Decrypto

Readiness

ready (confidence=0.90, rules=0.90, fun=0.90). BGG rank: 117; year: 2018; weight: 1.83; playtime: 45 min

Core Loop

Decrypto is a team-based word game for 3-8 players split into two teams. Each team has a screen hiding four keyword cards (words visible only to their team). Each round, one player from each team is the "Encryptor" — they receive a secret 3-digit code (e.g., 3-1-4) and must give clues that help their teammates guess which keyword positions the numbers refer to, without being so obvious that the opposing team can crack it.

After clues are given and written on a public whiteboard, the opposing team tries to guess the code (called "Interception"). Then the Encryptor's own team guesses (they know the keywords, so they're matching clues to positions, not guessing the words themselves). Correct interceptions by the opposing team give that team a black token; failed guesses by your own team give your team a white token.

Turn Structure and State

• BGA tutorial is present; useful for exact turn flow and implementation gotchas.
• BGG description anchor: Players compete in two teams in Decrypto, with each trying to correctly interpret the coded messages presented to them by their teammates while cracking the codes they intercept from the opposing team. In more detail, each team has its own screen, and in this screen they tuck four cards in pockets numbered 1-4, letting everyone on the same team see the words on these cards while hiding the words from the [...]

Win Condition and Arc

Teams lose if they collect 2 white tokens (miscommunication = failure) or if the opposing team collects 2 black tokens (interceptions = cracked). Teams win if the opposing team reaches 2 black tokens first. The meta-game is the clue archive — each round, the whiteboard fills with more clues, and the opposing team uses the growing history to narrow down which keywords belong to which position.

Decision Primitives

BGG mechanisms: Communication Limits, Paper-and-Pencil, Targeted Clues, Team-Based Game

Memory-derived primitives:

• Two-team simultaneous play (each team encrypts and intercepts each round)
• Clue-giving under dual pressure (cryptic enough to avoid interception, clear enough for own team)
• Code system (3-digit codes from 1-4, referring to 4 keyword positions)
• Running clue history (all past clues written on whiteboard — patterns accumulate across rounds)
• Interception mechanic (opponents guess after hearing clues)
• Miscommunication penalty (own team fails to decode = white token = loss condition)

v4 controlled primitives: semantic_overloading, communication_constraint, info_asymmetry_stable, _other:intercept_decode_meta, attrition_clock

Top iOS archetype fits: coop 6.2, balatro 4.7, cozy 4.7.

Why It Is Fun

Decrypto starts easy and gets exponentially harder as the whiteboard fills. By round 5, opponents can often guess from clue patterns alone without knowing the keywords. The encryptor must give clues that are "obvious to partners but opaque to opponents" — which requires creative lateral thinking. Teams develop shared shorthand, and the moment you realize the opponent has cracked your code is genuinely tense.

Player-voice evidence:

• Me arrependi de ter vendido e recomprei! Que jogo party bom! O jogo é muito inteligente e os componentes são especiais para ele, um dos melhores jogos party
• Fairly lighthearted when played casually but will get tense with competitive teams.
• Codenames++. Outclasses it in general, but also breaks apart at very high player counts and makes a really bad first impression. Worth playing the second and third time, trust me on this.
• I like a lot the dynamics of this game, how clever you have to be to come up with more and more difficult hints as the game progresses. I am impatient to play some more with a bigger group so it will be even better!
• Very fun way to give your friends word puzzles with the added challenge of making clues that are simultaneously easy for one team but hard for the other. Can lead to funny moments where each side is confused by the seemingly unrelated...

Friction and Failure Modes

• Treat Sonnet-memory edge rules as draft until confirmed by manual, BGA, or transcript.

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

• Clues cannot be a form of the keyword word itself (e.g., can't clue "Bicycle" with "Bike")
• All previous clues are permanently visible on the whiteboard to both teams — there is no way to reset the information trail
• The codes are drawn randomly each round (3 chips from a bag of four) — you don't choose which keyword to clue
• If both teams would lose in the same round, the round continues until one team has fewer tokens
• Decrypto can be played with teams of just 1 person per side as a 2-player variant

Sources Used

[
  {
    "kind": "bga_tutorial",
    "path": "data/bga_tutorials/225694.md",
    "quality": 0.85,
    "note": "BGA implementation rules summary"
  },
  {
    "kind": "bgg_comments",
    "path": "data/bgg_comments/225694.txt",
    "quality": 0.75,
    "note": "positive/player-voice sample"
  },
  {
    "kind": "llm_memory",
    "path": "data/llm_memory_sonnet/225694.md",
    "quality": 0.65,
    "note": "sonnet-self-rated-8"
  },
  {
    "kind": "wikipedia",
    "path": "data/wikipedia/225694.md",
    "quality": 0.55,
    "note": "bare-title"
  }
]