Sanity-check against known extremes

Test your estimate against the clearly too-high and too-low bounds to calibrate your range.

Why it works

Even when the precise estimate is unknown, the extremes are often knowable: the number of piano tuners in Chicago cannot be zero and cannot be ten million. Bounding the estimate from both sides constrains the plausible range and catches answers that drift outside the plausible region due to an error in the decomposition. The bounds provide a consistency check without requiring the answer itself.

How to do it

1. After your initial estimate, ask: what is the lowest it could plausibly be? What is the highest?
2. Identify what would have to be true for the answer to be at each extreme.
3. Check whether your estimate is inside the plausible range.
4. If it is near an extreme, re-examine the decomposition step that is driving it there.

Evidence

Bounding techniques are a standard element in expert estimation and risk analysis. They reduce the impact of overconfidence and anchoring by forcing engagement with the tails of the distribution before settling on a central estimate. (mechanistic)

Sanity checks catch large errors but not small ones — an estimate can pass the extreme test while still being significantly off. They are a minimum check, not a verification.

Common mistake

Setting the bounds symmetrically around your central estimate (e.g., ±50%) rather than thinking independently about what would need to be true for the extreme cases to hold.