## Democratic Primaries: Election Fraud Probability Analysis

05 May

Democratic Primaries: Election Fraud Probability Analysis

Richard Charnin
May 5, 2016

This is an analysis of estimated probabilities of fraud in the Democratic primaries.

The probability is a function of the 2-party exit poll share, 2-party recorded vote share and the margin of error  (MoE). An exit poll cluster effect (30%) is added to the MoE formula:

MoE =1.3*1.96*sqrt (EP*(1-EP)/N),
where EP is the 2-party exit poll share, N is the number of respondents.

The probability P that Sanders would do better in the exit poll than the recorded vote in AT LEAST 24 of 26 primaries  is P= 1 in  190,000 = 1-BINOMDIST(23,26,0.5,true)

There is a 95% probability that the exit poll will fall within the MoE. The probability P that AT LEAST 11 of 26 would exceed the MoE is calculated using the Binomial distribution.

P = 1 in 76.8 Billion = 1-BINOMDIST(10,26,.025,true)

The probability of fraud is calculated using the Normal distribution. The probability  is based on the difference (DIFF)  between the exit poll share (EP) and the recorded share (RS) less the MoE. If DIFF is equal to the MoE, the probability is 97.5%. The average probability of fraud for all primaries is 97.4%.

Prob = NORMDIST (EP, RS, MoE/1.96,true)

This is how the exit pollsters forced a match to the IN recorded vote.

 Indiana exit poll Unadjusted 1323 resp Clinton Sanders Men 42% 40% 60% Women 58% 48% 52% Total 100% 44.64% 55.36% Final Adjusted 1323 resp Clinton Sanders Men 41% 43% 57% Women 59% 50% 50% Total 100% 47.13% 52.87%

Note the WI 63.6% exit poll in my calculation differs from the 56.0% exit poll result provided by CNN. A screen shot indicated that Sanders had a 68% share of white voters. His share of whites was reduced to 59% in the adjusted exit poll. In addition, according to the 2014 census, whites comprised 92.4% of registered WI voters The CNN adjusted poll indicated that 83% were whites. I used a conservative 88% in the calculation.

 Race Percent Sanders Clinton White 88.0% 68.0% 32.0% Non-white 12.0% 31.0% 69.0% TOTAL 100.0% 63.6% 36.4%

Sanders 2-party  Recorded Vote, Exit Poll shares. Diff = Exit Poll – Recorded vote

 State MoE Recorded Exit Poll Diff – MoE Prob AL 3.9% 19.8% 25.9% 2.2% 99.9% AR 4.0% 31.0% 33.3% -1.7% 87.3% AZ 3.9% 40.9% 63.0% 18.2% 100.0% CT 3.6% 45.6% 47.2% -2.0% 81.3% FL 3.0% 34.1% 36.0% -1.0% 90.2% GA 3.4% 28.3% 33.8% 2.1% 99.9% IL 3.5% 49.1% 51.2% -1.4% 87.5% IN 3.5% 52.8% 55.4% -0.5% 92.9% MA 3.5% 49.3% 53.3% 0.5% 98.7% MD 4.1% 33.3% 33.4% -4.0% 52.7% MI 3.3% 50.8% 53.2% -0.9% 92.1% MO 4.4% 49.9% 51.9% -2.4% 81.0% MS 3.4% 16.6% 21.3% 1.3% 99.7% NC 3.0% 42.8% 43.7% -2.1% 72.3% NH 2.6% 61.4% 60.4% -3.7% 22.7% NY 3.5% 42.1% 48.0% 2.4% 100.0% OH 3.1% 43.1% 48.1% 1.9% 99.9% OK 4.5% 55.5% 50.9% -9.1% 2.1% PA 3.5% 43.6% 45.1% -2.0% 80.6% SC 3.1% 26.1% 31.3% 2.1% 100.0% TN 4.0% 32.9% 35.5% -1.4% 90.0% TX 3.5% 33.7% 37.9% 0.7% 99.1% VA 3.3% 35.4% 37.4% -1.3% 88.4% VT 2.3% 86.3% 86.5% -2.1% 55.5% WI 3.0% 56.7% 63.6% 3.9% 100.0% Average 3.48% 42.43% 45.89% -0.02% 97.42%

### 23 responses to “Democratic Primaries: Election Fraud Probability Analysis”

1. May 5, 2016 at 10:45 am

Where did your data for exit polls come from?

• May 5, 2016 at 7:52 pm

• May 7, 2016 at 2:03 pm

Can you link that CNN source? I’m not seeing it, and on another page here, it says that unadjusted exit polls are not for public viewing. I would love to spread this message, but hopefully you understand that I must be absolutely sure about everything.

• May 8, 2016 at 12:17 am

2. May 5, 2016 at 11:22 am

Richard, got this update email notice an hour ago and came here to log in to comment and it appears my log-in info, my name and/or my email address, isn’t recognized any more. Not sure why, so I’m using my Twitter account to comment.

As usual, I’ll be tweeting this link far and wide, and thank you for your hard work in making plain what many of us know. I also am glad to see Bev here (I trust it’s the same Bev Harris who runs BlackBoxVoting.org, which I went to in 2008 when I first sensed there was election fraud going on) and just a couple days ago stumbled on Lee Camp’s excellent work in initiating #ExitPollGate, which I’ll be using. I wasn’t seeing much about the US elections on RT so wasn’t tuning them in so much until his show came up in an automatic YouTube playlist.

I was also embroiled in a mini-twitter-storm with Joshua Holland over his besmirching of your work here, not to mention your character and credentials, as I took it personally too for having spread the word about your analyses. I take the position that it’s morally criminal for he and the other self-appointed opinion shapers (like your term “presstitutes”) to react with ridicule any time someone questions official accounts of controversial events and issues (i.e. the JFK assassination, 9/11, election theft in 2000 and 2004, etc) and I stand by that. I don’t condone physical violence, but I think folks that do that need what I term a “cyber-flogging” for undermining truth-seeking and the hard work of research. That may be too strident a position for some, but it makes me positively livid when I think of how easily these criminals hack our destiny by hacking our votes and yet no-one in a position of authority has done anything about it, such as order a re-vote. That’s why I’m asking you what you think it would take to make your work “actionable” (i.e. eliminate Optiscan machines without a paper trail, hold re-votes, etc) and also – what do you think will come of this fraud, other than to hand the nomination to Hillary on a silver platter?

I was indignant with Jane Sanders for not having used the opportunities she had recently to speak up for the thousands of voters who were defrauded in Arizona, let alone the other states. Despite having tweeted to her numerous times with links to your articles and spreadsheet, as well as sending her and other campaign officials emails (and I do understand how hectic it is at the helm of a national presidential campaign, having been one of Sen. Gary Hart’s first dozen volunteers in ’84), she publicly said she hadn’t seen it – even after Tim Robbins’ tweets about it were brought up by Neil Cavuto during her Fox interview. By stating she had “not seen it”, she could then move on, as she did, to casually acknowledge there “seems to have been some voting irregularities”. I’m sorry but that riled me, as I thought she would’ve, and should’ve, at least shone a brief light on the massive chaos produced in the Arizona primary. That alone should have elicited more than such a casual reaction, but perhaps she was too concerned with coming across as “whining” or “complaining”, though I think the issue could have been accelerated in the corporate media had she done so.

In any event, thanks for your contributions to our (fraying and warped) democracy, and I look forward to any comments or answers you can offer.

Don

• May 6, 2016 at 11:14 pm

I am sorry to disappoint, but I am not THE wonderful Bev Harris.

I have been aware of the issues of election fraud due to real evidence (hand counted paper ballots) having been eliminated with e-voting machines and hidden with e-scanning machines. It was automatically a red flag. I knew what removing evidence or hiding evidence meant: it meant election theft. So, I have for over a decade read, donated, and engaged with some of the most honest, decent people on the planet, election integrity activists and reporters. They are truly wonderful because they trust regular citizens with a provable democracy to create a better future for all, over the powerful self servicing leaders who have rigged the system. Election integrity people trust you and me to shape our future. I am so grateful for them.

• May 18, 2016 at 4:39 pm

An update from the real Bev Harris…wonderful work.
http://www.nakedcapitalism.com/2016/05/200pm-water-cooler-5182016.html

Oregon, Kentucky
“Sanders is considering seeking a recount in Kentucky, where Clinton was clinging to a lead of a half percentage point” [Bloomberg]. Probably wise.

My comment:
Bernie RECOUNT Kentucky. And, all the other states, where Unadjusted Exit Polls show Fraud.

And, supporting evidence for the need for Bernie to count our votes, a new investigation by Bev Harris of Black Box Voting:

http://blackboxvoting.org/fraction-magic-1/

Fraction Magic – Part 1: Votes are being counted as fractions instead of as whole numbers

By Bev Harris May 12, 2016
tags COMPANIES, ELECTIONS INDUSTRY, MAKE AND MODEL, RESEARCH

1 – Summary –

This report summarizes the results of our review of the GEMS election management system, which counts approximately 25 percent of all votes in the United States. The results of this study demonstrate that a fractional vote feature is embedded in each GEMS application which can be used to invisibly, yet radically, alter election outcomes by pre-setting desired vote percentages to redistribute votes. This tampering is not visible to election observers, even if they are standing in the room and watching the computer. Use of the decimalized vote feature is unlikely to be detected by auditing or canvass procedures, and can be applied across large jurisdictions in less than 60 seconds.

GEMS vote-counting systems are and have been operated under five trade names: Global Election Systems, Diebold Election Systems, Premier Election Systems, Dominion Voting Systems, and Election Systems & Software, in addition to a number of private regional subcontractors. At the time of this writing, this system is used statewide in Alaska, Connecticut, Georgia, Mississippi, New Hampshire, Utah and Vermont, and for counties in Arizona, California, Colorado, Florida, Illinois, Indiana, Iowa, Kansas, Kentucky, Massachusetts, Michigan, Missouri, Ohio, Pennsylvania, Tennessee, Texas, Virginia, Washington, Wisconsin and Wyoming. It is also used in Canada.
snip

All:
Part 1: Votes are being counted as fractions instead of as whole numbers http://blackboxvoting.org/fraction-magic-1

Part 2: Context, Background, Deeper, Worse
http://blackboxvoting.org/fraction-magic-2

Part 3: Proof of code
http://blackboxvoting.org/fraction-magic-3

Part 4: Presidential race in an entire state switched in four seconds
http://blackboxvoting.org/fraction-magic-4

Part 5: Masters of the Universe
http://blackboxvoting.org/fraction-magic-5

Part 6: Execution capacity – coming –
http://blackboxvoting.org/fraction-magic-6

Part 7: Solutions and Mitigations – coming –
http://blackboxvoting.org/fraction-magic-7

• May 10, 2016 at 12:27 am

I appreciate your efforts and totally agree with your comments. I am just as frustrated with Sanders for not speaking out on the clear evidence of fraud by voter disenfranchisement, corrupted unverifiable voting machines and central tabulators.

What can we do?
– Demand full disclosure from the exit pollsters: they must immediately publish the exit poll and vote counts for ALL precincts polled.
– Demand open source non-proprietary systems to confirm hand-counted ballots posted on the wall at every precincts.
– Flood the media with the overwhelming evidence which proves systemic fraud.

3. May 5, 2016 at 11:52 am

Reblogged this on deinvestiture.

4. May 5, 2016 at 12:38 pm

Do you have this study for the Republicans so we can compare?

• May 5, 2016 at 7:46 pm

No, only for the Dems

5. May 5, 2016 at 1:15 pm

I would think if the exit poll and recorded vote were within the margin of error, the probability of election fraud would be close to zero.

• May 5, 2016 at 7:45 pm

Not true. If the difference between the recorded vote and the exit poll = margin of error, the probability is 97.5%.
If the difference is less than that, the probability is less than 97.5% but still > 50%.

6. May 5, 2016 at 5:25 pm

Richard, you are one of my heros, along with John Brakey, Debbie Lusignan, and others on the forefront of exposing the rampant election fraud in this country. What is happening MUST BE STOPPED…. right now! I have some ideas on how to get this much more exposure and get well beyond the “tipping point” that John Brakey spoke of so eloquently at the end of his second interview with Debbie! Can you please contact me at the email I submit with this post? I want to get in direct contact with you, John, and Debbie. We need to coordinate on this!

I’m posting this to Facebook today. OMG!! This is MUST SEE public testimony in New York on the election fraud that went on. The New York City Board of Elections is certifying the NY vote today, even though these issues have NOT been resolved, including counting the affidavit ballots! AND you have to see the condescending and threatening way citizens are spoken to!!! We HAVE to get more people to see this and understand that the evidence of election fraud is OVERWHELMING at this point, including the fact that voting machines have been rigged in at least 20 states!!! It is simply a fairytale that HRC is ahead…. BERNIE is the true winner so far if you remove the FRAUD that has occurred in this primary!!!

Here’s the public testimony from these brave citizens being threatened left and right with expulsion and arrest.

It’s a video you and all the admirers of your work should definitely see!

7. May 5, 2016 at 6:38 pm

I have a question.

I’m looking over this and the math seems off. Look at MD, there the recorded results are 33.3% and the exit poll is 33.4%, meaning the exit poll was off by only 0.1, but you still gives a 52.7% chance of election fraud. Also, look at OK, where the recorded was 55.5 and the exit poll was 50.9 with an MoE of 4.5. In this case, the difference in results was outside the MoE, but you only gives this case a 2.1% chance of fraud.

I am an engineer, not a mathematician so it is entirely possible I am missing something… but don’t you need to use |diff-MoE| (absolute value) and not (diff-MoE). Additionally, I don’t think you should be using diff-MoE to determine where on the probability curve you fall, but I would have to dig into my old textbooks to remind myself.

• May 5, 2016 at 7:33 pm

You are looking at an old table which I just deleted.
Look at the table starting in row 183.
MD in row 193 and OK in row 2o1.
I will check the absolute value: you may be right

• May 9, 2016 at 9:36 am

@Rick and @Josh— You are correct to think there is something strange about the numbers Charnin has posted. His “probability of fraud” is nothing of the kind; that labeling seems to indicate a profound confusion on his part. As he does document somewhere above, he is giving the cumulative distribution for the normal (Gaussian) curve (also called the error function). The way that game is played, for a Gaussian of given mean and width, you name a value, and the CDF is the integral of the distribution up to that value, i.e. the probability that a random number drawn from that Gaussian will be smaller than the number you name. So if for example you name a number say 2 sigma (that being the standard “margin of error” in the polling game) above the mean, then the CDF says you have a 97.5% chance that a new random number will be less than the number you named. Or if you name a number equal to the mean, the the odds are exactly 50% that a new random number will be less. Or in the case of Oklahoma if you name a number about 2 sigma below the mean, the odds that a new drawing will yield a number even smaller will be about (100-97.5)%. But again, none of this is a measure of “fraud”. It is absurd, for example, to believe that when the exit polls and recorded vote match perfectly you have somehow demonstrated 50/50 odds of fraud. There are more sophisticated statistical tests one can make to decide if a given distribution really is Gaussian, but I wouldn’t expect Charnin to be familiar with them. All in all I recommend taking what Richard writes to be an invitation for others to explore on their own. I’ve found his spreadsheets to be mostly reliable, but one should always double check against the primary sources.
BTW let me be clear: I completely believe that there are thumbs on the scale here;
I simply point out that Richard’s overly naive analysis is nowhere near a “proof”.
By all means let’s raise a ruckus and see if someone will release the full exit poll data we need to do a more sophisticated analysis.

• May 9, 2016 at 5:20 pm

Greg,

As usual, you attempt to refute my analysis by misinformation just as you have done with Cumulative vote shares.
I claimed to have ESTIMATED the probability of fraud based on the exit poll discrepancies.
If you have a better calculation, let’s see it.

If you want to believe that the exit poll discrepancy beyond the MoE is due to bad polling or just coincidental, that is your choice.
I believe that an exit poll deviation at the tail of the normal distribution represents a 97.5% probability of fraud.

If you have a different probability calculation, let’s see it.

8. May 11, 2016 at 6:28 pm

So to summarize, I pointed out above that your “fraud probability” function is completely bogus, and instead of saying “Oops, thanks, I’ll fix it” your response is “Hey my broken function gives me what I expect at one limit”. Kind of like saying “what do mean my clock is broken, it is right twice a day!”

But to respond to your invitation to provide a more cogent analysis, I point you to the paper at http://electionmathematics.org/ucvAnalysis/US/Exit_Polls_2004_Mitofsky-Edison.pdf
something I expect you have read before. It is the paper which woke me up 10 years ago.
The figure on their page 6 give a histogram of exit poll disparities. My version of the same plot for the current primary data is here: https://www.dropbox.com/s/0je7o1tpc0r6i7i/ExitPoll.png?dl=0
Like them I can report that a Shapiro-Wilk test strongly rejects (p=.00015) the hypothesis that the data are normally distributed, and for what it is worth, after culling two outliers (AZ and WI in this case), the remaining 23 data points do pass the Shapiro-Wilk test (p=.17) but only with a shifted mean.
Here is the simplest analysis I would offer people: forget the margins of error etc. and just look at the sign of the discrepancy as a coin flip. Apparently Bernie won no more than 2 of 25 coin flips, something which should happen only 1 time in 100000 trials. Perhaps someone will be able to ask the pollsters to explain this apparent systematic bias.

9. May 11, 2016 at 6:37 pm

One attempt to refute your suggestion of fraud involves exit polls not counting early voting. I want to know what you think of my analysis (http://marionumber1.blogspot.com/2016/05/could-early-voting-explain-exit-poll.html), which attempts to show that the early vote can’t always explain the discrepancies.

10. June 22, 2016 at 7:30 am

You’re a good American. Can you do a piece on experimental patches that change totals in voting machines? Or the use of Social Psychologists by political parties to form people’s opinions for them subconsciously? Thanks

• June 22, 2016 at 9:05 am

Not my area of expertise.