Formal Specification
Syrian Ring Passing Game — Formal Specification
Syria · guessing-hidden · Confidence: medium
← Back to game pageImplementation Suitability
Suitability: engine-readyFidelity: directGap-fills: 1
# Lu'bat al-Khatem — AI-optimized specification
## General idea
A pure bluffing and deduction game played between two team agents. One team hides a secret token at an index in their array, and the other team must sequentially eliminate empty indices without striking the token, or make a direct accusation. The mathematical tension lies in the risk-reward ratio of continuous elimination (which increases potential payout if the opponent fails, but risks hitting the token and losing instantly).
## Formal model
- **Abstraction taken:** The group of players is abstracted into two team agents, Team A (Hider) and Team B (Guesser). The physical row of sitting players is represented as a 1D array of size (where is the number of players on Team A, typically ).
- **Player/agent count:** 2 team agents.
- **Turn model:** Alternating phases: setup phase (Team A hides) followed by sequential guessing phase (Team B makes moves).
## Formal rules
R1. Setup: Team A selects a secret index representing which player holds the ring. Team A also generates a "bluff profile" vector of size where represents the level of visual nervousness/tells displayed by player . [gap-fill: bluff profile modeling is added to provide a digital proxy for physical body language reading].
R2. Game State: A set of active players representing players whose hands are still closed. Initially, .
R3. Guesser Actions: Every turn of the guessing phase, Team B's agent must choose exactly one action:
3.1. `eliminate(i)` where .
3.2. `accuse(i)` where .
R4. Action Resolutions:
4.1. If Team B executes `eliminate(i)`:
- If , Team B has eliminated the ring holder. The round ends immediately. Team A wins the round and scores points.
- If , node is removed from (). If (only the ring holder is left), Team B is forced to accuse them on the next turn.
4.2. If Team B executes `accuse(i)`:
- If , Team B successfully finds the ring. Team B wins the round, scores 1 point, and gains the hider role for the next round.
- If , Team B has accused the wrong player. The round ends immediately. Team A wins the round, scores points, and retains the hider role for the next round.
R5. Win / Loss / Draw:
5.1. Scoring is cumulative.
5.2. The game ends as soon as either team's score reaches points. The team with the highest score wins.
## Edge cases
- Tie at the end condition: Not applicable, since turns and scoring are sequential; the first team to cross the threshold wins immediately.
- A player with no legal move: Not applicable; Team A must always select a valid index in R1, and Team B must always select an action from R3.
- Repetition / cycles: Bounded by the finite size of (at most guessing steps per round) and the strictly increasing score, ensuring the game provably terminates.
- Simultaneous claims on the same resource: Not applicable; turns are strictly sequential.
- Player elimination mid-round: Handled directly by the reduction of the active set .
- Resource exhaustion: None.
- Termination guarantee: Guaranteed by the finite score limit of 101 and the fact that every round awards at least 1 point to one of the teams.
## State representation
- Persistent state: Scores of both teams, identity of the current hider team, size of hider line , set of active hider indices , secret index , bluff vector , current round.
- Per-turn ephemeral state: Last action executed and its outcome.
- Observable information: Team A sees all variables. Team B sees and the bluff vector (representing their observation of physical tells), but is strictly hidden.
## Action space
- `hide_ring(index)`: Team A selects the secret ring holder .
- `eliminate(index)`: Team B eliminates an active index .
- `accuse(index)`: Team B accuses an active index .
## Playability notes
- Single-player-vs-AI: Highly suitable. An AI Team A can hide the ring using a neural net or simple probability distribution, generating realistic "bluff tells" containing a mix of true and false signals.
- Group games (n>4): The team-based structure natively collapses any players into a 2-agent formalism, meaning zero loss of playability or structure.
- Physical-skill games: Physical bluffing and observation are elegantly captured by the stochastic vector and the decision tree of Team B.
- What is lost: The rich social banter, intense staring contests, and collective team shouting.
## Implementation hints
- Very simple state machine. The primary challenge is designing an AI that can balance risk (eliminating more players to drive up Team A's penalty on failure) against safety (accusing early to secure the hide role).
- First milestone: Implement the state machine, action resolutions, and basic score updates with a human guesser and random hider.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Syrian Ring Passing Game (Syria). (n.d.). Games of Countries. https://gamesofcountries.com/games/sy-syria/lubat-al-khatem
@misc{gamesofcountries-sy-syria-lubat-al-khatem,
title = {Syrian Ring Passing Game (Syria)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/sy-syria/lubat-al-khatem}
}