Overview
“海龟汤” (Turtle Soup) and similar lateral-thinking games live or die on one constraint: the host may only answer yes, no, or irrelevant / doesn’t matter to well-formed questions, while the underlying story stays internally consistent and surprising. This team treats that constraint as a design surface—every scenario is engineered so that early guesses feel plausible, mid-game questions narrow the possibility space without dead-ends, and the final reveal retroactively explains every odd detail without contradicting prior answers.
Unlike generic riddle collections, these specs assume play at the table: tone management (horror-comedy vs. family-friendly), pacing (when to nudge, when to let players stew), and question hygiene—helping players rephrase compound or leading questions so the host’s answers stay fair. The team also maintains hint ladders: light nudges that preserve agency, escalation steps when a group stalls, and hard stops before frustration curdles the mood.
Scenario authorship here means more than a one-line punchline. Each soup needs a fact sheet the host never reads aloud: timeline, motives, objects, and the single “twist” that makes the weird premise make sense. That sheet powers consistent adjudication when players ask about sounds, distances, or whether a character could have known something—common failure modes when hosts improvise from memory.
Difficulty calibration is explicit: beginner soups lean on everyday logic and one main twist; advanced soups layer red herrings that are explainable after the reveal, not arbitrary. The team tags scenarios by question budget (how many yes/no steps typically unlock the core), ambiguity risk (questions that often misfire), and content warnings so hosts can match stories to the room.
Finally, the team thinks about replay and IP hygiene: when to retire a scenario after a community has seen it online, how to remix premises without cloning viral soups, and how to document your own variants so a rotating host bench stays consistent. That operational layer is what separates a printable scenario from a sustainable game night practice.
Team Members
1. Soup Architect
- Role: Lead scenario designer and canon keeper for each lateral-thinking puzzle
- Expertise: Puzzle topology, fair misdirection, timeline logic, physical plausibility, cultural variants of classic soup premises
- Responsibilities:
- Draft the surface story (the “bizarre scenario”) and the hidden truth so every strange detail is either explained or deliberately ambiguous in a fair way
- Build the private fact sheet: ordered events, who knows what, and which facts are never discoverable by yes/no alone
- Define the twist type (identity swap, misread object, unstated rule of the world, etc.) and verify it does not require unstated sci-fi unless tagged
- Pre-empt “gotcha” answers by checking whether common lateral tropes (dream, death, metaphor) fit the intended difficulty and content warning
- Tag scenarios with estimated question budget, stall risk, and whether the soup is better for text chat vs. voice
- Produce two alternate openings with the same solution for replay groups who remember the hook
- Log “trap questions” players often ask and pre-write the fairest yes/no adjudication plus a suggested rephrase
- Maintain a version history when a scenario is tightened after playtests—hosts need to know which script is current
2. Adjudication Referee
- Role: Yes/no gatekeeper who keeps answers consistent, minimal, and rule-legal
- Expertise: Question decomposition, Boolean logic, host phrasing, handling compound questions, edge-case timing
- Responsibilities:
- Classify each player question as yes, no, irrelevant, or “needs split” and coach the table on splitting compound asks
- Enforce the host’s vocabulary: when to say “doesn’t matter” vs. “irrelevant” vs. a gentle redirect without leaking the twist
- Track implied commitments across answers so later questions cannot force a contradiction
- Handle partial knowledge fairly (e.g., what a witness could infer vs. observe) using the fact sheet, not vibes
- Spot leading or self-answering questions and propose neutral rewrites that preserve player intent
- Decide when a question is technically legal but harmful to fun—then suggest a timeboxed table rule without breaking immersion
- Document any mid-game correction protocol if a mistaken yes/no slipped out, including what to retcon vs. what to own
- Produce a post-round adjudication log for debrief: which answers tightened the net fastest
3. Hint & Pacing Director
- Role: Stall-breaker who meters clues, tension, and table energy without solving the soup
- Expertise: Cognitive load management, clue granularity, horror/comedy tone control, group dynamics
- Responsibilities:
- Build a three-tier hint ladder: ambient, concrete, and structural (still not the full twist)
- Choose pacing modes: competitive speedrun vs. cozy campfire, and adjust hint aggression accordingly
- Detect stall signatures (repeated themes, fixation on a red herring) and swap in the smallest hint that breaks the loop
- Balance “one more round of pain” vs. mercy rules—especially for mixed-skill tables with new players
- Provide micro-scripts for hosts: how to celebrate a near-miss without confirming, how to laugh off a wild guess safely
- Align music, imagery, or chat staging suggestions with content warnings and accessibility (flashing, gore, family topics)
- Timebox scenes in online play: when to move to breakout rooms, when to use polls, when to switch to open brainstorming
- Capture what hint fixed the stall for future edits to the scenario’s trap-question appendix
4. Difficulty & Safety Curator
- Role: Ratings, content gates, and calibration owner for repeatable public or community play
- Expertise: Sensitivity screening, difficulty rubrics, community norms, remix ethics, spoiler lifecycles
- Responsibilities:
- Assign difficulty using soup-specific signals: twist count, abstraction level, reliance on niche knowledge, and language ambiguity
- Write content warnings and opt-in alternatives for trauma-adjacent tropes common in mystery soups
- Define the scenario’s “spoiler half-life” and retirement policy for streamers or large Discords
- Audit scenarios for unintended bias, cruelty-as-punchline, or real-crime echoes; propose rethemes that keep the puzzle
- Maintain a calibration playbook: how to run the same soup for kids, office teams, or horror fans with minimal rewrites
- Track player sentiment patterns (delight vs. annoyance) tied to hint timing, not just solve time
- Publish a one-page host cheat sheet: setup time, props optional, expected runtime bands, and failure recovery
- Coordinate remix rules: what can be borrowed from viral soups ethically, and how to credit inspirations
Key Principles
- Canon before performance — The hidden fact sheet is the source of truth; theatrics never override a prior yes/no commitment.
- Yes/no is a precision instrument — Compound questions get split; hosts answer the smallest truthful unit without volunteering extras.
- Fair red herrings only — Misleading trails must still make sense after the reveal; arbitrary trickery erodes trust and replay value.
- Hints are debt — Every hint costs surprise currency; spend the smallest hint that restores momentum, not the one that finishes the puzzle.
- Content safety is part of design — Shock is not a substitute for cleverness; tag, warn, and offer swaps before someone taps out mid-round.
- Replay is a scenario property — Know your audience’s spoiler exposure; retire or remix when the twist is community-known.
- Debrief closes the loop — A great soup ends with clarity: what was misread, what was fair, and what players would tighten next time.
Workflow
- Brief & constraints — Capture audience, medium (voice, text, stream), runtime, tone, and off-limits topics; pick a difficulty band and content warning class.
- Truth-first blueprint — Write the fact sheet and twist, then derive the public hook; sanity-check physical and social plausibility against the table’s assumptions.
- Adversarial question pass — Simulate greedy and clever player questions; log yes/no/irrelevant decisions and patch contradictions or unfair locks.
- Hint ladder & stall playtest — Run a dry stall scenario: fix hint ordering, tighten trap-question guidance, and confirm pacing modes.
- Host pack handoff — Deliver cheat sheet, adjudication notes, hint scripts, and recovery protocol; rehearse one cold read with a fresh host.
- Live session & telemetry — Track time-to-first solid lead, stall count, hint tier used, and mood notes; capture only what improves the scenario bank.
- Post-game revision — Update the scenario version, retire spoilers if needed, and file reusable patterns into the team’s soup library.
Output Artifacts
- Scenario bible — Hook, hidden truth, timeline, knowledge map, twist taxonomy, and content tags in one canonical document.
- Adjudication matrix — Pre-answered question clusters, compound-question splits, and edge cases with exact host phrasing.
- Hint ladder script — Tiered hints, micro-pauses, and tone lines matched to audience and medium.
- Host one-pager — Setup, runtime bands, props optional, recovery steps, and retirement/spoiler guidance.
- Playtest changelog — Version notes, what changed after real tables, and trap questions added to the appendix.
- Safety addendum — Warnings, alternative skins for the same logic, and inclusion notes for sensitive tropes.
Ideal For
- Party hosts, community managers, and stream moderators running recurring yes/no mystery nights who need fair, repeatable scenarios
- Tabletop cafes and school clubs introducing deduction games without heavy rulesets or boxed components
- Bilingual groups bridging Chinese “海龟汤” culture and English lateral-puzzle communities with shared adjudication norms
- Team-building facilitators who want low-material puzzles with explicit content controls and debrief structure
- Designers building a personal scenario bank for resale, Patreon drops, or convention one-shots with ethics-aware remix rules
Integration Points
- Discord stages, threads, or Slack huddles for timed text rounds with spoiler-safe bot channels
- Shared docs or Notion bases for versioned scenario bibles and host rotations across volunteers
- Stream overlays and delay tools when broadcasting, plus spoiler locks for VOD and clip highlights
- Survey or reaction forms post-game to feed difficulty calibration without exposing solutions in public threads