A look at Benford’s Law.
In certain numerical data sets, the leading digit is most likely to be a “1,” with each succeeding number being less likely. It can be used to see if someone unaware of this has manufactured data.
swans on tea
Just another SFN Blogs site
Seems like he uses very black and white examples in this demo. Surely there’s a progression somewhere between “fits perfectly” and “non-arbitrary data” where it’s just not as clear cut, right?
The video makes a sort of casual mention that the stock market follows Benford’s Law, but this case actually is much more like Nick suggests above, that is, there are more 1s than 4s, but the agreement isn’t terribly good. For the Dow Jones and the S&P 500, there are many more 8s and 9s than one expects from Benford, while the Nikkei index has roughly 60% starting with 1, instead of the 30% predicted by Benford’s Law. My analysis is here.