Exploring Benford's Law

[TheRandomSkillz]

Someone should do this with like 5,000 random youtube videos.

[ragragdoll]

Nature blows my mind.

[susantam25]

amazing

[davekirkwood]

If you pick a random number between 0-9, the chance of the number starting with a 1 is 10%. A random number between 0-19 has a 50% chance of starting with a 1. 0-29 is a 33% chance.

So, can this law be explained because natural things are constrained by a random upper-limit? If the upper limit is not 9, or 99, or 999 etc. then there will be more chance of the number starting with a 1.

[lukeseti]

Excellent, next time I defraud thousands of people... I'll make sure I use Benford's Law.

[SpicyHam]

pfft, Benford is a fool.

[SolidArch]

I found this on stumble

[scpeer]

It is more than just the prevelance of leading-digits in counting sets. Much natural growth happens expedentially, where totals spend the least amount of time in the leading-digit "9" position.

[lucb1e]

5/5 really interresting!

[Rodrush62]

Interesting video, the narrator also had a great voice

[luckyw4ss4bi]

He sounded hot

[analyst2004]

why u say that

[luckyw4ss4bi]

Just listen to his smooth and sexy voice

[bmwalshe]

wow, that was REALLY neat

[RagingGeekazoid]

Why don't you explain the reason for the law? It's simple enough. Just show a logarithmically even distribution of numbers over a full order of magnitude, display the bins corresponding to each leading digit, and graph the size of each bin.

[jestertru]

An account is based on a standard fee or contractual amount could be legal yet not follow Benford's law at all. The data source has to be "sufficiently" random for Benford's to apply at all.

The case mentioned could have been completely innocent (it might well have been real fraud - I don't know all the details), but to say "because the ledger didn't follow Benford's, so there is fraud" is completely insufficient as proof and complete garbage.

Faking Benford dists is trivial also..

[CypherVirus]

This is some cool shit... these kinds of mathematical phenomena are always fascinating. I feel things like this and the numbers pi and e hold some amazing insight to the universe we haven't discovered yet...

[royb673]

doesn't seem that strange to me

the lower numbers obviously occur more in a list of sequences eg.

1

12

123

1234

12345

123456 etc....

[HoboNetWeaver]

This isn't fascinating. Think about it for 20 seconds.

In an infinite data set, it wouldn't hold true.

[foodeater]

like?

[56jmoney]

Lets run an example of the population of the US states... see what comes up. Anyone have the figures?

[mendelbot]

In 20 seconds, I wasn't able to comprehend infinity and I still find it fascinating.

[HoboNetWeaver]

I promise you that if you think long enough about WHY this occurs, you will figure it out. Start by counting 1,2,3, and then think.. what happens when the sequence increases so much that I must PREFIX A NEW DIGIT.. What DIGIT will I prefix?  It's not a 9...

This law works because it is based on numbersets such as (1) stock prices, (2) street addresses, (3) census data, (4) lengths of rivers.. these data sets contain a FINITE number of values ranging from ZERO ON UP... (but not to infinity)

[othertopics]

fascinating

[roflroflomgomg]

lala