Trains: Fracture recovery — the ability to re-establish shared reality when two players have diverged.
Setup: Two players begin a scene. A coach secretly whispers a different context to each player — Player A is told "you're at a funeral," Player B is told "you're at a surprise party." They begin playing. The audience watches the fracture form. At any point, either player can attempt to repair: yield to the other's reality, or find a justification that bridges both. This manufactures the fracture so you can focus purely on the repair muscle. 2-3 minutes per pair, rotate through the room. Audience task: spot the exact moment the fracture becomes visible.
What to notice: The natural instinct when you sense a fracture is to push harder on YOUR version. This always makes it worse. The repair only happens when someone yields — simply drops their version and joins their partner's — or when someone bridges by finding the justification that makes both realities true. Yielding is the faster repair and players need permission to do it. You can just drop your thing. That's not failure; that's support.
Side-coaching: "Notice the gap." (Early, until players self-detect.) "You don't have to be clever — you can just go with theirs." "What does your partner think is true?"
The deeper lesson: In performance, nobody whispers to you. You'll create these fractures yourself by accident — a missed offer, a misread tone. Now you know what repair feels like. This drill builds the diagnostic skill (noticing the fracture) and the repair skill (yielding or bridging) as a practiced reflex.
Variation — Tonal Fracture: Instead of whispering locations, whisper genres or emotional tones. One player gets "tender," the other gets "competitive." Tonal fractures are the most common real fractures in experienced players and the hardest to name in the moment. This variation trains the subtler diagnostic: "What does my partner think the tone is?"
Variation — Triple Reality (advanced): Three players, three whispered contexts. The repair requires finding coherence across three perspectives — much harder, and a direct training for group scene dynamics.