You have probably been told pi is 3.14, and that it has something to do with circles. But what is it, really? It is just how many times the distance across a circle fits around its edge — a little more than three, always, for every circle there has ever been. Here is the surprising part: you can discover pi without measuring a single circle. Throw darts at random into a square that has a round corner painted in it, count what fraction land on the round part, multiply by four, and out pops pi. Press the buttons in the simulator and rain down darts — watch the number wander at first, then settle right onto 3.14.
Most people think pi is just a number about circles that you memorise as 3.14. In fact it is the ratio every circle forces on us, and it turns up far from circles: you can even conjure it from random darts.
What's actually happening
Almost everyone meets pi as a thing to be memorised — 3.14159, maybe a few digits further if you are showing off. That hides what it actually is. Pi is the answer to a measurement: take any circle at all, a coin or a planet, and see how many times its width fits around its rim. The answer is always the same, a hair past three, and the fact that it is exactly the same for every circle in the universe is the real marvel. It is not a label we assigned. It is a number the shape forces on us.
The strangest thing about pi is how many doors lead to it that have nothing to do with circles. One of the prettiest is pure chance. Picture a square with a quarter-circle drawn inside, its rounded edge sweeping from one corner. Now scatter points completely at random across the square. Because the rounded region takes up exactly π/4 of the square's area, the fraction of points that happen to land inside it will, given enough points, come out near π/4. Multiply that fraction by four and you are holding an estimate of pi, conjured from nothing but randomness. This trick is called the Monte Carlo method, named after the casino, and it was used in earnest on the first nuclear weapons work in the 1940s when the sums were too hard to do any other way.
There is a catch that the simulator makes you feel. Randomness is a slow teacher. A hundred darts might give you 3.0 or 3.3; to pin pi down to 3.14 reliably you need tens of thousands, and to add each further digit you need about a hundred times more darts again. That is why nobody actually computes pi this way in practice — there are far faster formulas. But as a demonstration it is unbeatable, because it shows that pi is not really about circles at all. It is woven so deeply into the mathematics of area, of probability, of waves and bell curves, that it keeps turning up uninvited, the same 3.14159 every time.
Pi is not really about circles but woven into area, probability and chance, which is why scattering random darts in a square reveals the same 3.14159 every time.
- 1Find any round tin or jar. Wrap a piece of string once around its rim, mark where it meets, then straighten the string against a ruler — that length is the circumference.
- 2Now measure straight across the top of the tin, through the middle — that is the diameter. Divide the circumference by the diameter.
- 3You will get about 3.14, no matter which tin you pick. Try three different-sized ones and watch the same number appear every time.
Common questions
A quarter-circle covers exactly π/4 of its bounding square, so the share of uniformly random points landing inside, multiplied by four, estimates pi. This is the Monte Carlo method, used in earnest on the first nuclear weapons work in the 1940s.
Randomness is a slow teacher: accuracy grows only as the square root of the number of darts, so each extra correct digit costs roughly a hundred times more throws. Far faster formulas exist, but the dart trick is unbeatable as a demonstration.
Only about fifteen to navigate spacecraft, even though pi has been computed past one hundred trillion digits. Forty digits would size a circle the width of the visible universe to within an atom.