(Huge) delegate vote anomaly in Alabama verified

[The Man]

/

How do you group the 20 precincts together? Are they sorted?
The reason I am asking is that when you look at Paul's share of votes vs Santorum's per county, which should cancel out quite a bit of noise, there is no sign of correlation whatsoever. None. The 5% "skimmer" is not apparent here either, as plenty of counties score above 5% and Paul's best score is almost all the way up to 8%. Some more head scratching at my end, I'm afraid.

I used a running average, n=20, precincts arranged ascending totals as usual. You still end up with 1800+ data points but each averaged total is the average of the 20 previous precincts. I call it a running average (not positive that's technically the proper term).

It's interesting that you say Paul scores almost 8% in some counties... but his cumulative % statewide stays at 5% +/- 0.3%. Hey Liberty1789, I think it's healthy that you see Bama from a different viewpoint. I don't have all the answers but there's simply no other logical explanation why Paul loses in almost every precinct other than fraud. Please explain differently about the "no correlation between Paul and Santorum's vote totals between counties." I believe that the manipulation is done at the state tabulation level so that Paul is maintained at 5%. Again, it's the "can't see the forest for the trees" concept.

 

[drummergirl]

These are just my thoughts on the matter at this point, so they are subject to change in the light of more information.

I have good reason to suspect that the "flipper" virus dates back to the late 1980s. Part of the reason it's so crude is it's age. I also suspect we can find the author by following the bodies, but that's a rabbit trail for another day.

The vote "sucker" (I don't know virus, object, function?) mechanism has a more modern and sophisticated feel to it. It would still need to be simple enough to avoid breaking the program to stay under the radar. Perhaps it takes a fixed percentage rather than capping a candidate at a certain amount. Perhaps it uses a random number generator to take a range of percentages (like 3.4 to 5.2 percent or something).

In any case I suspect that the sucker and the flipper were both expecting to utilize the cushion provided by the difference between PPE votes and delegate votes to hide their activities. And like 2 people on a joint checking account both using the last $100 before pay day, they blew past the cushion because they both used it.

 

[Liberty1789]

EVM Data - Jefferson County

Here we go, nice excel file from The Man's massive listing. Haven't analysed anything out of it yet. Best of luck.

http://www.filedropper.com/aljeffersonevmdata

 

[dr.k.research]

About 25-30% didn't vote in any delegate races

You sure know how to waste peoples' time. Whenever anyone ANYWHERE begins to demonstrate anomalies in vote counts, with reliable data, with significant evidence of tampering, you show up and start your babbling. Don't you have anything better to do than attempt to undermine this work? They are even quoting your babble on the DailyPaul, that is when threads about the Alabama delegate anomaly are posted.

 

[dr.k.research]

I tried to answer the question by looking at the data

Why didn't you look at the data before you made that terminally fraudulent post about Iowa? Have you cleaned up that post yet?

 

[The Man]

At first glance, I do NOT see a simple trend where Romney receives votes at a certain count. There does appear to be some correlation between precincts with a high "votes minus delegates" discrepancy AND % difference between 2 EVM's in a precinct, BUT the difference does not seem to favor a single candidate. Just digging in though.

 

[RonRules]

Here we go, nice excel file from The Man's massive listing. Haven't analysed anything out of it yet. Best of luck.

http://www.filedropper.com/aljeffersonevmdata

Thank you so much for converting the big PDF file. I know how much work that is.

Here what I rushed to look at:

Difference with EVM listing: 225

Let me emphasize once again. THERE SHOULD NOT BE A SINGLE VOTE DIFFERENCE between the EVM results and the State Tabulated results. NOT ONE VOTE.

This is not a question of vote interpretation by humans or vote machine scanning errors. This is a question of electronic vote counts incorrectly transferred between two computers. THERE IS NO MARGIN OF ERROR HERE. Those counts MUST match exactly.

There are 67 counties in Alablama. 225 votes in one county could mean as many as 15,075 state wide.

This is a HUGE problem and I thank "The Man" and "Liberty1789" for doing the initial work. We have more analysis to do, but most importantly, we need to follow up with the Secretary of State and the Atty General of Alabama.

 

[Liberty1789]

Difference with EVM listing: 225
Let me emphasize once again. THERE SHOULD NOT BE A SINGLE VOTE DIFFERENCE between the EVM results and the State Tabulated results. NOT ONE VOTE.

RonRules, I am a pretty safe hand with Excel, but the discrepancy error could be mine. It was not an easy conversion. I did not go and verify each of them manually, feeling a bit of speadsheet nausea after so many hours...

 

[Liberty1789]

parocks having the last laugh?

Jefferson's County listing allows to look at votes per machine. Guess what: 18 machines have... 1 vote! So we get to look at 18 single ballots! Fabulous!

-> 14 ballots followed the rule: presidential vote matches delegate race votes
-> 2 Romney voters violated the rules with 1 delegate race overvote, by voting once for Gingrich 1st delegate and once for Santorum 1st delegate
-> 2 Santorum voters violated the rules with massive overvote:

2 out of 18 near-vote-for-all: quite high...

Below are not ballots anymore, but addition of ballots. Simple logic allows to identify rule violations and to speculate on what went wrong. Have a go: it's kind of fun:

From all the machines where I would argue that you can identify a vote-for-all pattern in all likelihood, I derive a voting error rate of 7% (i.e. 12/165). In Jefferson County, Paul scored 4% in the presidential race and more like 10% in the delegate races.

parocks, feel good

The Man, if I may be so bold as to ask you 2 things that I would love to squeeze out of your Jefferson election official contact:

(1) the answer to the following question: was the enforcement of the delegate overvoting rule any different in 2008, through software or poll worker checks?

(2) the Jefferson EVM listing for 2008. I love pdf conversion so much

 

[RonRules]

Those Santorum voters sure like to vote! Why would they miss Paul P17 and a few Romneys? Were their names Devil and Lucifer?

I like how you've laid this out, but it took me about a two minutes to understand the color code. Red is an overvote for delegate and black is an expected delegate vote. The title "PPP Votes, 1 Santorum" will most likely confuse people, and you'll be flooded with questions. May I suggest the first line say: "One Vote for Candidate Santorum", second line: "Delegates picked on that ballot. (Red= incorrect vote)"

The result above (except for the dumb Santorum voter theory) can only be explained with disappearing candidate votes. This is not vote flipping because the total number of votes is still 1.

The Man: Would it be possible to get a copy of the poll tapes for machine #138 and #299? More would be even better (#197, #270, #131, #165 above), and particularly machines with high counts.

 

[The Man]

Hey RonRules, Liberty1789- I have forwarded your request to the official. I will post any kind of response received from him. I am paying attention to your posts. Liberty1789, in order to begin to consider this "idiot voter" theory, there are some very difficulty assumptions that must be made.

 

[The Man]

Hey Liberty 1789, Any chance that I could interest you in analyzing the POSSIBLE correlation between "Romney's votes minus delegates" AND Romney's % difference from EVM1 to EVM2 within a precinct?
Update- I'm working on it already. No need to bother.

 

[drummergirl]

OF COURSE!!!
<facepalm and I coulda had a V8 moment>

Why stuff the ballot box when you can just delete your opponent's votes?

Sorry, having a bit of a brain rush because this explains SO MUCH!

Jefferson's County listing allows to look at votes per machine. Guess what: 18 machines have... 1 vote! So we get to look at 18 single ballots! Fabulous!

-> 14 ballots followed the rule: presidential vote matches delegate race votes
-> 2 Romney voters violated the rules with 1 delegate race overvote, by voting once for Gingrich 1st delegate and once for Santorum 1st delegate
-> 2 Santorum voters violated the rules with massive overvote:

2 out of 18 near-vote-for-all: quite high...

Below are not ballots anymore, but addition of ballots. Simple logic allows to identify rule violations and to speculate on what went wrong. Have a go: it's kind of fun:

From all the machines where I would argue that you can identify a vote-for-all pattern in all likelihood, I derive a voting error rate of 7% (i.e. 12/165). In Jefferson County, Paul scored 4% in the presidential race and more like 10% in the delegate races.

parocks, feel good

The Man, if I may be so bold as to ask you 2 things that I would love to squeeze out of your Jefferson election official contact:

(1) the answer to the following question: was the enforcement of the delegate overvoting rule any different in 2008, through software or poll worker checks?

(2) the Jefferson EVM listing for 2008. I love pdf conversion so much

 

[RonRules]

It's also kinda bizzarre that SO many machines would have only ONE vote total. If this were random, you'd expect about the same number of machines with TWO votes, THREE votes and that vote count per machine increasing like the left tail of a Normal Distribution up to the median #of votes per machine.

May I have a histogram please!

I'm expecting a fat left tail. And I know tails!

Edit: My theory here is that the vote flipping algorithm is set up in such a way as to prevent the count from going negative or even zero. It just stops subtracting when it reaches 1. This is to prevent some rather embarrassing situations like Volusia County, FL, where the vote count dropped to negative 16,022.
http://en.wikipedia.org/wiki/Volusia_error

I also recently discovered that this "alleged" algorithm tops off at a certain count. I found that when you separate out the last top 10-20 percentile of the vote, the results flat line! That was quite a surprise until I realized that the likely reason is so the fraud can go undetected and below the margin of error count.

It is important to DEMAND that ZERO is the only acceptable tolerance of error with electronic transfer of vote counts.

 

[RabbitMan]

Great work guys! That's pretty messed up!

 

[RonRules]

Great work guys! That's pretty messed up!

And girls

 

[RonRules]

Wrong thread!

 

[RonRules]

Wrong again!

 

[RonRules]

I moved them to the Mother thread:
http://www.ronpaulforums.com/showth...occurence-of-algorithmic-vote-flipping/page67

 

[The Man]

Hey RonRules, Liberty1789- I have forwarded your request to the official. I will post any kind of response received from him. I am paying attention to your posts. Liberty1789, in order to begin to consider this "idiot voter" theory, there are some very difficulty assumptions that must be made.

So I sent this email to Jefferson County Alabama Election Commissioner:

"Thank you for your response. I have a request from a couple of analysts that I will relay to you. Any response to these is greatly appreciated:
(1) the answer to the following question: was the enforcement of the delegate overvoting rule any different in 2008, through software or poll worker checks?
(2) the Jefferson EVM listing .PDF for 2008 like the one to which you sent me the link.
(3) Would it be possible to get a copy of the poll tapes for machine #138 and #299? More would be even better (#197, #270, #131, #165 above), and particularly machines with high counts.
Sincerely,"

Here is the response received from him just now:

1) We changed no procedures from 2008 to 2012 regarding delegates. Realize though that in 2008, the presidential race was the only thing on the ballot. The presidential primary was held in February, and the statewide primary was held in June.
2) ftp://ftp.jeffcointouch.com/elections/SummaryReport.html
3) No. However, they would match the reports. We post those tapes here at the courthouse as well as at each poll location so anyone can view them and compare our published numbers to the actual tapes. Many of the candidates will have people assigned to various polls to obtain those numbers before the poll workers can drive downtown and would quickly notice a disparity. Also, the reports are compared to the tapes through a canvassing procedure after each election. For a primary election, the party is responsible for conducting the canvass and we facilitate it. For a general election, we conduct the canvass ourselves.