PAYCHECK ZONE
“models cluster where the funding is. that's the paycheck zone.”
The Lesson
When you see models clustered in a 'probability band,' ask about funding. Models are expensive to build and maintain. If your model falls outside the consensus, you don't get paid. So models inevitably cluster in what I call the paycheck zone-the range that keeps funding flowing. Models that drift outside get tweaked back in or discarded. This guarantees the appearance of consensus regardless of underlying truth. The lesson applies beyond climate: any field where models need continuous funding will show this clustering effect. Always ask: where's the paycheck zone?
Real-World Example
A founder evaluates consulting forecasts for their industry. Every firm predicts growth in a narrow range. They ask: what happens to a consultant who predicts dramatically different numbers? They don't get hired again. The forecasts cluster in the paycheck zone. The range clients want to hear. The founder learns to discount consensus forecasts and look for analysts with skin in the game.
Watch Scott explain this lesson
Watch Lesson on Locals →