[R] analyzing results from Tuesday's US elections
mccorm@ck @end|ng |rom mo|b|o@mgh@h@rv@rd@edu
Mon Nov 16 07:14:41 CET 2020
I really like this guy's video as well. (He also has another nice
video critiquing a statistical analysis of vote results from Kent
county, Michigan that was presented by a Massachusetts Senate candidate,
who has some impressive academic credentials. )
And continuing in this same vein of the complexities of statistical
analysis by intelligent people here is a video by Mark Nigrini using
Benfords analysis on Maricopa County vote results.
If you search for Mark Nigrini on Amazon you will see that he has
written a major text on Forensic Analysis, specifically forensic
accounting investigations, that is now in its second edition as well as
an additional two books on analysis with Benford's Law for accounting,
auditing, and fraud detection (He plugs the text in the last part of the
video). All four books have 4-5 star reviews with 2-48 reviewers. From
the tiny amount of reading I have done on Benford's Law, it seems that
Nigirini is a leading figure in the use of Benford's Law. In the video
he shows that voting results for both Trump and Biden from Maricopa
county AZ both agree with Benfords Law. However, he uses the last digit
and not the first. A word of caution before you click on that link: he
uses Excel !
On 11/13/20 9:59 PM, Rolf Turner wrote:
> External Email - Use Caution
> On Thu, 12 Nov 2020 01:23:06 +0100
> Martin Møller Skarbiniks Pedersen <traxplayer using gmail.com> wrote:
>> Please watch this video if you wrongly believe that Benford's law
>> easily can be applied to elections results.
> Just watched this video and found it to be delightfully enlightening
> and entertaining. (Thank you Martin for posting the link.)
> However a question springs to mind: why is it the case that Trump's
> vote counts in Chicago *do* seem to follow Benford's law (at least
> roughly) when, as is apparently to be expected, Biden's don't?
> Has anyone any explanation for this? Any ideas?
> Rolf Turner
More information about the R-help