What is a Galton board?
A Galton board (also called a bean machine or quincunx) is a triangle of pegs above a row of bins. A ball dropped in at the top hits one peg in every row and bounces either left or right. Sir Francis Galton built one in the 1870s to show something surprising: although no one can predict where a single ball will go, the pile made by many balls has almost the same bell shape every time.
How to use the simulator
- Choose the number of rows (4 to 20) and the chance of bouncing right, p. A fair board has p = 0.5.
- Drop 1 or 10 balls to watch them fall, or 100 to 10,000 to fill the bins instantly. Step moves one ball a row at a time.
- Compare the blue bars (what happened) with the orange marks (the binomial prediction) and the green dashed curve (the normal approximation).
- Tick Path counts on pegs to replace each peg by the number of different routes that reach it.
Counting paths: Pascal's triangle
A ball's journey is a list of choices such as L R R L. With n rows there are 2ⁿ possible lists. The bin a ball ends in depends only on how many of those choices were “right”: k rights puts it in bin k (counting from 0 on the left).
How many routes lead to a particular peg? A ball can only arrive from the peg above-left or the peg above-right, so the number of routes is the sum of the two numbers above it. That is exactly the rule that builds Pascal's triangle, and its entries are the binomial coefficients
C(n, k) = n! ÷ (k! · (n − k)!) = the number of ways to choose which k of the n bounces go right.
You can compute them with the combination calculator. For 10 rows the counts are 1, 10, 45, 120, 210, 252, 210, 120, 45, 10, 1 — there are 252 ways to reach the middle bin but only one way to reach each end, which is why the middle fills up.
The binomial distribution
Each particular path with k rights has probability pᵏ·qⁿ⁻ᵏ, where q = 1 − p. Multiplying by the number of such paths gives the probability of landing in bin k:
P(k) = C(n, k) · pᵏ · qⁿ⁻ᵏ
Example: n = 10, p = 0.5, k = 5: P(5) = 252 × (1/2)¹⁰ = 252/1024 ≈ 0.2461. About a quarter of all balls land in the middle bin.
This is the binomial distribution. Its mean is np and its standard deviation is √(npq). For the fair 10-row board that is a mean of 5 and a standard deviation of √2.5 ≈ 1.58. The readout shows these next to the mean and (population) standard deviation of the balls you actually dropped.
Why a bell curve? The central limit theorem
A ball's final position is the sum of n small independent random steps. The central limit theorem says that sums of many independent random quantities are approximately normally distributed, whatever the individual steps look like. So as the number of rows grows, the binomial bars hug the normal curve with the same mean np and standard deviation √(npq) ever more closely. This special case — the binomial tending to the normal — was found by de Moivre in 1733, long before the general theorem.
The approximation is best when p is near 0.5 and n is large. With few rows, or with p close to 0 or 1, the distribution is lopsided and the symmetric normal curve fits poorly; a common rule of thumb asks for np and nq both at least 5.
There are two different kinds of “getting closer” here. More balls make the blue bars match the orange binomial marks (the law of large numbers — see the probability simulator). More rows make the orange marks match the green curve (the central limit theorem).
Experiments to try
- Reset and drop 10 balls. Is the pile bell-shaped? Add 100, then 1,000, then 10,000. When do the bars settle onto the orange marks?
- Turn on path counts with 6 rows. Check that each number is the sum of the two above it, and that the bottom row adds up to 2⁶ = 64.
- Set p = 0.8. Predict the mean bin for 10 rows before dropping any balls. Where does the normal curve fit badly?
- With 4 rows and p = 0.5, work out all five bin probabilities by hand (1/16, 4/16, …) and compare with 10,000 balls.
- Compare 5 rows with 20 rows at p = 0.3. For which does the green curve match the orange marks better, and why?
Frequently asked questions
What does a Galton board demonstrate?
It shows that many small independent random events add up to a predictable pattern. Each ball bounces left or right at random, but the bins fill according to the binomial distribution, which looks like a normal bell curve when there are many rows.
Why do most balls land in the middle of a Galton board?
Because far more paths lead to the middle. To reach an end bin a ball must bounce the same way every time, which can happen in only one way, while there are many different left/right sequences with roughly equal numbers of lefts and rights. With 10 rows, 252 of the 1,024 paths end in the centre bin.
How is Pascal’s triangle related to the Galton board?
The number of different paths to each peg equals the sum of the path counts of the two pegs above it, which is the rule that generates Pascal’s triangle. The entry in row n, position k is the binomial coefficient C(n, k), the number of paths to bin k.
What is the formula for the probability of each bin?
For n rows and probability p of bouncing right, the probability of landing in bin k is C(n, k) × p^k × (1 − p)^(n − k). The mean bin is n × p and the standard deviation is the square root of n × p × (1 − p).
What happens if the pegs are biased?
If the chance of bouncing right is not 0.5 the peak shifts towards one side, to about n × p, and the distribution becomes skewed. It is still binomial, and with enough rows it still approaches a normal curve.
How does the Galton board illustrate the central limit theorem?
A ball’s final position is the sum of many independent left-or-right steps. The central limit theorem says such sums are approximately normally distributed, so with more rows the binomial distribution of the bins gets closer and closer to a bell curve.