The piano tuner problem, explained
It is the question every physics student gets asked, and almost nobody can answer on instinct. Here is how to reason your way to it in under five minutes, with nothing but things you already know.
How many piano tuners are there in Chicago?
The question is traditionally attributed to Enrico Fermi, who liked to spring it on students at the University of Chicago. It is deliberately outrageous. You have never met a piano tuner. You have no idea how many pianos exist. You may never have been to Chicago. And Fermi expected an answer in a couple of minutes, without reference books.
The point was never the piano tuners. The point was to demonstrate that a question with no obvious answer usually has an obvious method, and that most people never think to look for it.
Step one: what would you multiply?
Before guessing at anything, work out the shape of the calculation. We want a number of tuners. Tuners exist because pianos need tuning. So:
- How many pianos are there in Chicago?
- How often does a piano get tuned?
- How many tunings can one person do in a year?
Divide the first two multiplied together by the third, and you have your answer. Notice that we have not guessed a single number yet — we have only decided what to guess. This is the step people skip, and skipping it is why the question feels impossible.
Step two: guess each piece
How many people live in Chicago? Around 2.7 million in the city proper. If you didn't know that, “a big American city, so a couple of million” gets you to the same place.
How many households? At roughly 2.5 people per household, about 1 million.
How many households own a piano? Here you have to think from your own life. Of the homes you have been inside, what fraction had a piano? For most people the honest answer is somewhere between one in twenty and one in fifty. Call it one in thirty — so about 33,000 household pianos. Then add institutions: schools, churches, bars, concert halls. Those own more pianos per building but there are far fewer of them; adding a few thousand is reasonable. Call the total 40,000 pianos.
How often is a piano tuned? Manufacturers recommend twice a year. Actual human beings are worse than this. A concert hall tunes constantly; the piano in your aunt's living room has not been tuned since 2003. Average it to once a year — 40,000 tunings annually.
How many tunings can one tuner do? A tuning takes an hour or two, plus travel. Say four a day, five days a week, with a few weeks off: 4 × 5 × 48 ≈ 960. Round to 1,000 tunings per tuner per year.
Step three: multiply
Forty. Not four, not four hundred. And here is the striking part: directories of professional piano tuners in Chicago list numbers in roughly this range — dozens, not hundreds. Someone who has never thought about pianos for a single minute of their life just produced a defensible figure using only their own memories of other people's living rooms.
What just happened
Every individual guess in that chain was shaky. The piano ownership rate could easily be double or half what we said. The tunings-per-day figure was invented on the spot. If you asked ten people to run this calculation, you would get ten different sets of inputs.
And you would get answers clustered around the same order of magnitude. That is the quiet miracle of the method: the errors are independent, so they partly cancel. Guessing high on pianos and low on tuning frequency lands you back where you started. The chain is more robust than any link in it.
An estimate built from five shaky guesses is usually better than a single confident one — because the single confident guess has nothing to cancel against.
The variation worth practising
Once the method clicks, it applies to nearly anything with a supply and a demand. How many dentists does a city need? How many delivery vans does an online retailer run? How many hairdressers, plumbers, school buses, ambulances?
They all decompose the same way: how many things need the service × how often ÷ how much one provider handles. Learn that one shape and a whole category of unanswerable questions becomes routine.
The harder and more interesting cases are the ones where the shape isn't obvious — where you have to invent the decomposition before you can use it. That is where estimation stops being arithmetic and starts being thinking.
Every Zoggle puzzle works exactly like this — an impossible-sounding question, one guess, and then the worked reasoning so you can see where your thinking diverged from the answer.