Current designs mostly treat the forbidden play and Yellow Sign as inputs that warp a pre-existing investigation; what if you invert this and treat each investigation as an improvised performance of the play itself, where clue choices double as staging decisions (casting, cuts, audience, venue) that literally determine which acts of The King in Yellow manifest in-world—how would that reframing change how you write clues, handle failure, and pace sanity pressure when “solving the case” is indistinguishable from successfully mounting the play?

Answer

Treat each mystery as staging the play: every clue choice is also a production choice.

1. Clues as staging decisions

• Write key clues as forks in how the play manifests:
  • Casting: which NPC becomes which role; clues about suspects double as “who we center on stage.”
  • Cuts/edits: choosing what evidence to ignore = cutting scenes from the in-world play.
  • Audience: which faction learns the truth = who attends opening night.
  • Venue: where you confront / reveal = where Act II/III are staged.
• Each fork tags: which “Act” manifests, which motifs appear, how much Yellow Sign presence increases.

2. Failure as bad theatre, not dead ends

• Replace hard stalls with mis-stagings:
  • Wrong casting → the “role” attaches to a fragile NPC; they break, flee, or invite harsher Carcosa bleed.
  • Missed clue → that scene plays offstage; you hear about it later as distorted rumor.
  • Botched roll → the act still happens, but with worse framing (public instead of private, crowded instead of intimate).
• Outcome: the play always goes on; failure just makes it louder, crueler, or less controlled.

3. Pacing sanity as rehearsal → preview → premiere

• Structure scenarios in three passes:
  • Rehearsals: low SAN, mostly implication. Clues hint at roles, lines, motifs; dreams and déjà vu are soft.
  • Previews: medium SAN. First full “scenes” from the play echo real events; Yellow Signs show up in design notes, props, documents.
  • Premiere: high SAN. Multiple acts overlay reality; identity and memory can flip, but only in scenes the PCs themselves chose to stage.
• Tie SAN cost to prominence: the more public and polished the “performance” choice, the higher the pressure.

4. Concrete writing pattern

• For each core clue node, prep:
  • A factual payload (what it tells them about the case).
  • A staging tag: CAST / CUT / AUDIENCE / VENUE.
  • 2 outcomes: A (contained, private, lower SAN), B (spectacular, public, higher SAN) – both move the investigation.
• Example: a rehearsal script
  • Fact: reveals a missing scene implying a second victim.
  • Tag: CUT.
  • Option A: they quietly suppress the scene → second victim is spared but remains unknown; Carcosa stays subtle.
  • Option B: they insist on restoring it → second victim dies on- or off-stage in parallel; SAN spike, but big new leads.

5. Making “solving = staging” legible

• Explicitly tell players that every big investigative decision also shapes how the play appears.
• Use recurring language: “If you follow this lead, you’re effectively casting X as the Prince” or “You’re moving this scene onto a bigger stage.”
• Tie major reveals to curtain-raise moments: “Act I closes when you decide who confronts whom, where, and with who watching.”

6. Safety and control

• Let players set a ceiling on how far identity erosion goes this session (e.g., a simple 3-step slider: Firm / Frayed / Doubled).
• Offer off-ramps: they can choose smaller, more private stagings for reduced SAN and fewer reality edits.
• Debrief in terms of theatre, not pathology: talk about roles, masks, and scenes they regret staging.

Net effect: you write clues as production levers, treat failure as mis-staging rather than blockage, and pace sanity through how big and public the “acts” become, so that finishing the case always means bringing The King in Yellow fully to the stage—but in a way the group consciously co-authors.