Skip to main content
Formal Specification

Fang Target Disc Pitching Game — Formal Specification

Equatorial Guinea · dexterity-throw · Confidence: low

← Back to game page

Implementation Suitability

Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 4
# Mbot — AI-optimized specification > **Reconstructed source.** The entry carries a provenance warning: its rules were written from > general knowledge, not read from the two works it cites. Ndongo Bidyogo (1977) and Panyella (1959) > are real books, but the entry does not evidence that either describes this game. ## General idea A disc-pitching game with a **steep reward gradient**: hitting the target peg outright is worth 3 points, while merely resting nearest to it is worth 1. That three-to-one ratio is the design. Unlike the closely related Ndölö entry from the Central African Republic — where touching the stake pays double the proximity score — here the direct hit pays triple, which pushes optimal play toward aggression at a distance of 6 metres where accuracy is genuinely hard. A player who consistently plays for the safe single point will lose to one who converts hits at even a modest rate, and working out where that threshold lies is the only interesting question the game poses. ## Formal model - **Abstraction taken:** Throwing accuracy over 6 m is collapsed into a per-disc outcome distribution over {hit, nearest-candidate, neither}, with proximity resolved among the discs that did not hit. The entry specifies the distance, disc budget, point values, and target exactly, and gives no accuracy model. - **Player/agent count:** 2-6 (entry §Players). - **Turn model:** Sequential throws within a round; scoring resolves once all discs are thrown. ## Formal rules R1. Target (entry §1): a carved **wooden peg** (*Nkan*) set at **6 metres** along a cleared dirt alley. R2. Disc budget (entry §2): each player throws **2 discs** per round — polished stone or hardwood. R3. Scoring (entry §3): a **direct hit** on the peg scores **3 points**; the **closest resting disc** scores **1 point** [gap-fill: the entry does not say whether a hit also counts for the proximity point, nor whether two players hitting in the same round both score 3. Both are treated as independent and additive-per-event: every hit scores 3, and the proximity point goes to the closest disc among those still resting on the alley]. R4. Victory (entry §4): the first player to accumulate **15 points** wins. R5. Displacement: **unstated** [gap-fill: the entry does not say whether a disc may strike and move another player's disc, which the closely related Ndölö entry explicitly permits. Since this entry is silent, no displacement tactic is asserted, and discs are treated as resting where they land. This is a substantive difference between two otherwise similar games and it is not harmonised]. R6. Throw order: unstated [gap-fill: fixed, rotating, or result-determined order is not specified; a fixed rotation is assumed. Order matters less here than in Ndölö precisely because R5 gives no displacement]. R7. Peg persistence: the entry does not say whether a struck peg is knocked down or stays standing [gap-fill: the peg is assumed to remain a valid target throughout the round, since R3 implies repeated hits are scorable]. ## Edge cases - Tie at the end: not applicable at the victory condition, since R4 fires the moment a total reaches 15. Two players crossing 15 in the same round is possible — a 3-point hit each — and is unaddressed by the entry. Within a round, two discs equidistant from the peg would tie R3's proximity comparison; with continuous positions this is negligible, and no rule is invented. - A player with no legal move: not applicable — every player has `pitch` for each of their 2 discs in every round, and discs are recovered between rounds. - Repetition / cycles: rounds repeat until R4 is satisfied. No cycle risk exists, because R3's proximity point is awarded in every round to some player, which is what bounds the game below. - Simultaneous claims on one position: **excluded by R5's silence.** Two discs cannot occupy the same point physically, but with no displacement rule the entry offers no resolution for a disc landing where another rests. This is precisely where the Ndölö entry has a rule and this one does not, and the difference is preserved rather than filled in. - Elimination mid-round and turn-order repair: no player is ever eliminated, and discs are recovered each round. The rotation is fixed under R6's assumption and needs no repair. - Resource exhaustion: not applicable — the 2 discs per player are recovered each round and the peg persists (R7). Nothing depletes; the only growing quantity is the score. - Termination guarantee: **guaranteed deterministically.** Scores are monotone — nothing subtracts — and R3 awards at least one point per round, since some disc is always closest to the peg. With `n` players, after at most `14n + 1` rounds some player must hold 15 points by the pigeonhole principle. This is a hard bound rather than a probability-1 argument, and it holds even if no disc ever hits the peg. ## State representation - Persistent (match) state: `score[p]` (monotone non-decreasing, target 15), `current_round`, `throw_order`. - Per-round ephemeral state: `disc_positions[p][2]`, `hit[p][d]` (bool), `distance_to_peg` derived per resting disc. - Observable information: fully observable — a cleared alley where the peg, every disc, and all scores are visible. The only asymmetry is temporal: earlier throwers commit with less information about the round's eventual geometry. ## Action space - `pitch(aim, force)`: the active player; precondition: discs remain this round (R2). This is the only action, and — unlike Ndölö, where R4 gives throws a second purpose — the decision here is purely how much accuracy risk to accept in pursuit of R3's 3-point hit. There is no target selection, because R5 provides no opponent-directed play. ## Playability notes - Single-player-vs-AI: moderate. The risk-reward choice between playing for the hit and playing for proximity is genuine, but it is a single scalar decision repeated every throw, and with no interaction between players (R5) an AI opponent is effectively a second sample from the same distribution rather than an adversary. - Reducing n > 4 to 2 agents: mechanically lossless, since players do not interact. Fewer players means the proximity point is contested by fewer discs, which slightly raises each player's expected score per round and shortens the match. - Physical-skill games: throwing accuracy at 6 m is the entire contest and collapses into a single outcome draw. A polished stone disc thrown along a dirt alley will skip and roll rather than stop where it lands, and none of that ground interaction is modelled — which matters, because it is what determines final resting position and hence R3's proximity point. - What is lost: the Río Muni rainforest setting and the Fang clan communities the entry names, and the team-duel variant it records with alternating throwing turns — a structurally different game the entry mentions only in passing. ## Implementation hints Model the disc's **resting** position rather than its impact point: R3 scores where discs come to rest, and on a dirt alley a thrown disc travels well past first contact, so a model that scores the landing point will systematically misjudge proximity. Expose the hit probability as a function of force, so that R3's 3-versus-1 gradient produces a real optimisation — the interesting output of any implementation here is the hit rate at which aggression beats safety, and that number should fall out of the model rather than being assumed. Do **not** import Ndölö's displacement rule to fill R5: the two entries differ on exactly this point, and adding displacement would make them the same game when the corpus records them as different. R7's peg persistence should be checked against whatever scatter model is used, since a knocked-down peg would end proximity scoring for the round. Suggested first milestone: the throw loop with a force-dependent hit probability and a roll-out resting model, verifying the pigeonhole bound — every round awards at least one point, and matches always end.

Download as JSON

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

Download JSON
Cite this page

Fang Target Disc Pitching Game (Equatorial Guinea). (n.d.). Games of Countries. https://gamesofcountries.com/games/gq-equatorial-guinea/mbot-fang

@misc{gamesofcountries-gq-equatorial-guinea-mbot-fang,
  title = {Fang Target Disc Pitching Game (Equatorial Guinea)},
  year = {n.d.},
  publisher = {Games of Countries},
  url = {https://gamesofcountries.com/games/gq-equatorial-guinea/mbot-fang}
}