Skip to main content
Formal Specification

Sack Racing — Formal Specification

Netherlands · strength-endurance · Confidence: low

← Back to game page

Implementation Suitability

Suitability: engine-readyFidelity: discretizedGap-fills: 3
# Zaklopen — AI-optimized specification ## General idea A linear spatial race where agents progress along a discrete track by hopping with both feet restricted inside a sack. Each turn (time tick), agents choose a hop intensity that trades off forward speed against fall probability. A fallen agent suffers a recovery delay before resuming forward progress. ## Formal model - **Abstraction taken:** Continuous physical jumping and balance dynamics are discretized into a 1D track of length L (default 30 units). Time progresses in discrete simultaneous ticks where each active agent selects a hop power level. - **Player/agent count:** 2 to n individual agents racing simultaneously. - **Turn model:** Simultaneous discrete ticks. ## Formal rules R1. **Track setup:** The course is a 1D linear track of integer length L=30 spaces. All agents start at position x=0 in the `standing` state. R2. **Action selection:** In each tick, every standing agent simultaneously chooses an action from the legal action set: `hop(power)` where `power` ∈{1,2,3}. R3. **Stochastic outcome:** - `power = 1` (Cautious): Advances 1 space. Fall chance P(fall)=0.05. - `power = 2` (Normal): Advances 2 spaces. Fall chance P(fall)=0.20. - `power = 3` (Sprint): Advances 3 spaces. Fall chance P(fall)=0.45. R4. **State transitions:** - If no fall occurs, agent position updates to x←min⁡(L,x+power). - If a fall occurs, the agent does not advance (x unchanged) and transitions to `fallen` state with `recovery_ticks_remaining` =2 [gap-fill]. R5. **Recovery:** While in `fallen` state, an agent cannot hop; each tick decrements `recovery_ticks_remaining` by 1. When it reaches 0, the agent returns to `standing` state at the current position. R6. **Win condition:** The first agent(s) to reach position x≥L wins. If multiple agents reach x≥L on the same tick, the agent with the highest overshoot wins; if exactly tied, a shared victory or tie-breaker sprint is declared. ## Edge cases - **Tie at the end condition:** If two or more agents cross x≥L on the exact same tick with identical distance, they share victory [gap-fill: standard simultaneous-tick dead heat]. - **A player with no legal move:** Not applicable; an agent in `standing` state always has the full set of hop power choices, and an agent in `fallen` state automatically executes `recover()`. - **Repetition / cycles:** Impossible; position x is monotonically non-decreasing (Δx≥0), and maximum ticks can be capped at Tmax⁡=100 [gap-fill]. - **Simultaneous claims on the same resource or position:** Lanes are independent parallel tracks; agents do not collide or block each other. - **Player elimination mid-round; turn order repair:** Not applicable; all agents race concurrently until completion without elimination. - **Resource exhaustion (pieces, tokens, space, time):** Not applicable; the course length is fixed and actions require no finite consumable resource. - **Termination guarantee:** Because P(advance≥1)≥0.55 per tick for power=1, the expected advancement per tick is strictly positive, guaranteeing almost-sure convergence to x=L within finite ticks. ## State representation - **Persistent state:** Track length L, tick counter t. - **Per-agent state:** Position xi∈[0,L], status ∈{standing,fallen}, `recovery_counter` ∈{0,1,2}, total wins. - **Observable information:** Fully observable; all agent positions and states are public knowledge. ## Action space - `hop(power)`: Parameter `power` ∈{1,2,3}. Precondition: status ==standing. - `recover()`: Automatic / default. Precondition: status ==fallen. ## Playability notes - **Single-player-vs-AI:** Natural; human racer vs. 1 or more AI agents with parameterized risk preferences. - **Group games (n>4):** Scales cleanly without structural modifications since lane interactions are independent. - **Physical-skill games:** Physical leg power and balance recovery are modeled as a discrete risk-reward decision problem (Markov Decision Process). - **What is lost:** Real-world neuromuscular coordination, slip dynamics, and uneven turf irregularities are simplified into uniform probability distributions. ## Implementation hints - Trivial complexity; state is a set of scalar tuples (xi,statei). - Optimal policy under known probabilities can be solved exactly via dynamic programming or value iteration. - First milestone: Implement the discrete tick loop with 2 agents and verify Monte Carlo win rates across aggressive vs. conservative policies.

Download as JSON

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

Download JSON
Cite this page

Sack Racing (Netherlands). (n.d.). Games of Countries. https://gamesofcountries.com/games/nl-netherlands/zaklopen

@misc{gamesofcountries-nl-netherlands-zaklopen,
  title = {Sack Racing (Netherlands)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/nl-netherlands/zaklopen}
}