Current designs focus on gates and escalating involvement with the play; what would a tightly-scoped ‘bottle’ scenario look like where the investigators are already deep inside a single ongoing production (cast or crew), and the mystery is which specific rehearsal changes, script edits, or backstage rituals are making this version uniquely dangerous—and how do we structure clues, rehearsals-as-encounters, and sanity beats so the puzzle is to diagnose and excise the malignant elements without killing the show outright?

king-in-yellow-rpg | Updated at

Answer

Outline: a single-theater bottle mystery where the active case is: “What change made this production lethal, and can we cut it without ending the show?”

  1. Core bottle frame
  • One location: a theater in tech week / final rehearsals.
  • PCs: all cast/crew (or attached roles: producer, critic embedded, consultant doctor, etc.).
  • Premise: earlier productions of this script were weird but survivable; this one is wrong in a new way.
  • Win state: identify and remove/alter the malignant element while still getting to opening night (or a safe “preview”).
  1. Define the malignant element upfront (GM side) Pick 1–2 concrete levers:
  • A new scene or monologue.
  • A specific line revision.
  • A blocking choice / stage geography.
  • A prop, mask, or set piece (bearing the Yellow Sign).
  • A backstage “focus ritual” introduced by someone. Prep 3–4 possible suspects but only 1–2 are truly dangerous; others are red herrings or low-grade weird.
  1. Clue map: three veins A) Text
  • Script pages, edits, notes, marginalia, new draft from a mysterious source.
  • Clues show: which lines repeat like an incantation; where the Yellow Sign appears in stage directions or logos; who is pushing for which edit. B) Performance
  • Rehearsal runs, scene work, table reads, vocal warm-ups.
  • Clues show: which scenes trigger fugues, shared hallucinations, bleed-through to Carcosa; who improves when they lean into it. C) Backstage
  • Costumes, masks, props, lighting cues, sound design, pre-show rituals.
  • Clues show: objects that don’t photograph right, light cues that open “windows,” rituals that sync people’s dreams.

Structure the clue map so:

  • Every node belongs to at least one vein.
  • Each wrong suspect has 2–3 strong clues pointing at it.
  • The real malignant element has both “text” and “performance” fingerprints, so players must triangulate, not just pick the creepiest prop.
  1. Rehearsals as encounters Treat each rehearsal block as a repeatable encounter type with variants:
  • Table read: low physical risk, high clue yield about text (weird phrasing, compulsive rereads).
  • Staged run of a trouble scene: medium risk; triggers shared visions, identity blur, emotional spikes.
  • Full act run / tech: high risk; multiple harmful elements coincide (line, blocking, lighting, sound). Procedures:
  • Before each rehearsal, players can propose one change: cut/soften a line, alter blocking, swap a prop, skip a warm-up, or insist on “as written.”
  • GM adjusts risk and clue output accordingly: more untouched elements = richer visions but higher SAN cost; more safety edits = safer but foggier.
  • Treat each rehearsal like an investigate+stress scene: everyone there rolls for insight (clues) and for mental strain (SAN or stability).
  1. Sanity beats and pressure Use three main SAN triggers:
  • Recognition: realizing a mundane theater choice is also an occult pattern (blocking mirrors a Yellow Sign spiral; a mask is “known”).
  • Dissociation: losing track of where the scene stops and self begins (PC called by character name off-stage; lines bleed into arguments).
  • Contagion: seeing the text spread (crew quoting the play outside work; posters shift, critic emails echo lines). Make SAN costs small but frequent. Emphasize:
  • Group checks when the cast as a whole hits a dangerous beat together.
  • Personal checks when a PC’s specialty is weaponized (director realizes their staging is the ritual; designer sees their set as gate geometry).
  1. Puzzle: diagnose without killing the show Frame key choices around three axes instead of “shut it down vs not”:
  • Reduce intensity: soften, paraphrase, or re-stage beats instead of cutting whole scenes.
  • Contain context: limit who sees what (closed rehearsal, rewriting programs, changing ad art).
  • Redirect symbolism: move the Yellow Sign from sacred center to marginal, mocked, or broken positions on stage. Mechanically:
  • Track two simple meters: SHOW INTEGRITY and CARCOSA LOAD (0–5 each).
    • Cuts/defiance lower LOAD but also lower INTEGRITY.
    • Leaning in raises both.
  • Offer 2–3 “safe-ish” endings:
    • Opening night happens with LOAD below a threshold: weird but survivable; Carcosa scars remain.
    • Dress rehearsal as final performance; show “dies in the womb” but the cast walks away.
    • The show mutates into something else (a farce, a protest piece), diffusing the ritual while saving some careers.
  1. Example scene flow (3–5 session bottle) Session 1
  • Cold open: mid-rehearsal incident (minor harm, scary bleed-through).
  • Orientation: map roles, relationships, first suspicions.
  • Scenes: script meeting (text vein), costume fitting (backstage), short table read (low-risk encounter). Session 2
  • Focus on 2–3 suspect elements (new scene, mask, pre-show meditation).
  • Run each in rehearsal once with minimal edits; harvest strong visions and first major SAN hits. Session 3
  • Midpoint choice: cast vote / producer ultimatum—what must change to keep the show alive?
  • PCs propose a “safe” version of the production: cuts, rewrites, blocked symbols.
  • Rehearsal to test the new plan; see which weirdness remains. Session 4–5 (if desired)
  • Escalation: sabotage, true believer pushback, external pressure (investors, critic, censor) to restore the “brilliant” dangerous version.
  • Final rehearsal / invited preview as climax: play out the last run with chosen edits, resolve meters, show fallout.
  1. Player safety
  • Lines/veils for real-world theater trauma, workplace abuse, gaslighting, and mental health.
  • X-card or equivalent; encourage step-out talk as normal between beats.
  • Keep identity erosion clearly fictional: no using player names, no referencing real jobs/diagnoses.
  • Allow any player to “fade out” their PC between acts (illness, quitting the show) without penalty.

This gives a tight, replayable bottle design where the mystery is about diagnosing which artistic choices are Carcosan, the encounters are rehearsals, and success is cutting or transforming the ritual just enough that the show can go on without becoming a gate.