ViewWiki

Lesson 4 of 4

The bell curve.

About 25 minutes · 5 practice questions

A Galton board is many coin tosses in a row. How many paths there are, why most balls land in the middle, what moves when the pegs are not fair, and why the average result is not the result you usually get.

Step 1 of 5 · Every peg is a coin toss

Step 1 of 5

Every peg is a coin toss

Balls fall from the hopper through a triangle of pegs. At every peg a ball bounces left or right, half and half, like a coin toss. At the bottom it lands in a bin, and the bin is set by how many times the ball bounced right: the leftmost bin is no rights at all, the rightmost is right at every row.

Here there are 44⁠ rows of pegs, and so 55⁠ bins. Every ball is four coin tosses in a row, just like the tree in the last Lesson.

To the question

Step 2 of 5

The middle bin

Only one path leads to each bin at the edge: right at every row, or left at every row. Many paths lead to the middle bin, two rights and two lefts, because the order can change: right-right-left-left, right-left-right-left, and so on.

All the paths are equally likely, so a bin’s chance is the number of paths into it out of all the paths. Look inside is set to Bell curve: the orange line shows how many balls to expect in each bin.

To the question

Step 3 of 5

The edges

Now there are 1212⁠ rows and 1313⁠ bins. To reach the rightmost bin, a ball has to bounce right at every one of the 1212⁠ rows. Predict how likely that is.

To the question

Step 4 of 5

Pegs that are not fair

Chance of bouncing right sets the chance at every peg. On average a ball bounces right at that share of the rows. With nn⁠ rows and a chance pp⁠ of bouncing right (as a fraction), the centre of the hump sits at bin

n×pn \times p⁠

This is called the expected value: the number of rights you get on average. With fair pegs and 1212⁠ rows it is 12×0.5=612 \times 0.5 = 6⁠, the middle bin. Change Chance of bouncing right so that the centre of the hump is at bin 99⁠.

To the question

Step 5 of 5

The average is not the rule

Drag Rows of pegs up to 1616⁠ and look: the more rows, the smoother the bell curve the bars trace, a curve called the normal distribution. It turns up wherever many small, independent effects add up: people’s heights, measurement errors, the results of many tosses.

Now there are 1010⁠ rows: ten coin tosses. The average result is five rights, in the middle bin. But how often do you really get exactly 55⁠ out of 1010⁠?

To the question

Interactive scene

Loading the interactive version…

Each step sets up the Scene for you, and a control marked “fixed” stays put for that step.

To the experiment

Practice · Calculate

The board has 44⁠ rows of pegs, and at every peg the ball goes left or right. In how many different paths can a ball travel from the top to a bin?

To the experiment

Practice · Calculate

What is the chance that a ball lands in the middle bin, two rights and two lefts? Write it as a percentage, the number only.

To the experiment

Practice · Predict, then look

With 1212⁠ rows of fair pegs, what is the chance that a ball lands in the rightmost bin?

To the experiment

Practice · Reach the target

Rows of pegs is locked at 1212⁠. Change Chance of bouncing right so that the centre of the hump, the bin expected on average, is bin 99⁠. Press Check.

To the experiment

Practice · Calculate

1010⁠ rows of fair pegs, like 1010⁠ coin tosses. What is the chance that a ball lands exactly in the middle bin, 55⁠ rights out of 1010⁠? Write it as a percentage, the number only, to one decimal place.