← all games
Treasure Island

Treasure Island

#825BGG ↗

2018 · 2-5 players · 45min · weight 2.20 · 7,101 ratings

port: no portdifficulty: Hard
BGGv4 widev4v4 deep5Sources2Rules cardCandidateDeep dive
Bayes
6.75
Users rated
7,101
Owned
12,729
Wishing
3,599

At a glance — v4 wide

Controlled-vocabulary primitives + 8-axis MDA aesthetic vector. Vocab v1.

Core loop (micro)

Question Long John, draw a clue card, mark a search area on the map, then narrow the treasure's location.

MDA aesthetic vector (0–3)
Sensation
1
Fantasy
3
Narrative
3
Challenge
2
Fellowship
3
Discovery
3
Expression
1
Submission
1
Mechanic-interaction primitives (5)
  • [3]info_asymmetry_stable— “one player embodies Long John, trying to mislead the others in their search for his treasure
  • [3]bluff_layer— “Long John trying to mislead the others; question him about location of his treasure — or perhaps his misdirections
  • [3]partial_observability— “deduction mechanism; round after round they question him and explore the island following his directions
  • [2]escalation_then_resolution— “hunt reaches its climax with Long John's escape, when he will make a final run to get the booty for himself
  • [2]variable_player_powers— “one player embodies Long John with secret unit deployment; variable player powers mechanism
discovery_score: 0.427

Archetype fits — v4 deep

How well this game shape maps to mobile archetype templates. Composite is a weighted sum of the 10 fit dimensions.

ArchetypeCompositeLTFSessionComboArcShare5inOnboard
Coop
Crew coordinating against Long John — semi-coop with asymmetric traitor-shape; info_asymmetry_stable strength 3 is load-bearing.
5.307.0748344
Balatro
Storytelling deduction has tight climax but no engine, no run structure, asymmetric roles.
4.205.0427344
Snap
Strong bluff_layer but session is 45min and asymmetric one-vs-many doesn't compress to 3-min Snap match.
4.004.0327343
Wordle
Storytelling-deduction needs human Long John; doesn't shrink to 90s puzzle.
3.905.0326354
Cozy
`bluff_layer` strength 3 → Hard tier. Cap 3 applied. Adventure tone is exciting, not cozy.
3.903.0426344

Rules card

Synthesized from sources below. Readiness: draft-ready. Confidence: 0.72.

Readiness

draft-ready (confidence=0.72, rules=0.65, fun=0.85). BGG rank: 825; year: 2018; weight: 2.20; playtime: 45 min

SourceQualityRoleNote
bgg_comments0.75player voicepositive/player-voice sample
llm_memory0.65draft synthesissonnet-self-rated-7

Core Loop

Treasure Island is a one-vs-many deduction game where one player is Long John Silver — a pirate who has hidden treasure somewhere on the island map — and all other players are adventurers searching for it. Long John Silver was captured at the start but will eventually escape; the race is for the adventurers to find the treasure before he breaks free and recovers it himself.

Each round, Silver is compelled to give clues: he draws a clue card that specifies what type of hint he must give (e.g., "the treasure is not in the forest," "the treasure is within 3 spaces of a palm tree," "point in the direction of the treasure"). Silver must give the clue honestly but chooses how to phrase it within the card's constraint. Adventurers use these clues to narrow down the island map and commit searches (placing search markers in specific areas). After a fixed number of rounds, Long John Silver escapes; he then reveals which hidden tile he marked at game start (his secret), and all players who haven't found it lose while Silver wins.

Turn Structure and State

  • No manual/BGA/transcript source is present; card relies on memory plus BGG context.
  • BGG description anchor: Long John Silver's crew has committed mutiny and has him cornered and tied up! Round after round, they question him about the location of his treasure and explore the island following his directions — or perhaps his misdirections? Who knows... The old sea dog is surely planning an escape, after all, after which he will definitely try to get his treasure back. Treasure Island is a game of bluffing and [...]

Win Condition and Arc

Adventurers win if any of them places a search marker on the treasure's hidden location before Silver escapes. Silver wins if he escapes (fixed round trigger) and the treasure is not yet found — he immediately reveals the treasure and claims victory. The arc is a narrowing funnel: early clues eliminate large regions; later rounds focus searches on small target areas.

Decision Primitives

BGG mechanisms: Deduction, Hand Management, Line Drawing, Measurement Movement, Secret Unit Deployment, Storytelling, Team-Based Game, Variable Player Powers

Memory-derived primitives:

  • Constrained honest clue-giving (Silver must satisfy card conditions but has some phrasing latitude)
  • Deduction via cumulative clues (players circle/mark the map over multiple rounds)
  • One-vs-many asymmetry (Silver is a single constrained player vs. cooperative adventurers)
  • Secret hidden location (Silver marked the treasure spot in secret at setup)
  • Escape countdown (Silver escapes and self-searches after X rounds)
  • Physical map annotation (players mark their own maps with each clue's implications)

v4 controlled primitives: info_asymmetry_stable, bluff_layer, partial_observability, escalation_then_resolution, variable_player_powers

Top iOS archetype fits: coop 5.3, balatro 4.2, snap 4.0.

Why It Is Fun

The tension of constrained clue-giving is unique — Silver must reveal truth but can still misdirect within the card's rules. Players debate clue interpretation around the table. The personal map annotation creates satisfying "aha" moments when multiple clues intersect. The escape countdown creates urgency.

Player-voice evidence:

  • Thematic deduction game, with a very tense and close ending. We liked it a lot at Essen and bought one copy.
  • My family loves playing this game. No other games like it in our collection. We've gotten 10+ plays and still having great time.
  • Super fun game of finding booty! Not sure how well balanced it is, but maybe that doesn't matter - as long as everyone has a good time.
  • Wonderful experience; going to love playing this every time, but that's partly because the Treasure Island theme and the cartography mechanics are exactly what I wanted. I think it's also because of the relaxed attitude I take into this...
  • Fun game. Quick to pick up. Some of the colors on the board can be hard to see on the map. 8/10

Friction and Failure Modes

  • Treat Sonnet-memory edge rules as draft until confirmed by manual, BGA, or transcript.
  • Needs at least one stronger rules authority before final extraction use.

Translation and Design Hooks

  • Use this card to ask: which primitive carries the fun if theme/licensing is removed?
  • For iOS, look for short-session compression, clear state visualization, and a digital-only twist.
  • For new tabletop design, look for the tension source and decide whether to preserve or invert it.

Edge Rules and Gotchas

  • Silver cannot give false information — the card dictates what must be true about the clue
  • Some clue cards require Silver to physically gesture (point direction), which is harder to be subtle about
  • Adventurers may disagree about map interpretation — no consensus mechanism, each marks their own copy
  • When Silver escapes, he searches the map himself: he goes to his secret tile, and if unclaimed he wins immediately
  • The map is a fixed shared island; Silver's secret tile is marked on a personal sheet at game start before anyone sees it

Sources Used

[
  {
    "kind": "bgg_comments",
    "path": "data/bgg_comments/242639.txt",
    "quality": 0.75,
    "note": "positive/player-voice sample"
  },
  {
    "kind": "llm_memory",
    "path": "data/llm_memory_sonnet/242639.md",
    "quality": 0.65,
    "note": "sonnet-self-rated-7"
  }
]

Sources (2)

Inputs to rules-card synthesis. Click any pill with ↗ to open the original source.

BGG comments0.75LLM memory0.65