Formal Specification
Emirati Sea Diving and Retrieval Game — Formal Specification
Uae · water · Confidence: low
← Back to game pageImplementation Suitability
Suitability: adaptation-requiredFidelity: discretizedGap-fills: 2
# Al-Khuwaisa — AI-optimized specification
## General idea
A competitive underwater search and retrieval game. Agents manage limited breath-holding capacity (oxygen timer) while navigating a 3D discretized water volume (surface, mid-water, seabed) to locate a hidden sunken token and surface cleanly before rivals intercept.
## Formal model
- **Abstraction taken:** The water volume is discretized into a 3D grid `(x, y, z)` where `z ∈ {0: Seabed, 1: Midwater, 2: Surface}`. Time progresses in discrete ticks (`1.0s`). Agents have an active `breath_capacity` that depletes while `z < 2`.
- **Player/agent count:** `N` individual agents (`3 <= N <= 8`).
- **Turn model:** Simultaneous real-time ticked execution.
## Formal rules
R1. Aquatic Grid:
- `x ∈ [0..15]`, `y ∈ [0..15]`, `z ∈ {0, 1, 2}`.
- Start line: `y = 0`, `z = 2` (all agents start at surface).
- Hidden target object placed on seabed at `(x_target, y_target, 0)` with uniform distribution `x, y ∈ [4..12]`.
R2. Breath and Diving Mechanics:
- Agents start with `breath = max_breath` (default 25 ticks).
- While submerged (`z < 2`), `breath` decreases by 1 per tick.
- If `breath <= 0`, agent suffers hypoxia, drops any held item, and is forced to float to surface `z = 2` for 5 ticks of recovery.
R3. Vision and Discovery:
- When an agent is at `z = 0` (seabed) within Euclidean distance `dist <= 2.0` of `(x_target, y_target, 0)`:
- Discovery check: `P(spot) = vision_skill * (1.0 - dist / 3.0)`. [gap-fill]
- If spotted: Agent executes `pickup_object()` and gains item possession.
R4. Underwater Interception:
- An agent possessing the object moves toward `z = 2`.
- If another co-located submerged agent executes `intercept_carrier()`:
- Contest check: `P(steal) = 0.5 + (stamina_thief - stamina_carrier) * 0.01`. [gap-fill]
- If steal succeeds: Item transfers to the interceptor.
R5. Scoring and Round End:
- An agent holding the object who reaches `z = 2` (surface) and executes `claim_surface()` scores 1 point.
- First agent to reach 3 points wins the match.
## Edge cases
- **Tie at the end condition:** First agent to surface and claim object triggers exclusive round win.
- **A player with no legal move:** Submerged agents can always swim toward surface or search.
- **Repetition / cycles:** Bounded by finite breath pools forcing regular ascents and eventual object discovery.
- **Simultaneous pickup claims:** Broken by highest diving skill rating.
- **Player elimination mid-round:** Out-of-breath agents recover at surface without disrupting active competitors.
- **Resource exhaustion:** Single persistent token per round.
- **Termination guarantee:** Guaranteed by finite points target and monotonic round progress.
## State representation
- **Persistent state:**
- `scores: Array[N] of int`.
- `round_number: int`.
- **Ephemeral round state:**
- `agent_coords: Array[N] of Tuple[int, int, int]`.
- `agent_breath: Array[N] of int`.
- `target_pos: Tuple[int, int, int]`.
- `carrier_id: Optional[int]`.
- `target_discovered: bool`.
- **Information set:** Partially observable (object position is hidden until agent enters vision radius).
## Action space
- `swim(dx: int, dy: int, dz: int)`: Move 1 cell in grid.
- `pickup_object()`: Precondition: co-located with target at `z = 0`.
- `intercept_carrier(target_id: int)`: Precondition: submerged and co-located with carrier.
- `claim_surface()`: Precondition: at `z = 2` holding target object.
## Playability notes
- **Single-player-vs-AI:** Highly playable. Player balances seabed exploration time against breath limits while tracking rival diver bubbles.
- **Group games:** Scales naturally from 3 to 8 players.
- **Physical-skill model:** Apnea endurance and underwater spatial navigation are mapped to discrete breath timers and stochastic perception bubbles.
- **What is lost:** True hydrodynamic fluid dynamics and ambient water clarity fluctuations.
## Implementation hints
Model as a 3D grid with simple Manhattan/Euclidean search. Ensure breath depletion forces AI agents to make realistic risk-managed ascents to surface.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Emirati Sea Diving and Retrieval Game (Uae). (n.d.). Games of Countries. https://gamesofcountries.com/games/ae-uae/al-khuwaisa
@misc{gamesofcountries-ae-uae-al-khuwaisa,
title = {Emirati Sea Diving and Retrieval Game (Uae)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/ae-uae/al-khuwaisa}
}