please empty your brain below

I believe this was how Greece were caught lying about their economy - the data they submitted to the relevant European institution (ECB? Not sure) didn't satisfy this law, rather had a roughly even distribution of first numbers!
Re Greek economic numbers, see

My goodness,DG, you are getting stir crazy.
Have you got patterned wallpaper? Really like to know how many dots/swirls there are.
I thought you *did* have hundreds of old tickets lying around at home. I'm disappointed.
As a former collector of bus tickets I was interested to view the link to the LT collection. But I am disappointed to find there a ticket described as 189d which is clearly for 1/2d. I cannot explain this.
Didn't work for the houses in my street. They all just have names and no numbers.
You could handle house names by turning them into numbers using a=1, b=2 etc then summing the numbers for all letters in the name.
Fascinating stuff. I don't have a collection of bus or train tickets but I do have access to a database of PCs at work. HP use serial numbers starting with a country code (desktops mostly made in CZ, laptops in CN) I've just pulled out the desktops and get the distribution below which is a bit disappointing! I don't know if HP use specific digits to indicate something about manufacture. (I do know that they sometimes reuse serial numbers; doing warranty checks is often a pain because you get "more than one device has that serial number"!)

0 3086
1 2280
2 3160
3 3058
4 2482
5 4030
6 4673
7 4152
8 5773
9 3593

dg writes: You need to remove leading zeroes to give this a chance of working.
Super interesting read.

"I found an old copy of the East London phone book" - love it !
Has anyone tested the Governments daily death toll numbers yet?
With reference to house numbers, back in the 70's we lived on a new estate in Northampton with house numbers that went into the high 700's. The estate had a spine road with numerous cul-de-sacs off each side. All the houses on one side of the spine had even numbers and all the houses on the other side had odd numbers. Thus in our cul-de-sac, we lived next door to 703 & 707 with 683 & 681 opposite.
We would get strangers looking for even numberered houses which were about half a mile away.
Apparently it saved having to think up lots of names for the individual cul-de-sacs!
You do have the ability to explain a complex statistical subject logically and without maths speak, and not for the first time.

With time on your hands, I think you should write 'Geezers Introductory Guide to Statistics'
"Far more numbers start with the digit '1' than the digit '9'."
Good grabby headline which made me think, 'hang on a minute!' But as you go on to explain, the distribution of starting digits across all numbers evens out (doesn't it?!), but not across all the values in a typical data set.
As your introduction suggests, Benford’s law is a nice example of Stigler’s law of eponymy.
The numbers in my street run from 4 to 14 on one side, and 7 to 19 on the other. Given that it was built at the same time as, or later than, the surrounding streets, and wasn't bombed, this is a mystery I have yet to solve.
"Pick a set of data", you say? Hmmm...


The station usage passengers stats for all 2563+ stations from the ORR for the last period. I've just been through 'em, only took ten minutes. Breaks down as follows

1) 771 - 30.12%
2) 441 - 17.23%
3) 339 - 13.24%
4) 253 - 9.88%
5) 184 - 7.19%
6) 170 - 6.64%
7) 151 - 5.90%
8) 114 - 4.45%
9) 118 - 4.61%

It was going perfectly right until the end! There are slightly more '9' stations than there are 8's! Heh.

Although, sorry ... yes, it should of course be noted that 30.1% is spot on, as is 4.6%. w00t!
The numbers in my road are based on the distance from the town hall. Thus house number 849 is 849 metres from the town hall.
Very interesting DG :) I now know what this is called as I see it every month when I do a mailshot to my customers...
When did East London have its own phone book?
I now feel very happy every house I have ever lived in has started with a 1.
Also i would disagree with the opening statement! it should be...

"Far more numbers that are in everyday use start with the digit 1 than the digit 9".
Charles... Looking on the relevant page in the BT Shop, London East still has a separate book, as does London South East, London Central, London West, London South West and London North.
Look Around You...
Good stuff. Perhaps next an investigation of whether a dataset's deviations from Benford's Law are statistically significant. A Kolmogorov–Smirnov test, perhaps.
Fascinating - numbers have always had a meaning or a memory for me of things not necessarily numerical. Such as old BOAC flight numbers - seeing those digits unexpectedly in an irrelevant context sends my thoughts to where those planes flew (eg '591' means to me London - New York - Los Angeles - Honolulu - Nadi - Sydney - Melbourne, flown in later years by glorious VC10s); same for London's bus route numbers, for where they go now or once went. Loved this; to the roving and curious mind this is fantastic food.

More please, at the outsider's level of course!
Nice. Happy Face emoji
This would be fascinating in any case (I've never heard of this) but kudos on doing the examples.
In binary 100% of numbers over 1 start with a 1 - Jimmy’s Law.
Interesting stuff. As others have said, the most surprising part of this is that you didn't have a sufficient stash of tickets at home to create a dataset example.
Barry S -- thank you. Things have changed since I left London in 1997. Four books for the whole Greater London area, A-D, E-K, L-R, S-Z. A new one left free on your doorstep every year or so.

TridentScan | Privacy Policy