Galton Board
Chance, curves, and a stranger thing — the 1889 machine that makes the bell curve visible, and the TV lab that made us wonder whether it could be beaten.
Clip 1 — Stranger Things: Eleven and the Peg Board
Rotate your device for the best video experience.
Clip 2 — A Real Galton Board in Action
Rotate your device for the best video experience.
Sir Francis Galton and His Bean Machine — A Brief History
Sir Francis Galton (1822–1911) was a Victorian polymath — a half-cousin of Charles Darwin who contributed to meteorology, geography, psychology, statistics, and fingerprint science. Among his many contributions to mathematics, one stands out for its elegance: in 1889 he introduced a physical device for demonstrating probability in a book titled Natural Inheritance. He called it a quincunx (from the Latin for a five-peg arrangement), though his version used many more rows.
The device is simple: balls are dropped from a single point at the top and fall through a triangular array of pegs, bouncing left or right at each one. Because each bounce is independent and equally likely to go either way, the balls accumulate at the bottom in a shape that traces out the binomial distribution — which, as the number of peg rows grows, converges to the famous bell-shaped normal (Gaussian) distribution.
Galton’s device — now widely known as the Galton Board or bean machine — became one of the most celebrated demonstrations in probability and statistics. Physical versions exist in science museums worldwide, including the Smithsonian and the Natural History Museum in London. Its appeal is timeless: it turns an abstract mathematical theorem into something you can watch with your own eyes in real time.
The second clip above shows a modern commercial Galton Board — the kind now sold as a desktop demonstration device. Watch closely: no matter how many times the balls are dropped, the bell curve appears. It is not a trick. It is mathematics.
Stranger Things — When Eleven Met the Peg Board
Stranger Things (Netflix, 2016–) is set in Hawkins, Indiana, in the 1980s. At its center is Eleven, a young girl raised in Hawkins National Laboratory by the enigmatic Dr. Brenner, who has been running experiments on children with psychokinetic abilities for the U.S. government’s MKUltra-inspired program.
Among the most memorable lab scenes are the experiments in which Eleven is asked to influence the outcome of a falling-disk apparatus — a device in which small disks or balls drop through a series of pegs and are expected to fall randomly. The implication is profound: if Eleven can shift the distribution of outcomes, she can demonstrate that her mind can overcome the laws of probability. If the bell curve appears, the laws held. If the bars skew in one direction, something “stranger” is at work.
This scene was the spark that led to the TNT Galton Board Simulator. The question it raised was irresistible: could we build the device that Eleven was trying to defeat? The answer was yes — in p5.js, with a physics engine, a binomial PMF overlay, and a stats panel that would have told Dr. Brenner in real time whether the results were matching theoretical predictions.
The TNT Connection — Building the Simulator
The TNT Galton Board Simulator was built in p5.js following the S.P.A.R.K. methodology, starting from a student’s loading-icon animation (David’s HBO Loading Icons) as a structural template. The simulator features:
- Triangle and Rectangle (quincunx) peg layouts
- Fully adjustable physics: gravity, bounce factor, ball and peg radii, drop rate
- A gold bell curve overlay showing the exact theoretical binomial PMF — not a best-fit curve, but a pure mathematical prediction calculated before any ball drops
- Empirical vs theoretical statistics panel, updated live
- Optional ball–ball physics toggle, demonstrating how independence violations change the distribution
- A snapshot feature for capturing the board mid-run
A companion Bell Curve Tutorial guides Algebra 2 students from coin flips through Pascal’s Triangle, the binomial expansion, the binomial PMF, and the normal distribution equation — all with images captured directly from the simulator. No calculus required.