When the Unusual Thing Stalls, Try Opposites. And Why Relationships Still Come First.
- Claudia Dot

- Jun 3
- 2 min read

There's a concept from UCB training that I keep coming back to, even though I've never set foot in their building. Not yet, anyway. But their philosophy adds useful depth to my own, so I pay attention.
The idea: find the unusual thing in your scene. The bank robber who cares more about the tellers' feelings than the money. The doctor who's afraid of blood. That one offbeat detail that makes the scene play.
"Unusual" is vague on purpose. It's whatever you notice as unexpected.
But here's what I've found teaching this. Some unusual things land and some don't. A quirky habit can feel arbitrary. A random weirdness can stall. When the unusual thing doesn't spark, the scene goes nowhere.
That's when I reach for a different tool: opposites.
When Unusual Stalls
Not "what's weird about this character." Instead, "what's the opposite of what this role expects?"
The CEO who can't make eye contact.
The firefighter who's scared of heat.
The stand-up comic who hates attention.
The drill sergeant who whispers.
Opposites create instant friction against the audience's built-in expectations. You don't have to guess whether it's interesting. The tension is baked in.
But Here's Where I Take It Further
I teach improvisers to build from relationships first. Not from a gimmick. You start with two people who genuinely need something from each other. When you've got that foundation, the unusual thing stops being a quirky detail and becomes a clue about the relationship.
The bank robber who cares about feelings? That's not just unusual. It tells you something about how he sees himself in relation to others. The unusual thing becomes a window into why this character relates the way they do. Suddenly you're not playing a game of "spot the weird detail." You're playing game theory with real stakes.
The unusual thing is the setup. The relationship is the scene. And the game (the pattern of behavior that emerges when those two collide) is where the comedy lives.
Real-World Application
This isn't just for improv. In a pitch, a design, a tough conversation, name the expected move. Then imagine the opposite. But don't stop there. Ask what that opposite reveals about the relationship. That's where the insight goes from clever to useful.

The AI Angle
This maps onto how I think about prompt engineering. A model trained to produce the statistically likely answer is like a scene played straight. Predictable, safe, low friction. The opposites trick is asking: "What's the least expected useful answer?" But the relationship trick is giving the model a persona with wants and needs, not just a query pattern. That's where the outputs stop being clever and start being alive.
Try This Yourself
Next time you're stuck in a scene, a pitch, or even just a conversation, flip it. Name what's expected, then play the opposite. Build it into a relationship with stakes. Let the game emerge from the friction. You might be surprised what shows up when you stop looking for the unusual thing and start looking for the opposite one instead.





Comments