I noticed yesterday that I was having lots of strangers show up to scream at me whenever I posted any information about election fraud, but they were all low information types just barfing up “fact checks” which was basically whatever the news had just told them, but none of them had the basic knowledge of how fraud works to even sorta discuss any of the actual data. So I got curious and posted the following on facebook:
One quick question, only answer if you have worked in auditing/stats/fraud/investigations/or
other data analysis type fields. In your entire career, have you ever seen a
case that threw up this many flags that DID NOT turn out to be fraud?
Again, flags are not proof. They are merely anomalies which would
cause an auditor to check. Nor am I claiming this is a scientific poll (though
I’d bet I’m still more accurate than Nate Silver!). There is of course a
sampling bias as I know many of these people in meat space (and their resumes
on this topic are killer) but it was also open to the public so anyone could
comment and it got shared a hundred times.
The consensus thus far is overwhelming. No. Not only no but hell
no. There have been a few hedging their bets (but they are still suspicious)
and zero saying that there is nothing wrong (like in every single other thread,
where I get screamed at by Dunning-Krugerands). I’m not claiming this is an
accurate sampling of every professional of this type in America, but it is
pretty telling.
#
Big Six accounting
for four years. #4 global investment bank, running the antifraud unit, also for
4 years.
No.
#
Worked in
immigration fraud for many years and currently an Intel analyst at a different
agency. Just the sheer number of statistical anomalies covered here and
elsewhere raises so many red flags, it’s like all the coaches in the NFL
started frantically tossing their challenge flags. Then there’s the poll worker
shenanigans in Georgia, Michigan and Pennsylvania that tells me the Democratic
city machines in those states were going “Fuuuck, he’s behind! Quick, find more
Biden only ballots!”…..
#
No. 25 yrs investigating
financial fraud and money laundering. Where there is smoke there is fire. When
you have this many unconnected witnesses saying the same/similar things, that
is very strong corroboration. Not evidence of made up stories.
#
I was a CPA for 45
years and never saw anything as suspicious as this election.
#
25 years of
investigating white collar crime, primarily complex DoD contracting fraud.
Something with this many allegations would absolutely deserve a very thorough
preliminary investigation. The Hunter Biden situation would have already gotten
an accepted referral to an AUSA for gj subpoenas for records.
#
18 years insurance
work comp fraud investigator.
This many red
flags. Id be able to retire on the billing
#
Investigation
background.
No. Not a single
time. The flags here show clear cut illegal activity being concealed by
multiple sources. At this point I’d reach out to others I trust/know and begin
a more broad based approach to the investigation including into as many of the
secondary involved parties as possible.
#
Not much but in 3
years of AP/investigations; no.
#
I’m a workers comp
investigator. I follow people committing fraud every single day. In my
experience, when something seems this far off, its because it is.
#
I’ve had lots of
red flags where I couldn’t prove in court either that there was fraud, or I
could prove the fraud but not who committed it.
But there’s
definitely enough to trigger a thorough investigation.
#
Certified Fraud
Examiner since 1992 here. I have never seen such an oversupply of red flags. 3
or 4 might be explicable or coincidence, but dozens all pointing the same way?
This would be too implausible for fiction, let alone a case study.
#
Insurance analysis
here. And the answer is “No”.
#
Investigated a case
at a casino with this many flags, approxamaty 60% of the leads turned up
theft/fraud. There were some that were anomalies but they were able to be
checked out via witness accounts.
#
I’ve assisted in
some fraud investigations doing the digital forensics. These many red flags and
you’d want to line the investigator’s chairs and desks with plastic.
#
In 28 years as a
certified fraud examiner I’ve had many similar cases where I had multiple huge
red flags but I couldn’t prove it in court. Fraud cases are hard to prove.
#
Former Army CID.
Answer is a hard no. Smoke= Fire. This is damn near Napalm.
#
IT Forensics. I’ve
never had to do a full blown fraud investigation that involved more than one
individual at a time.
But if I was
looking at a single computer, and saw this all, I would be telling my bosses
that there’s some problems needing to be addressed.
#
(My credentials:
Fellow of the Institute of Chartered Accountants in England and Wales. Former
senior manager in the audit methodology and audit quality control department of
a big 4 firm. I used to run the audit quality review programme for the whole of
Europe for that firm. I’m also pretty much neutral as far as US politics is
concerned – I’m not American, and I’m no fan of either Trump or Biden.)
In a financial
statements audit situation, I would always tell people that discrepancies are
much more likely to be errors than fraud. You’ll often get excitable young
auditors convinced they have found some devious director-level fraud, when
actually they’ve just found a cock-up by Doreen in the accounts department.
However…
When you’re a
financial statements auditor, you’re mostly looking for material errors –
errors which are singularly large enough to change someone’s perception of the
accounts. But you also bear in mind that smaller errors might add up to make a
material difference. So what auditors do is they keep track of those smaller
differences. This list is sometimes called a “scoresheet”, sometimes an “overs
and unders schedule” and sometimes something else, but they are all the same
thing.
On the scoresheet,
the auditor records the double entry required to correct the difference, and
the effect that the error has had on accounting profit. Typically you would
expect some errors to increase profit and some to decrease profit, because that
is the nature of errors. But if you get a situation where almost all of the
‘errors’ are in the same direction – for example, they almost all increase
profit, and the directors have a clear incentive to overstate profit – then you
start to think it looks suspicious.
As an auditor, you
wouldn’t instantly shout “FRAUD!”, but you would investigate further. You’d
look more carefully at areas of the accounts that previously you had considered
lower risk. You’d maybe think about doing some computer-based auditing on
routine transactions. If you were concerned that perhaps some sales invoices
were fictitious, then something you might do (among many other possible tests)
is see if the distribution of significant initial digits in the invoice numbers
and / or the invoice values fitted the distribution suggested by Benford’s Law.
Yes, _that_ Benford’s Law.
So if I was running
an audit that had this many discrepancies, would I have concluded that this was
because of fraud? No, at least not yet. But I definitely would want to
investigate further. (And if the directors tried to prevent me from doing so,
then I’d get even more suspicious…)
#
Fraud manager,
banking and finance and a stint in credit reporting. I can’t recall a single
case with the sheer number of red flags attached, probably because most
organized bad actors take some pains to hide their tracks. Cases with multiple
reporting parties and multiple data anomalies, always turn out to be fraud. The
only question is how widespread it turns out to be when you start pulling on
the threads (usually turns out to be long running and widespread).
#
No. In fact I would
say that in about 34 years of this work it is my professional opinion that at
this level it is mathematically more likely that our sun blinks out of
existence as a result of every particle in it spontaneously “blinking” into
another state than it is that fraud did not take place on the order of millions
of votes.
#
Data Forensics.
No.
#
Accountant/Small
Fry Auditor/ MS in Fraud Investigations/ working on CPA licensure.
The example cases
in my textbooks weren’t this level of ham fisted. Anyone with the sense that
God gave a billy goat could see those. This is a flashing sign that can be seen
from space.
#
Former military
intelligence professional here (yes, we did investigations without any
evidentiary standards to get in the way). I’m insulted at how sloppy with was
carried out.
Information
operations depend on not being noticed. They have to be part of what looks like
the background noise to nudge people toward a conclusion. If the hand is seen,
the magic is lost and the mark spots the trick.
#
Nope – bank auditor
for 30 years – damn I am getting old
#
Fucking newb IT
security chiming in with my .02 Cents (depreciated for lack of experience) IF
I’M seeing shit thats fucky the experts must be having a field day.
#
I’ve been in
infosec for over 10 years now (at least I’ve had my CISSP that long). It’s just
standard practice to assume that any hole in information security will be
exploited eventually. I can think of a number of ways to game the election
system in the US without having to think hard. Or at all, really. Seems like
there’s solid evidence to suggest each technique I’ve come up with was used
somewhere.
#
Larry, about ten
years of Engineering Failure Analysis/Materials, did vendor
investigations/stats. It smells bad…
#
No. Patterns,
Patterns. forensic analysis in cases where men try to mimic natural causes
always reveals improbable elements. Even showing a pattern of trying to avoid
patterns. But this is far more obvious than any case I’ve examined in the last
12 years of expert witness/ consulting work.
#
Insurance Fraud
Special Investigator for 10 years (personal and commercial lines)… One possibly
two red flags could be attributed to stupidity, mistakes, etc. This many,
screams rampant fraud and intenent to commit a fraud. The statistical
improbabilities are also a great indicator that something is wrong. The fact
that the media completely dismisses any of these claims without any interest or
investigation is sure curious. It should not matter what party you are for. If
the there is an indication of wrongdoing shouldn’t everyone be outraged when the
possibility of affecting the outcome of an election is at hand?
#
This many flags all
in one direction? No. But I’ve seen places where it was just that screwed up
that we THOUGHT they were stealing but once we dug in they were just that
terrible. Ended up just closing the whole facility. It was easier rather than
trying to fix all the incompetence.
#
Last case I
investigated, there were only small red flags that were actually not things
that jumped out. I knew at once something was very wrong. There is also gut
feelings, which seem to be right more than wrong. Yeah, I getcha!
#
I teach financial
accounting not fraud detection. That said, I still cover some of the basics in
my classes. One thing that is not discussed much is the idea of controls. In
any business, there are procedures (controls) to limit and detect fraud. In our
elections, there are almost no controls. And, to directly answer your
questions, no there are no textbook examples that with this much smoke without
fire.
#
I used to run the
front end for a RE investing team. You’d be amazed how much fraud there is
therein.
We busted an entire
drug smuggling operation – multiple players, multimillion$/year… on maybe 2% of
the evidence in this case.
#
No and all of my
very liberal friends who have done the grind is statistical analysis with me
are silent…because they know this is all BS.
#
I’m a criminal atty
and I’d tell my client to pray for a good plea bargain offer
#
Not accounting or
fraud investigation, but significant amounts of research statistics. Not
exactly the same, but once the p-values get out beyond the 0.0000000x range,
it’s easy enough to point at something and say that fuckery has been happening.
#
Ex-military intel,
prosecutor and trial attorney. I’m gonna be a bit contrarian, but just a bit.
What we have right
now is a lot of data and data analysis. Certainly the data analysis is *a lot*
of smoke. But I don’t feel good about there being a fire yet because, right
now, the fishy data looks like magic. I want somebody to flip. I want somebody
to indict. Where are those people? With this much information out in public,
how is it that someone hasn’t lost his nerve and confessed his involvement?
So far, there is no
explanation for *how* the fraud was accomplished, which is very useful to know
when you’re trying to figure out who accomplished it.
So…my investigatory
hackles are up, all right, but I won’t really feel confident until we have some
information on *how* the large scale, result-flipping work must have been done.
#
Former Air Force
OSI agent, practicing attorney. I have more respect for people who still
believe in Santa Claus and the Easter Bunny than those who claim to believe
there wasn’t massive amounts of vote fraud in the battleground states where
Trump’s substantial leads suddenly evaporated after counting was paused.
#
Former auditor here
– I actually can think of one time, but only because we didn’t/couldn’t find
evidence the errors were intentional so it couldn’t be called fraud.
And it might not
have had nearly this many red flags – it was 20 years ago when I was just
starting my career
#
PhD in finance and
major stat nerd here. And I do this stuff both professionally and for fun (nerd
– Duh).
Here’s some stats
goodness. When you have a binomial distribution (i.e. two outcomes, like vote
for Biden or for Trump), it’s pretty easy to calculate probabilities of unusual
outcomes.
Let ‘b” be the
“true” probability of Biden getting a given vote (i.e. his normal or expected
share of the vote), and ‘t’ be Trump’s probability. The standard error (think
of it as the standard deviation of the deviation from the expected) is Square
Root of “the sample size * b * t”.
Here’s an example:
Take a sample of 1,000 voters and assume (let’s be generous) that Biden’s True
percentage of the votes is 60%. That means you expect to see 600 Biden votes
out of 1000, with a standard error of Square Root (1000 x 0.6 x 0.4) = 15.5
votes, or 1.55%.
So Biden getting
70% of the votes (a 10% deviation) on a sample of that size is something
something like 6 standard deviations away from the expected 60%. And as the
sample size increases, even smaller percentage deviations are highly unlikely.
When you look up
the probability of a 6 standard deviation event occurring by chance in a states
table, all the table says is “Here There Be Feckery” or, in probabilities,
about 1 in 500 million.
Or, as Sir Pterry
once said, “pull the other one – it’s got bells on it”.
#
Lawyer, and this
election has taught me that I’ve clearly been underestimating the genius of the
op party who used white out and a photocopier to falsify employee records for a
compliance audit and the other op party that embezzled money by writing checks
directly to their personal account.
#
I spent 17 years as
a private investigator, and part of that time I spent investigating and
building fraud cases on behalf insurance companies to submit to local law
enforcement for prosecution. Just what is known in the media on voter fraud is
more probable cause than I had in any of my cases which resulted in a
conviction.
#
I used to work in
welfare fraud for Fresno county and damn this is all very fishy I’ve also been
a voting center captain which means I was in charge of a designated voting area
and at the end of the night had to take the locked up ballots to the designated
counting center. There are so many chances for fraud to happen it’s ludicrous
to think it couldn’t happen
#
Investigations, and
still in the game. You have one of my coins actually.
So once upon a
time, we were looking at a guy that was throwing out flags, smoke signals, and
hand gestures. We kept poking and poking at him, and couldn’t find what we were
looking for. Finally reached out to some colleagues, turns out guy was deep
into shenanigans, but on a different playing field than we were on*. That’s
about as close as I’ve come to the question you asked.
*we ended up
talking to the guy, and he ended up giving us a ton of information that was
later used to bury him. He had no problem talking to us, felt safe because we
weren’t directly asking about the stuff he was involved in. Apparently unaware
that professionals often talk to each other across fields.
#
I have worked in
corporate investigations for over 10 years, and only once when a server holding
data got corrupt, showed 100,000,000,000,000 dollar loss for a store that
earned no more then 15 million a year. Other then that no. Especially if you
make a spreadsheet of pa voters. Only reveiwing 1/3 of mailed in votes, and
look up same day mailed out, and received that same same day for over 8,000
votes, and 8,000 votes that were received before being mailed out. And about
20,000 more votes that were received after November 4th.
#
I’ve worked in
insurance claims for 25 years. Over that entire time, I have been involved in
exactly 1 group of related claims involving an organized fraud ring with NICB
and local law enforcement involved as well. That type of fraud exists, but is
relatively rare. The bigger issue in my industry is opportunistic fraud.
Someone has an accident and sees it as a way to make a little extra by
exaggerating things. Even that level of “soft” fraud is relatively rare
compared to legitimate claims. Now, the election results…they smell like
carefully orchestrated and organized fraud or a large concentration of
opportunistic fraud situations to me. Everything has become so partisan now
that it does not take that much of a stretch to see poll workers in
battleground states putting their thumbs on the scales to favor their preferred
candidate.
#
15 years with
insurance and safety compliance. Some fraud. Lots of stats and analysis for a
large multinational corporation with hundreds of locations. Used to catch
people fairly often trying to cheat at injury statistics or workers comp
claims.
Mistakes are
generally random, or, they are caused by one error and replicated in a
predictable way. Creating a consistent pattern that you can track, identify, trace,and
fix, usually instantly.
Multiple
inconsistencies all making a number bigger only in one column, that’s usually
someone doing something who can’t hide it too well.
You can get away
with it once. Or maybe, a little, once in a while. But I met a few fools who
thought they figured out how to ice skate uphill. Like they were the smartest
people ever.
It looked about as
blatant and stupid as this.
When it gets this
blatant it is usually “systemic” .
Real errors go both
ways. I had to call a lot of people because they were being too hard on
themselves and over reporting; ignorance & incompetence creates errors both
ways.
When there were
lots of “errors” that resulted in better stats in one location across all work
cells… thats a problem in mgmt at the top of the location. Usually someone at
the top of their little food chain was sending the message/ motivating their
people to try to show initiative and hide things. But you can’t hide that very
long. It shows up at the corporate level when things don’t jive.
What I am getting
at is this variety of errors going all one way means it is systematic across
the entire organization. Different errors all going one way means that it isn’t
one state, one software company, one voting method… this is everyone in the organization
getting the message to move the stats one way. And they did it sloppy and
across the board because although the message was sent and received, it wasn’t
*organized* from the top. It was handled from the local level. It was
impossible to be slick and smart, the front line knew what the top level wanted
as a result and no one knew how much it would take so it became super obvious…
The late night
miracles? That was the front line folks scrambling to make magic happen when
the usual methods weren’t sufficient.
Tl:dr its too many
error types across too many locations with all the errors going one way to be
unintended. It is way too widespread to be the usual background fraud that
always happens. It was done too many ways on too many places to be locally driven.
But it was so sloppily done that it wasn’t coordinated as a policy, but more as
a goal because it relied on local initiative to implement.
#
I do a lot of
statistical analysis and groupings at work. If i was given this data as a
sample set. I’d throw it out and require the submitter to correct their data.
#
I need more
specifics and details. But mail in ballots have a great potential of fraud. And
given the emotional and media disgust for this great President, large volume of
fraud would not surprise me.
#
No. When we found
this many red flags, someone went away in handcuffs courtesy of the RI State
Police. Including a Finance Director for a town.
#
And the thread is still going on. As of yet
nobody has showed up to scream about how we’re all big stupid moron idiots who
should shut up and listen to our betters in the totally unbiased news media,
like in every other post I make about the election (It’s almost like the left
has made social pressure compliance shaming into an art form or something). And
nobody claiming to have an audit/fraud/investigation background has chimed in
to gaslight everyone yet, probably because they realize they would get eaten.
