# Fermi Estimation Activity — How Many Balls Fill This Room?

Source: https://decka.dev/classroom/physics/fermi-estimation

Nobody can look it up, and nobody is meant to know. Each student breaks the question into a room and a ball, answers as a power of ten and says how sure they are. The projector shows the room's spread, overall and by confidence, before a worked example checks the reasoning.

## What it teaches

Students break an unknowable quantity into estimates they can defend, answer on a powers-of-ten scale, and compare how sure people said they were with how closely their answers agreed. They learn that agreement is not accuracy.

## A Fermi estimation activity your whole room answers

A Fermi problem asks for a number nobody can look up and nobody is expected to know, and it rewards getting the size right over getting it exact. Students break the question into pieces they can estimate, here the volume of the room and of one ball with an allowance for gaps, and answer as a power of ten.

Each student also says how sure they are: within a factor of ten, within a factor of a hundred, or honestly no idea. The projector shows the spread of the room's estimates, its extremes, quartiles and middle, for the whole room and for each confidence answer, with counts. Each of those is an estimate somebody gave, so one person's estimate can appear on the board, but never with a name.

Then comes a worked example: an illustrative 15 × 10 × 5 m room holds roughly 15 million 40 mm balls after allowing for gaps. It is a reference to check reasoning against, not an answer key, and you can rework it for your own room before class.

## Classroom debrief

Which assumption moved your estimate most: the room, the ball, or the gaps?

Did the people who said “within a factor of ten” agree more closely? Were they nearer the worked example?

What would convince you an estimate is right, not just popular?

Redo the worked example for this room. Which way does the answer move?

## Fermi estimation questions, answered

### Why a power of ten?

Because the lesson is about size, not precision. A power of ten lets a student commit to an estimate they can defend, and lets the room compare answers that span many orders of magnitude on one scale.

### Is there a right answer?

There is a worked example, not a key. The size of a real room, a ball and the gaps between balls all vary, so students compare their estimates with a reference they can justify.

### Do students see each other's estimates?

Not beside a name. The room sees the shape of its estimates, overall and by confidence, including the lowest and highest.

### Can I change the example for my room?

Yes. The worked example is part of the lesson's wording, which you can edit before class so that it describes your own room.

### Why ask how sure they are?

So the room can see whether the people who felt surest actually agreed more closely, and whether agreeing made them any closer to the worked example. Agreement alone does not establish accuracy.

### How long does it take?

One estimate each and the discussion; the saved default shows under the headline as "About N min", and every timer is a setting you control before you run it.

## Run the lesson

Present the boards in PowerPoint, Google Slides and Decka. Students scan a QR code or open the session link to join from their phone browsers.
