Skip to main content
Formal Specification

Kamaj / Neighborhood Tactical Scout Tag — Formal Specification

North Macedonia · hide-seek · Confidence: low

← Back to game page

Implementation Suitability

Suitability: engine-readyFidelity: discretizedGap-fills: 5
# Камај — AI-optimized specification ## General idea A turn-based team tactical stealth and pursuit game played on a 2D obstacle grid with imperfect information (fog of war). Two teams operate from home bases, navigating cover to spot and ambush opponents with line-of-sight tags while attempting to rescue imprisoned teammates or capture the enemy base. ## Formal model - **Abstraction taken:** The urban neighborhood terrain is discretized into a 30×30 grid containing static obstacle walls (blocks line-of-sight and movement). Line-of-sight is computed via 2D raycasting. Tagging occurs when an agent enters an adjacent cell to an enemy without being flanked. - **Player/agent count:** 2 teams (TA,TB) of K agents (K∈[4,8]). - **Turn model:** Tick-based simultaneous movement and stealth actions with partial observability. ## Formal rules R1. **Grid Setup:** Grid 30×30 with static wall set 𝒲. Team A Base located at BaseA=(2,2), Team B Base at BaseB=(27,27). R2. **Agent State:** Each agent i maintains: - Position (xi,yi)∈[0,29]2∖𝒲. - `Status` ∈{Active,Jailed,SafeInBase}. - `Facing` ∈{N,E,S,W}. R3. **Action Space per Tick:** - `move(dx, dy)`: Δx,Δy∈{−1,0,1}, ‖Δ‖≤1, target cell ∉𝒲. - `scout(facing)`: Update vision cone in chosen cardinal direction (range 8 cells, 90∘ FOV). - `ambush_tag(target_id)`: Tag adjacent enemy agent (d≤1.5 cells). R4. **Vision & Fog of War:** An agent observes only cells within line-of-sight not obstructed by walls 𝒲. Information is shared among active teammates. R5. **Capture & Jail Mechanics:** - If agent Ai executes `ambush_tag` against adjacent active enemy Bj: - If Bj is not within BaseB: Bj is captured (StatusBj←Jailed), immediately teleported to BaseA, and immobilized. R6. **Prisoner Rescue (*Слобода*):** - If an active agent Bk enters the cell of BaseA where jailed teammates are held and executes `rescue_prisoners()`: - All jailed TB agents have Status←Active and gain 3 ticks of immunity to exit BaseA. R7. **Base Infiltration Victory:** If an active agent Ai steps into BaseB while no enemy defenders are present inside BaseB, Team A instantly wins by Base Infiltration. R8. [gap-fill] Full Team Imprisonment:** If all agents of a team are simultaneously `Jailed`, the opposing team wins the match. R9. [gap-fill] Maximum Turn Cap:** A match is capped at Tmax=150 ticks. If neither win condition is met, the team with more active (non-jailed) agents wins. ## Edge cases - Tie at the end condition: If both teams have equal active players at tick 150, the match is a draw [gap-fill]. - A player with no legal move: Jailed agents submit `wait()`; active agents always have legal directional moves. - Repetition / cycles: Bounded by the hard 150-tick limit. - Simultaneous claims on the same resource or position: If two opposing agents enter adjacent cells and both execute `ambush_tag` on the exact same tick, the agent who initiated movement from behind cover wins priority [gap-fill]. - Player elimination mid-round; turn order repair: Jailed agents remain in state and can be rescued, preserving full agent identities throughout. - Resource exhaustion: Grid space is static and non-depletable. - Termination guarantee: [gap-fill] Guaranteed by the 150-tick upper bound. ## State representation - **Persistent state:** Map layout, wall coordinates 𝒲, base coordinates, team rosters, match score. - **Per-tick ephemeral state:** Agent positions, status (`Active`, `Jailed`), orientations, shared team visibility maps. - **Observable information:** Partially observable; agents observe only cells within their team's combined line-of-sight (Fog of War). ## Action space - `move(dx, dy)`: Cardinal step. - `scout(facing)`: Rotate vision cone. - `ambush_tag(target_id)`: Tag adjacent enemy. - `rescue_prisoners()`: Execute jailbreak. ## Playability notes - Single-player-vs-AI: Excellent tactical stealth game; human commands 1 scout with AI squad members against an AI enemy patrol network. - Group games (n>4): Natively accommodates 4v4 to 8v8 teams on the 2D obstacle map. - Physical-skill games: Physical sprinting speed and physical stealth hiding are modeled as grid movement costs and vision occlusion. - What is lost in the adaptation: The heart-pounding thrill of hiding behind real neighborhood brick walls and the shouted chorus of *"Камај!"* echoing down cobblestone alleys. ## Implementation hints Medium complexity. Requires 2D raycasting for field-of-view, BFS pathfinding for AI bots, and a jail-state manager. First milestone: implement R1–R5 for 2v2 agents with basic line-of-sight capture.

Download as JSON

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

Download JSON
Cite this page

Kamaj / Neighborhood Tactical Scout Tag (North Macedonia). (n.d.). Games of Countries. https://gamesofcountries.com/games/mk-north-macedonia/kamaj

@misc{gamesofcountries-mk-north-macedonia-kamaj,
  title = {Kamaj / Neighborhood Tactical Scout Tag (North Macedonia)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/mk-north-macedonia/kamaj}
}