The case for the occurence of algorithmic vote flipping

[RonRules]

Chinese cookie

Look at the Chinese cookie I got at lunch:

"You will be successful in you work"

 

[drummergirl]

Look at the Chinese cookie I got at lunch:


"You will be successful in you work"

Eating desert first, I see

 

[Champ]

Just thought this would go in this thread..

Fraud Confirmed in Texas by Judge..

From the DailyPaul: http://www.dailypaul.com/239494/fra...-says-votes-were-switched-tonight-drkrbn-live

 

[affa]

Just thought this would go in this thread..

Fraud Confirmed in Texas by Judge..

From the DailyPaul: http://www.dailypaul.com/239494/fra...-says-votes-were-switched-tonight-drkrbn-live

Wow.

 

[RonRules]

She's being interviewed right now. Click on the upper right of the page:

http://ronpaulradio.com/

 

[RonRules]

Since Travis County is such a travesty, I did a couple more charts:

The slow curve at the left occurs when one candidate has 100% (or most) of the votes the few very small precincts. Then the curve takes a while to catch up to the flat line. I'd say both of the above are flat.

 

[wstrucke]

it seems like the only way this could get traction would be to pick a district that looks particularly bad and organize to go door to door and get people to sign affadavits that they voted for Ron Paul. if you can get enough signatures to show the reported counts are WAY off, then you'll have something.

 

[RonRules]

it seems like the only way this could get traction would be to pick a district that looks particularly bad and organize to go door to door and get people to sign affadavits that they voted for Ron Paul. if you can get enough signatures to show the reported counts are WAY off, then you'll have something.

If you like this approach, join Dr. K and others at this website:
http://www.ronpaulvotecount.org/

You can also help them through the pac:
http://www.libertyusapac.org/

If your county uses voting machines like the Sequoia Edge (and others), see if the "Poll Tape" produces a report of each candidate. Take a picture of each end of the tape. Collect that data in a spreadsheet and send it to me (post the link here). With that data I'll make a chart.

If you can get a 1,000 or more votes, I'll have chart with good resolution and strong statistical significance.

 

[RonRules]

Woo Hoo! I can start insulting again:

 

[RonRules]

Thought I'd let you all know that Riverside County will NOT allow me to take pictures of the poll tapes. I already have pictures of 9 tapes, which flat lined.

Today I asked to see the top 50 precincts and take pictures of them. They refused and said that they have to have the county lawyer approve first.

This is getting INTERESTING!

I also had asked for 9 other items: Libertarian, Green party, Riverside District Sup. 1,2,3,4,5, Congressional districts, 41, 42, US Senator.

I only got Libertarian and District 3. I'll put the charts up soon.

 

[bcreps85]

Thought I'd let you all know that Riverside County will NOT allow me to take pictures of the poll tapes. I already have pictures of 9 tapes, which flat lined.

Today I asked to see the top 50 precincts and take pictures of them. They refused and said that they have to have the county lawyer approve first.

This is getting INTERESTING!

I also had asked for 9 other items: Libertarian, Green party, Riverside District Sup. 1,2,3,4,5, Congressional districts, 41, 42, US Senator.

I only got Libertarian and District 3. I'll put the charts up soon.

Not that you couldn't be on to something, but I wouldn't take it as an admission of guilt either. They are probably just getting nervous about all the questions and of course your correspondences that they are probably taking as accusations. Public servants seem to go into "CYA" mode the second someone starts to take interest in their job. I imagine that it is because, unlike in the private sector, they aren't used to anyone taking interest in whether or not they are actually doing their jobs properly or not.

 

[RonRules]

Public servants seem to go into "CYA" mode the second someone starts to take interest in their job.

That's probably the case. Of the 30-40 people in that office I'm confident that nobody there is to blame.

However, I had met with them 3 months BEFORE the elections and communicated frequently with a plan to PREVENT the vote flipping problem. It got completely ignored.

 

[RonRules]

Libertarian Party, Riverside California.

This one is weird and deserves to be explained!

There are very few votes for Libertarians in Riverside, so the chart is going to be very choppy.

Here are the total votes for the entire county for each candidate:

Barbara Joy Waymire 72
Lee Wrights 14
Roger Gary 16
James Ogle 13
Scott Keller 35
Bill Still 35
Gary Johnson 227
RJ Harris 18
Carl Person 10

The slow downward curve for Gary Johnson can be explained because in small precincts, he was likely the only libertarian with any votes. So he got lots of small precincts with 100% of the votes, and this favorable ratio proportionally decreased as the precinct size increased. In the larger precincts, he got the ratio that he "deserved", close to the final ratio.

In addition, if any precinct has less than 10 votes, they are not published and the following statement is issued: "Insufficient Turnout to Protect Voter Privacy". There are a total of 839 precincts and they are separated in "Vote by Mail" "Election day", each of those is subject to the 10 vote minimum limit for each party. Since there are so few Libertarian votes anyway, usually less than 10 per precinct, I believe that I'm missing most Libertarian votes that way.

So that's my excuse for the lousy chart. Considering the limitations and looking at only the last 50%, I say it flat lines.

 

[RonRules]

This is the Riverside County Supervisor, District #3.

I know most of you won't care about this particular race, but I do. I can't stand Jeff Stone.

So you tell me since I'm a little biased, Is Jeff Stone a flipper? (We tend to disregard the first 10-20%).

 

[drummergirl]

Hard to say for sure. You'd need to put in confidence intervals for that one, but it does look suspicious.

This is the Riverside County Supervisor, District #3.

I know most of you won't care about this particular race, but I do. I can't stand Jeff Stone.

So you tell me since I'm a little biased, Is Jeff Stone a flipper? (We tend to disregard the first 10-20%).

 

[drummergirl]

Looks like the sample size is too small for the statistics of large numbers to work effectively.

Libertarian Party, Riverside California.

This one is weird and deserves to be explained!

There are very few votes for Libertarians in Riverside, so the chart is going to be very choppy.

Here are the total votes for the entire county for each candidate:

Barbara Joy Waymire 72
Lee Wrights 14
Roger Gary 16
James Ogle 13
Scott Keller 35
Bill Still 35
Gary Johnson 227
RJ Harris 18
Carl Person 10

The slow downward curve for Gary Johnson can be explained because in small precincts, he was likely the only libertarian with any votes. So he got lots of small precincts with 100% of the votes, and this favorable ratio proportionally decreased as the precinct size increased. In the larger precincts, he got the ratio that he "deserved", close to the final ratio.

In addition, if any precinct has less than 10 votes, they are not published and the following statement is issued: "Insufficient Turnout to Protect Voter Privacy". There are a total of 839 precincts and they are separated in "Vote by Mail" "Election day", each of those is subject to the 10 vote minimum limit for each party. Since there are so few Libertarian votes anyway, usually less than 10 per precinct, I believe that I'm missing most Libertarian votes that way.

So that's my excuse for the lousy chart. Considering the limitations and looking at only the last 50%, I say it flat lines.

 

[KingNothing]

For precincts that you suspect of being "flipped," have you done tests to show if the differences are statistically significant from the county, congressional district, etc mean?

You'd also have to consider how likely it would be to find a naturally occuring difference, given the thousands of precincts investigated.

 

[drummergirl]

For precincts that you suspect of being "flipped," have you done tests to show if the differences are statistically significant from the county, congressional district, etc mean?

A lot of those type of tests were done early on in this project when the charting technique was being developed. The condensed version is in about pages 22-29 of the summary here:

https://docs.google.com/document/d/1EokVx9tDsrjAJ-7H9XoPv3KmZYDvVjSFJ4cuxJTo1iE/edit

You'd also have to consider how likely it would be to find a naturally occuring difference, given the thousands of precincts investigated.

Those odds have been calculated in some states (in great detail in South Carolina, for instance). Generally speaking they come out so far outside the 99% confidence interval that they are quite laughable (or cryable depending on your perspective). There have been several where the odds that it "just happened" were equivalent to winning the Texas state lottery on one quick pick ticket purchased per week for 4 or 5 weeks in a row (sure to draw the attention of the lottery commission on the second week )

The charting technique we've developed is a quick way to look at results and see if there's any hanky panky. Analytically, it's similar to charting widely varying data on a logarithmic scale to look for patterns; they are still there on a regular scale, just difficult to visualize. At this point so many of us have seen so many charts, that we can generally call a flipper from them. Sometimes it's too close to call and you'll need to go back to the more traditional tests to know whether the results are within the 99% confidence interval or not. (like above, one case the slope is slight, another case too few data points)

 

[KingNothing]

I just did a quick search of that document and found no reference to any sort of statistical test to compare means. That needs to be done if it hasn't, and should have been the first thing done.

 

[RonRules]

Does anybody here believe that Romney would legitimately earn more than 7 times more votes than Ron Paul, where his home in Lake Jackson is located? If yes, go read something else.

For those interested in doing something about it, please read on and help investigate.

I finally got the data for Brazoria County, TX, the home of Lake Jackon:

Considering that Ron has gotten as high as 80% of the vote in his prior congressional elections, would you believe that Romney beat him in on his own turf by more than 700%:

Here's the cumulative chart and it definitely looks suspicious:

When a chart curves up like that, something is going on and it's important to look further. When the line slopes up straight and steadily, I believe that's the work of a nefarious central tabulator and a coded software algorithm. Such perfect lines don't occur naturally.

When the line curves up, like in the above, it's manual vote flipping, but only in a few precincts. Those few high precincts have to be caught up by the normal legitimate data and it produces a curve. (Remember, these charts are cumulative.)

Here's how you can visualize this better. I made a chart of the data by comparing only Romney and Ron Paul and adding a third line (blue) to show the ratio between Romney and Ron Paul. Note the incredible peaks in Romney's "performance". You won't see that in flat-lining data.

Note that in precinct #27 shows Romney beating Ron by a whooping 719%:

I suggest investigating the following precincts: #12, #27, #46, #47, #51, #61. Ask everything you can: Poll Tapes, Ballot Chain of Custody, Names of election workers at that location, Counting Process (Local Count or Central Count), #of signatures at the precinct compared to number of votes (it may match anyway because of flipping), Reports of broken seals, "corrected" vote tallies, etc.

Do you guys care about Ron? Do you care about this Revolution and Liberty? Please help investigate this. (I'm a bit busy in my own Riverside county)

If you have the ear of the campaign, show them these charts and see if they're willing to help you investigate.

The source of my data is right off the Brazoria website:
http://brazoriacountyclerk.net/recorder/assets/repprimarycanvass-20120529(1).pdf