Fermi Estimation
Order-of-magnitude reasoning: how to get usably close to the truth without having all the data
What is Fermi estimation and how does it help you make better quantitative judgments?
Fermi estimation is the practice of making rough but principled quantitative estimates by decomposing an unknown into knowable sub-problems, estimating each, and combining them. Named for physicist Enrico Fermi, who was renowned for accurate estimates from minimal data, it is used in science, engineering, and everyday decisions to calibrate intuitions and check whether a number is in the right ballpark — not to achieve false precision.
The classic Fermi problem — 'how many piano tuners are in Chicago?' — is solved not by looking up the answer but by breaking it into knowable pieces: Chicago's population, fraction of households with pianos, tuning frequency, and a tuner's daily capacity. Each sub-estimate has error, but the errors partly cancel when combined, and the result is usually within a factor of two or three of reality. Fermi estimation is not about exactness; it is about replacing 'I have no idea' with 'it is probably between X and Y' — a much more useful epistemic position. Here are the practices, with honest evidence.
Practices
- Decompose the unknown into knowable sub-problems
Break the question you cannot answer directly into smaller questions you can.
- Anchor on what you know and scale from there
Start from a number you are confident about, then reason to the unknown.
- Estimate in ranges, not point estimates
Instead of "my estimate is 500," say "I think it is between 200 and 2000."
- Check units and scales for consistency
Catch order-of-magnitude errors by ensuring your units are consistent across the estimate.
- Sanity-check against known extremes
Test your estimate against the clearly too-high and too-low bounds to calibrate your range.
- Track your estimates and calibrate
Compare your Fermi estimates to actual figures when you can, and use the gap to improve future estimates.
Practice this with IX Coach
IX Coach: 7 days free, then $40/month (about $1.30/day).
Related concepts
- Thinking in Bets
Probabilistic thinking, separating luck from skill, and building a better decision process
- Thinking, Fast and Slow, Made Usable
Two systems, the biases they create, and when to slow down
- Circle of Competence
Know what you know, know what you do not know, and act accordingly
- First-Principles Thinking, Made Usable
Reasoning up from fundamentals, with the mechanism behind why it beats analogy