Skip to main content
Formal Specification

Kloot Throwing — Formal Specification

Germany · dexterity-throw · Confidence: high

← Back to game page

Implementation Suitability

Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 1
# Klootschießen — AI-optimized specification ## General idea An athletic projectile distance contest where agents execute multi-phase biomechanical throws (runway sprint, springboard leap, and overhand release) with a dense lead-weighted ball. The physical throw yields two consecutive distance components: ballistic aerial flight (*Flug*) and surface roll (*Trüll*), subject to a foul-line step-fault check. In tournament mode, agents seek to maximize single best total distance across fixed attempt rounds. ## Formal model - **Abstraction taken:** The continuous sprint mechanics, springboard rebound dynamics, and ground bounce physics are abstracted into an agent decision over launch parameters (sprint_velocity,jump_timing,release_angle), which map stochastically via an execution model to a valid throw boolean and distance vector (Δxflight,Δxroll). - **Player/agent count:** 1 to N individual competitors (standard tournament: N≥2) or 2 teams. - **Turn model:** Sequential attempt rounds; each agent executes one attempt per round in fixed order. ## Formal rules R1. **Tournament Setup:** The contest consists of R=3 official rounds. A roster of N competitors is registered with initialized best distance Dbest(i)=0.0 for all i∈{1…N}. R2. **Turn Sequence:** In each round r∈{1…R}, agents take turns in ascending index order i=1,2,…,N. R3. **Action Submission:** The active agent submits action `throw(approach_speed, release_angle, risk_factor)`. R4. **Foul Determination:** A foul check is computed: Pfoul=f(approach_speed,risk_factor). If foul occurs, the attempt is scored as 0.0 meters (`invalid`). R5. **Distance Computation:** If valid, the resulting flight and roll are sampled: - Δxflight∼𝒩(μflight(speed,angle),σflight2) - Δxroll∼Exponential(λ(ground_hardness,Δxflight)) - Total throw distance: Dthrow=Δxflight+Δxroll. R6. **Score Update:** If Dthrow>Dbest(i), the agent's personal record updates: Dbest(i)←Dthrow. R7. [gap-fill] Tie-Breaking Rule:** If two or more agents have identical Dbest at the conclusion of round R, ties are broken by their second-best valid throw D2nd(i); if still tied, a sudden-death single throw round is conducted. R8. **Win Condition:** The agent with the maximum recorded Dbest at the end of round R (or after tiebreak) wins the competition. ## Edge cases - Tie at the end condition: Resolved via R7 (second-best attempt comparison, followed by sudden-death throw-off). - A player with no legal move: Not applicable; submitting a valid throw parameter tuple is always available. - Repetition / cycles: Bounded strictly by fixed round count R=3 and finite sudden-death rounds. - Simultaneous claims on the same resource or position: Not applicable; turns are strictly sequential from a single runway. - Player elimination mid-round; turn order repair: If an agent disconnects or withdraws, their previous scores are retained and subsequent rounds skip their turn slot. - Resource exhaustion (pieces, tokens, space, time): Not applicable; the single *Kloot* sphere is retrieved after each recorded throw. - Termination guarantee: why the game provably ends: Guaranteed because the outer round loop has fixed length R, each round executes exactly N throws, and tie-breaking has strictly positive resolution probability per sudden-death round. ## State representation - **Persistent state:** Total rounds R, current round r, agent roster, score table holding all historical throws and Dbest(i) per agent. - **Per-turn ephemeral state:** Active agent index i, chosen throw parameters, sampled outcome (valid,Δxflight,Δxroll). - **Observable information:** Perfect information; all historical throws, current rankings, ground conditions, and agent scores are globally visible. ## Action space - `throw(approach_speed, release_angle, risk_factor)`: - `approach_speed` ∈[0.0,1.0]: normalized approach velocity. - `release_angle` ∈[20.0∘,60.0∘]: launch trajectory angle. - `risk_factor` ∈[0.0,1.0]: tradeoff parameter increasing peak potential distance at the cost of higher Pfoul. Precondition: agent is active and currently on the approach runway. ## Playability notes - Single-player-vs-AI: Natural; 1 human user competes in a high-score leaderboard challenge or head-to-head match against parametric AI throwers. - Group games (n>4): Individual format naturally scales to any N competitors without rule modifications. - Physical-skill games: Physical sprint, jump coordination, and wrist whip are mapped to an optimization surface over (speed,angle,risk) balancing expected distance against foul probability. - What is lost in the adaptation: Muscular strain, sensory feel of the lead-cored wooden ball, and tactical foot placement on the physical wooden springboard. ## Implementation hints Very low computational complexity. Core engine is an iterative tournament evaluator with a 2D trajectory/roll sampler. First milestone: implement single-agent 3-throw trial and verify that risk-reward trade-off curves produce realistic distance distributions (approx. 60–100 m for elite adult Kloot throws).

Download as JSON

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

Download JSON
Cite this page

Kloot Throwing (Germany). (n.d.). Games of Countries. https://gamesofcountries.com/games/de-germany/klootschiessen

@misc{gamesofcountries-de-germany-klootschiessen,
  title = {Kloot Throwing (Germany)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/de-germany/klootschiessen}
}