Skip to main content
Formal Specification

Ogar-Ogar (The Sleeping Wolf Circle Tag) — Formal Specification

Turkmenistan · chase-tag · Confidence: low

← Back to game page

Implementation Suitability

Suitability: engine-readyFidelity: discretizedGap-fills: 4
# Ogar-ogar — AI-optimized specification ## General idea A traditional children's circle-approach and explosive-scatter tag game. A designated "wolf" (*möjek*) lies dormant in the center of a radial arena while a surrounding circle of children slowly contracts inward chanting traditional verses. The instant a runner dares to touch the sleeping wolf, the wolf awakens and triggers an immediate scatter sprint toward outer safety boundaries. ## Formal model - **Abstraction taken:** Physical creeping, touching, and scattering are discretized onto a 2D continuous or grid-based polar arena (r∈[0,Rsafe],θ∈[0,2π)). The game transitions deterministically through two distinct operational phases: the Contracting Circle Phase (r→0) and the Scattering Pursuit Phase (r→Rsafe). - **Player/agent count:** 1 Wolf agent, N Runner agents (N≥2, typically 9 to 39). - **Turn model:** Real-time-ticked with simultaneous action commitment per tick. ## Formal rules R1. **Arena & Setup:** The field is an arena of radius Rsafe=20 m. At t=0, 1 agent is assigned role `WOLF` and initialized in state `SLEEPING` at coordinate (0,0). N `RUNNER` agents initialize in a circular ring at radius r=15 m in `APPROACHING` state. R2. **Phase 1 (Contracting Approach Phase):** - On each tick during Phase 1, runners may move inward toward (0,0) at maximum speed vapproach=1.0 m/s. - The wolf remains stationary in `SLEEPING` state and cannot move or tag. - Any runner within contact distance d≤1.0 m of (0,0) may execute the action `TouchWolf()`. R3. **Phase 2 (Awakening & Scatter Phase):** - The instant `TouchWolf()` is executed on tick t∗: - The wolf transitions to `AWAKE` state with sprint speed vwolf=5.0 m/s. - All runners transition to `FLEEING` state with sprint speed vrunner=4.0 m/s, attempting to reach r≥Rsafe. R4. **Tagging & Role Transfer:** - The wolf tags a fleeing runner if ‖pwolf−prunner‖2≤1.0 m before that runner reaches r≥Rsafe. - [gap-fill] The first runner tagged is designated the new Wolf for the next round. - [gap-fill] If all runners reach r≥Rsafe without any tag occurring, the current wolf remains the wolf for another round. R5. **Scoring & End Condition:** - [gap-fill] The game is played across K=5 rounds. - Each runner earns +1 survival point for every round they successfully reach safety untagged. - [gap-fill] The agent with the highest cumulative survival score after K rounds wins. ## Edge cases - Tie at the end condition: Ties in cumulative score are shared equally as co-victors. - A player with no legal move: Not applicable — moving in any 2D heading or holding position is always legal. - Repetition / cycles: Bounded because Phase 2 runners have strictly positive radial velocity (dr/dt>0) toward Rsafe, terminating Phase 2 in at most ⌈Rsafevrunner⌉=5 seconds. - Simultaneous claims on the same resource or position: If multiple runners execute `TouchWolf()` on the exact same tick, the trigger fires normally; if the wolf simultaneously touches two runners, the closer runner is tagged. - Player elimination mid-round; turn order repair: Tagged runners are marked for next-round role assignment; active fleeing continues until all agents exit or are caught. - Resource exhaustion (pieces, tokens, space, time): Spatial boundary Rsafe is fixed; round count K is finite. - Termination guarantee: why the game provably ends: Guaranteed by monotonic outward radial displacement in Phase 2 and finite round limit K. ## State representation - Persistent state: Round counter k∈[1,K], survival scoreboard `Dict[AgentID, int]`, active Wolf ID. - Per-player state: `position: (x, y)`, `velocity: (vx, vy)`, `role: {WOLF, RUNNER}`, `state: {SLEEPING, AWAKE, APPROACHING, FLEEING, SAFE, TAGGED}`. - Per-turn ephemeral state: Simultaneous action vector a→t, `TouchWolf` event flag, tag contact events. - Observable information: Fully observable — all agent positions, role assignments, and phase states are public. ## Action space - `MoveTowards(target_x, target_y)`: 2D directional movement clamped by role speed limits. - `TouchWolf()`: Execute physical tap on dormant wolf. Precondition: Distance to wolf ≤1.0 m during Phase 1. ## Playability notes - Single-player-vs-AI: Natural — 1 human player controls the Wolf timing their awakening rush, or controls a daring runner pushing for close contact before fleeing. - Group games (n>4): Reduces cleanly to a 2-agent form (1 Wolf vs 1 Runner in a 10 m radial circle) without destroying the core risk-reward threshold. - Physical-skill games: Physical reaction time, sudden sprint acceleration, and peripheral spatial awareness are modeled via discrete speed differentials and reaction latency. - What is lost in the adaptation: The atmospheric choral chanting in unison (*"Ogar-ogar, ogar-a!"*), the theatrical pretending of sleep, and the sudden group adrenaline spike are converted into coordinate tracking. ## Implementation hints Low complexity. Standard 2D continuous/grid arena. Branching factor is continuous (heading angle) but easily discretized into 8 or 16 directional headings. Suggested first milestone: implement the 2-phase state machine (Contracting → Scattering) with 1 Wolf and 3 AI runners, verifying boundary safety detection and tag resolution.

Download as JSON

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

Download JSON
Cite this page

Ogar-Ogar (The Sleeping Wolf Circle Tag) (Turkmenistan). (n.d.). Games of Countries. https://gamesofcountries.com/games/tm-turkmenistan/ogar-ogar

@misc{gamesofcountries-tm-turkmenistan-ogar-ogar,
  title = {Ogar-Ogar (The Sleeping Wolf Circle Tag) (Turkmenistan)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/tm-turkmenistan/ogar-ogar}
}