Back to game page

Implementation Suitability

Suitability: engine-readyFidelity: abstractedGap-fills: 1
# Caber Toss (Tilgeil a' chabair) — AI-optimized specification ## General idea A sequential heavy athletic contest where agents balance a 150-lb tapered tree trunk, sprint forward, and heave it so the heavy tip strikes the turf and pivots over. Unlike distance throws, the objective is rotational completion and angular alignment against a 12 o'clock clock face relative to the runner's line of approach. ## Formal model - **Abstraction taken:** Physical lift balance, forward acceleration, and explosive hip heave are parameterized into three strategic control choices: `approach_velocity in [3.0, 7.0]` m/s, `heave_angle in [70.0, 85.0]` degrees, and `vertical_power in [0.5, 1.0]`. Dynamic physics resolves log angular momentum `L = I * omega` upon ground impact to determine if the caber rotates through the 90-degree vertical threshold to achieve a "turn". - **Player/agent count:** 1 active agent per heat in an N-agent tournament bracket (typically `N = 8` competitors). - **Turn model:** 3 sequential attempts per agent across 3 tournament rounds. ## Formal rules R1. **Tournament Setup:** Each agent is allocated exactly 3 attempts. Best single attempt score determines final ranking. R2. **Toss Declaration:** On each attempt, the active agent declares: - `toss(approach_speed: float, heave_angle: float, lift_force: float)` R3. **Turn and Rotation Resolver:** - Rotational momentum upon top-end turf contact: `omega_tip = f(approach_speed, heave_angle, lift_force) - Gravity_Resistance`. - If `omega_tip <= 0`: The caber fails to turn. Score is recorded as `Angle(degrees)` reached relative to ground (`score_type = ANGLE`, value in `[45.0, 89.9]` degrees). - If `omega_tip > 0`: The caber completes a successful turn (`score_type = CLOCK`). R4. **Clock-Face Alignment Calculation:** - Lateral deviation angle `theta_drift ~ Normal(0, sigma_balance)` in radians. - Minute drift from 12:00: `minutes_drift = abs(theta_drift) * (30.0 / pi)`. - Clock score representation: `Deviation = minutes_drift` (e.g. `0.0` is perfect 12:00; `5.0` is 12:05 or 11:55; `15.0` is 12:15 or 11:45). R5. **Ranking and Score Hierarchy:** - Any `CLOCK` score strictly outranks any `ANGLE` score. - Between two `CLOCK` scores, the one with lower `Deviation` (closer to 0.0 / 12:00) strictly wins. - Between two `ANGLE` scores, the one with higher angle (closer to 90.0 degrees) wins. R6. **Victory:** The agent with the highest single ranked attempt across the 3 rounds wins the competition. ## Edge cases - **Tie at the end condition:** If two agents tie with identical best clock deviations (e.g. both throw exact 12:00), [gap-fill] the tie is broken by comparing their second-best throw, then third-best throw; if still tied, a 1-throw sudden-death toss decides the winner. - **A player with no legal move:** Not applicable; `toss()` is always legal. - **Repetition / cycles:** Bounded by strict 3-attempt tournament limit. - **Simultaneous claims on the same resource:** Not applicable; competitors throw one at a time on an open runway. - **Player elimination mid-round:** If an athlete suffers a fault/injury on attempt 1, remaining attempts record DNF without invalidating previous scores. - **Resource exhaustion:** Exactly 3 attempts per athlete; log is non-depletable. - **Termination guarantee:** Guaranteed by finite attempt loops (`3 * N` total throws). ## State representation - **Persistent state:** `tournament_round: int (1..3)`, `current_thrower_index: int (0..N-1)`, `attempt_history: dict[agent_id -> list of 3 throw_records]`, `current_leaderboard: list of agent_id`. - **Per-throw ephemeral state:** `declared_inputs: tuple`, `is_turned: bool`, `clock_deviation: float`, `angle_achieved: float`. - **Observable information:** Fully public leaderboard and throw records. ## Action space - `toss(approach_speed, heave_angle, lift_force)`: Continuous 3-parameter input tuple. ## Playability notes - **Single-player-vs-AI:** Excellent strength-timing simulator; player attempts to find the perfect speed and launch window to achieve 12:00 under fluctuating balance noise. - **Group games reduction (2-agent form):** Natural 1v1 head-to-head duel over 3 throws each. - **Physical-skill games:** Physical leg drive and shoulder balance are abstracted into the angular momentum equation and balance noise `sigma`. - **What is lost in the adaptation:** The physical strain of cupping a 150-lb tree trunk with raw blistered hands and running in a Highland gale. ## Implementation hints - Simple mathematical physics model (rigid body rotational kinetics upon impact). - First milestone: Write the turn-evaluation function and run a Monte Carlo test across varying speed/force inputs to verify that 12:00 is rare and rewarding.

Download as JSON

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

Download JSON
Cite this page

Caber Toss (Tossing the Caber) (United Kingdom). (n.d.). Games of Countries. https://gamesofcountries.com/games/gb-united-kingdom/caber-toss

@misc{gamesofcountries-gb-united-kingdom-caber-toss,
  title = {Caber Toss (Tossing the Caber) (United Kingdom)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/gb-united-kingdom/caber-toss}
}