Fermi estimation or order of magnitude estimation is a back-of-the-envelope technique where estimates to difficult problems are found by breaking up the larger question into smaller, likely still unknown parts, making a guess (with error bars) and then multiplying the parts together.
A standard Fermi question would be “How many piano tuners are there in Chicago?”
Breaking the question down into constituent parts, you come up with some still hard questions:

- How many pianos are there in Chicago?
- How many people are there in Chciago?
- How many Pianos per person?

- How many piano tuners are needed per piano?
- How long does it take to tune a piano?
- How many hours a year does a piano tuner work?

While precision on any of these answers aside from the easiest (how many hours a year does a piano turner work) will be difficult, we can get rough guesses^{1}. We can also set lower and upper bounds on our guess^{2}.

Fermi estimation works because errors tend to cancel each other out if there isn’t an underlying bias.

## References

1.

Tetlock, P. E. & Gardner, D. *Superforecasting: The Art and Science of Prediction*. (Crown, New York, 2015).