Formal Specification
Murrumbidgee (Diving Retrieval Race) — Formal Specification
Australia · water · Confidence: low
← Back to game pageImplementation Suitability
Suitability: adaptation-requiredFidelity: probabilistic-proxyGap-fills: 4
# Murrumbidgee — AI-optimized specification
## General idea
A timing-and-depth skill contest where deliberately waiting before diving is itself a difficulty dial — the variant of delaying the dive to catch the object closer to the bottom turns a simple retrieval task into a genuine risk-calibration decision.
## Formal model
- **Abstraction taken:** the dive becomes a single `dive(delay)` action, where delay is how long the diver waits after the throw before entering the water; success is drawn from a distribution that gets harder as delay increases (closer to the bottom) but rewards more if so, per the variant's own framing.
- **Player/agent count:** 2-4 per group, acting in fixed thrower/diver or rotating roles.
- **Turn model:** sequential — one throw-and-dive attempt at a time.
## Formal rules
R1. The thrower performs `throw(object)` into the deep water.
R2. The diver performs `dive(delay)`, choosing how long to wait after the throw.
R3. Retrieval succeeds (a "clean" catch before the bottom) with `P(success) = f(skill, delay)`, decreasing as delay increases. [gap-fill] no explicit success-probability formula is given; added since a guaranteed catch removes all skill content, and the delay variant explicitly implies a difficulty gradient.
R4. [gap-fill] A successful catch scores points proportional to the delay risked, rewarding the harder, later catches more, since the source frames later diving specifically as "a test of skill" rather than a fixed pass/fail.
R5. Roles alternate, or a fixed number of attempts per diver are taken; the player or team with the highest total score wins.
## Edge cases
- **Tie at the end condition:** equal total scores after the agreed number of attempts; [gap-fill] resolved by a decider dive, since the source gives no explicit tie-break.
- **Player with no legal move:** not applicable — dive(delay) is always available.
- **Repetition / cycles:** not applicable — each attempt resolves independently.
- **Simultaneous claims:** the multi-diver variant has several agents diving for the same object; [gap-fill] resolved by nearest-to-object priority, with an explicit no-contact constraint per the source's own caution.
- **Mid-round elimination and turn-order repair:** not applicable — no elimination.
- **Resource exhaustion:** attempts, if a fixed count is used (R5), are the bounded resource.
- **Termination guarantee:** guaranteed by the fixed attempt count (R5).
## State representation
Persistent: water depth, delay-to-difficulty function, skill parameters.
Per-player: cumulative score.
Per-turn ephemeral: this attempt's chosen delay and resolved outcome.
Information set: fully observable — every dive's outcome is public.
## Action space
- `throw(object)` — precondition: actor is the thrower.
- `dive(delay)` — precondition: actor is the diver, a throw has just been made.
## Playability notes
- **Single-player-vs-AI:** natural as a risk/reward delay-calibration challenge against an AI diver's recorded scores.
- **Group games (n>4):** the simultaneous-dive variant extends naturally to more divers per object, subject to the no-contact caution.
- **Physical-skill model:** the delay-dependent success function (R3) is the entire model — a clean example of the framework's dexterity-with-difficulty-gradient pattern.
- **What is lost:** the historical specificity of the 1834 observation and its recorded social enjoyment have no gameplay analogue and remain in the entry's cultural framing only.
## Implementation hints
The delay-as-difficulty-dial (R2-R4) is what elevates this above a trivial fetch task — implement the success-probability curve so it strictly decreases with delay while the score reward strictly increases, creating a genuine risk/reward optimum rather than a dominant strategy at either extreme.
Download as JSON
Download the full game corpus entry for this game as structured JSON.
Download JSONCite this page
Murrumbidgee (Diving Retrieval Race) (Australia). (n.d.). Games of Countries. https://gamesofcountries.com/games/au-australia/murrumbidgee
@misc{gamesofcountries-au-australia-murrumbidgee,
title = {Murrumbidgee (Diving Retrieval Race) (Australia)},
year = {n.d.},
publisher = {Games of Countries},
url = {https://gamesofcountries.com/games/au-australia/murrumbidgee}
}