YINSH

In YINSH, the players each start with five rings on the board. Every time a ring is moved, it leaves a marker behind. Markers are white on one side and black on the other. When markers are jumped over by a ring they must be flipped, so their color is constantly changing. The players must try to form a row of five markers with their own color face up. If a player succeeds in doing so, he removes one of his rings as an indication that he has formed such a row. The first player to remove three of his rings wins the game. In other words, each row you make brings you closer to victory-but also makes you weaker, because you have one fewer ring to play with. Very tricky!

{
  "game_id": 7854,
  "name": "YINSH",
  "core_verb": "move ring, flip discs",
  "decision_shape": "combinatorial",
  "reward_schedule": "mixed:delayed+immediate",
  "aesthetics": [
    "Challenge"
  ],
  "core_loop_pitch": "Slide a ring leaving a marker; jumping flips colors. Make 5-in-a-row to claim a ring; first to remove 3 wins.",
  "mobile_translation_difficulty": "Easy",
  "translation_difficulty_reason": "Pure abstract on a fixed grid translates easily; small unofficial implementations exist (Shyring, BGA beta) but no canonical publisher port.",
  "direct_digital_port": "Shyring (iOS, unofficial)",
  "closest_loop_translation": "none yet",
  "primitive_tags": [
    "self_weakening_victory",
    "marker_polarity_flip",
    "pattern_completion_5_in_row",
    "ring_movement_constraint",
    "abstract_perfect_information",
    "win_by_resource_loss"
  ],
  "confidence": 0.7
}

{
  "game_id": 7854,
  "name": "YINSH",
  "mechanics": {
    "core_verb_long": "On your turn you slide one of your five rings in a straight line across a hexagonal grid of 85 spaces, dropping a two-sided marker (white-up or black-up) onto the space the ring just left. The ring may glide freely over empty spaces but the moment it lands beyond a contiguous run of markers, every marker it jumped is flipped to its opposite color. The cadence is slow and deliberate: pick up a ring, trace a line, plant a disc, and watch the board's color polarity change beneath you.",
    "core_loop_long": "Setup: players alternate placing five rings each on the truncated hex-star board. On each turn you must move exactly one of your rings in a straight line; you may glide over empty spaces, but you must stop on the first empty space after any markers, flipping every jumped marker to its opposite color. After moving, check whether the new flipping created a row of five markers of your color anywhere on the board — if so, you remove that row and one of your own rings as a victory token. The endgame trigger is asymmetric: the first player to score three rings wins, but each ring removed is one fewer mover, so a leading player has fewer tools to defend with. The game can also end if either player runs out of markers in the supply.",
    "decision_space": "Each turn you balance flipping for offense (building toward your own five-in-a-row) against flipping for defense (breaking up an opponent threat) while managing the geometry of your own ring placements, since rings cluster after each move and can crowd themselves into uselessness. The option set at a typical mid-game turn is roughly 15-30 legal ring moves, but only a handful are sharp; the difficulty is reading which moves create double threats versus which leave you exposed two turns later. The signature tension is that scoring is self-handicapping — every five-in-a-row you complete strips you of a ring, so leading players must convert with caution. Identifying a 'good move' requires forecasting whose color a chain of flips will eventually settle on, which is non-obvious because every traversal toggles state.",
    "skill_expression": "Spatial pattern recognition is dominant: strong players see latent rows two and three flips ahead, recognizing diagonal and orthogonal alignment threats before they materialize. Risk modeling is the second-largest skill — knowing when to score and accept the ring penalty versus when to defer and consolidate position. A third skill is tempo: the player who forces the opponent to defend rather than develop tends to dictate the geometry of the late game. There is essentially zero memory load and no hidden information, so the game is a pure visual-search and forecasting test, much like high-level Othello with a movement layer.",
    "tactile_dependency": "low",
    "tactile_dependency_reason": "All state is fully legible: 85 hex cells, ten rings, a finite marker pool. Flipping is mechanically pleasing in person but encodes a deterministic toggle that maps cleanly to a tap-and-animate digital interaction. Numerous online implementations (yinsh.io, BGA) demonstrate the game survives translation without losing essence."
  },
  "theme": {
    "promise": "A pure abstract duel of geometry and self-sabotage: every win condition you achieve weakens you, so you must sprint to a finish line that keeps moving away.",
    "setting": "abstract",
    "narrative": "none — abstract",
    "audience": "hobbyist Eurogamer, designer-game-aficionado",
    "art_direction": "Clean Bauhaus-flavored production from Don & Co.: matte black-and-white discs, lacquered wooden rings, a soft cream board with thin hex lines. Functional, austere, recognizable as part of Kris Burm's GIPF Project visual lineage."
  },
  "translation": {
    "digital_meta_layer_ideas": [
      "Daily puzzle mode: pre-set board states with 'win in 2 moves' or 'avoid loss' constraints, leaderboard by solve speed, à la Chess.com puzzles",
      "Asynchronous ranked ladder with Glicko rating, per-season cosmetic ring/disc skins, and an Elo-gated weekly tournament",
      "Roguelite 'gauntlet' overlay: face a sequence of AI opponents with escalating quirks (e.g., 'opponent starts with one extra ring', 'board has dead cells'); pick a relic between matches that bends the abstract rules",
      "Coaching/replay layer: post-game heatmap of move quality vs engine, with branching 'what if you'd played here' tutor"
    ],
    "closest_mobile_genre": "puzzle/zen",
    "live_service_potential": "low"
  },
  "confidence": 0.78,
  "extraction_version": "v3"
}