In lots of real-life lists of numbers, about 3 out of every 10 start with the digit 1. And fewer than 1 in 20 start with a 9. Numbers have favourites!
Guess First!
Imagine a huge list of real-life numbers: the populations of towns, the lengths of rivers, the prices in a giant supermarket. You look at just the first digit of each number. (For 4,832 that's 4. For 19 that's 1.)
What fraction of the numbers do you think start with a 1?
There are nine possible first digits (1 to 9, since numbers don't start with 0). If they're all equally likely, each gets a share of about 11%. That's what most people guess.
The real answer is about 30%.
Wait, That's Not Fair!
Why would a 1 get picked three times more often than it "should"? Let's play a game with growing money.
You have $1 and it keeps doubling.
$1 → $2 → $4 → $8 → $16 → $32 → $64 → $128 → $256 → $512 → $1,024
Look at the first digits: 1, 2, 4, 8, 1, 3, 6, 1, 2, 5, 1. The digit 1 keeps coming back!
Here's the secret. To get your money to leave a leading 1, you need it to double: from $1 to $2, or $10 to $20. That's a +100% jump. But once you're at $9, you only need $1 more to hit $10, and then you're back to a leading 1. That's just a +11% jump.
To leave a leading 1: needs +100% growth
To leave a leading 2: needs +50%
To leave a leading 5: needs +20%
To leave a leading 9: needs only +11%
Growing numbers spend ages in the "1" zone and zoom past the "9" zone. Since so many real things grow (populations, savings, sales), the 1s pile up!
The Official Scoreboard
This pattern is called Benford's Law. Here's how often each digit shows up as the first digit:
1 → 30.1%
2 → 17.6%
3 → 12.5%
4 → 9.7%
5 → 7.9%
6 → 6.7%
7 → 5.8%
8 → 5.1%
9 → 4.6%
The digits 1 and 2 together make up nearly half of all numbers (47.7%). The poor 8 and 9 together only get 9.7%.
Try It Yourself
We got a computer to check the first 1,000 numbers in the doubling game (1, 2, 4, 8, and so on up to 2 multiplied by itself 999 times, which is a number with hundreds of digits!). Here's what we found:
1 → 30.1%
2 → 17.6%
3 → 12.5%
4 → 9.7%
5 → 7.9%
That's almost exactly the scoreboard! Even tripling gives about 30% ones, and so does the famous Fibonacci sequence (1, 1, 2, 3, 5, 8, 13...).
Now try your own experiment: collect 30 numbers from real life (prices on a shop website, populations of towns, the number of views on videos) and tally up the first digits. How many 1s did you get?
When Benford's Law Doesn't Work
It isn't magic. It needs numbers that spread across lots of sizes, like $5 up to $5,000,000. It doesn't work for:
Adults' heights in centimetres. They're all squished into a small range, so nearly all start with 1.
Phone numbers or postcodes. They're just labels, not real amounts.
A neat list that stops at a round number. Take the whole numbers 1 to 1,999: more than half (55.6%) start with 1, but only because the list ran out before the 2,000s.
The Detective Trick
Here's why grown-up detectives love this. When people make up numbers (say, a cheat writing fake accounts), they tend to spread first digits out fairly evenly, because "it feels more random." But real numbers don't do that!
A researcher called Mark Nigrini turned this into a fraud-spotting method, and a famous article by the mathematician Theodore Hill reports that New York investigators used his tests to find fraud at several companies. A mismatch doesn't prove someone is cheating, but it's a great clue that something needs a closer look.
Who Spotted It First?
An astronomer called Simon Newcomb noticed it way back in 1881. He saw that the early pages of the library's big books of number tables (called logarithm tables) were much dirtier than the later pages. People were looking up numbers starting with 1 more often! Nobody paid attention. In 1938 a scientist called Frank Benford tested 20,229 numbers from rivers, baseball statistics and more, and the law ended up with his name.
The OddMaths Takeaway
Our brains think numbers start with each digit equally often. But growing things take a long time to get from 1 to 2, and hardly any time to get from 9 to 10, so 1s win. It's a pattern hiding in plain sight, and it can catch people who fake the numbers.