Flipping the vote against Ron Paul in South Carolina?

[DEGuy]

One more thing. Lets assume that this spreadsheet is correct. What about an exit poll? Exit poll shows almost the same results as official counts.

Exactly. They would have to be rigging the exit polls too. Sorry, but the burden of proof is high for a reason. I would like to see the same analysis with the point data and not just smoothed out lines.

 

[bcreps85]

Exactly. They would have to be rigging the exit polls too. Sorry, but the burden of proof is high for a reason. I would like to see the same analysis with the point data and not just smoothed out lines.

It's not uncommon for exit poll data to not match up very well with the actual results...in this country we just change it to match the official results after the fact, so yeah...

 

[socal]

The guy who did this analysis added an update,

http://www.dailypaul.com/214415/guess-what-sc-primary-results-in-question?page=4

****Analyst Adds Update****

Hey guys. I just want you all to know that I have graphed in detail all of the counties in NH, SC and Fla that have the precinct information available on the Election Commissions' website. I have amassed a couple of hundred graphs probably. The most difficult part of this is getting this information into a form that is brief but easy to understand. Please appreciate this.

There are surely exceptions to the following observations, but here are some generalities:

1. In any county where Ron Paul has more votes than Mitt Romney using the low vote total precincts, you get a ridiculous- looking curve like the one in Anderson County. (Anderson, Greenville, Spartanburg, and Oconee Counties). Mitt ends up at a vote total that could have been Paul's projected total and Paul crashes to the ground.

2. In the counties where Romney has more votes than Paul in the low vote total precincts, There is no ridiculous anomaly like the one in Anderson County.

3. In any race where Newt is ahead of Romney and Romney is anywhere close to Gingrich in vote total, Newt gets flipped by Romney (Richland, Charleston, and Beaufort Counties in SC) much like the maneuver in Anderson County where Mitt flips Paul. It appears to me that Newt actually won these counties as well as Polk and Duval Counties in Florida.
...

 

[affa]

This needs to get out there.

Has anyone ever seen Freakonomics? In it, they show how statistics uncovered deep rooted cheating in the world of sumo wrestling. It was fascinating stuff. This is not unlike this.

This is massive. Who the hell downvoted this thread?

 

[slamhead]

Another thing that jumped out at me. The algorithm seems to not kick in until Paul and Romney had over 1000 votes each...actually looks like 1100-1200 where the lines start to diverge. Let me know if I am reading this wrong.

Looking at it closer it looks like that right around 1000 is where the cheating begins. Is there a way you can do a blown up graph of the diverging point?

 

[socal]

Another thing that jumped out at me. The algorithm seems to not kick in until Paul and Romney had over 1000 votes each...actually looks like 1100-1200 where the lines start to diverge. Let me know if I am reading this wrong.

Looking at it closer it looks like that right around 1000 is where the cheating begins. Is there a way you can do a blown up graph of the diverging point?

The 1,000 vote figure is an artifact in the way the plot was constructed. The precincts are sorted by size, and the y-axis is the cumulative vote total. It might be more informative if instead of the plots that were given, colored scatter plots were provided: the X-axis = the precinct size, and the Y-axis = the % of votes carried by each candidate. Possibly that is what poster DEGuy meant above.

 

[slamhead]

My first thought was that the smaller precincts were experiencing a commonality where a small handful of supporters in each precinct were experiencing a sphere of influence and the 277 number was the statistical breaking point. Even under that scenario the break would statistically break amongst all the candidates. I think you are onto something. I would love to see NV and IA broken out the same way. Did they ever even publish the total results for Clark county. As we know, the Adelson caucus was counted in the open and Ron Paul won that caucus with twice as many votes as Romney. What were the other results in that count?

 

[SovereignMN]

I know in Minnesota Paul tended to win the lower attended precincts while Santorum cleaned up on the larger attended ones. This was mostly a demographics issue. The lower attended precincts were in the urban areas where there are less Republicans while the suburban ones have a huge base of evangelicals and more traditional Republican voters.

That wouldn't explain the massive drop in Newt's graph though.

One thing that has always struck me as odd about South Carolina is that their turnout was over 30% larger than in 2008. No other state has even come close to duplicating that. In fact most of them have declined and the ones that increased (Iowa, NH) were increases in turnout of less than 5%. Also, Romney's percentage increase in votes from 2008 was over 100% !!! I think he was around 25% increase in NH but never anything that approaches a 100% increase. Certainly seems odd that his biggest increase in total votes would be from a state like South Carolina, capital of the bible belt.

 

[gb13]

Wow. This is very suspicious looking to me. Seems too close to be coincidence.

 

[floridasun1983]

Exactly. They would have to be rigging the exit polls too. Sorry, but the burden of proof is high for a reason. I would like to see the same analysis with the point data and not just smoothed out lines.

Well said.

 

[J_White]

wha!!

 

[bbwarfield]

this is gonna sound terrible.... but is there any connections between diebold (the original makers of the voting machines) and bain capital? or there subsidiaries.... I have this sneaky sneaky suspicion.... and im running out of tin foil at my house

 

[affa]

Exactly. They would have to be rigging the exit polls too. Sorry, but the burden of proof is high for a reason. I would like to see the same analysis with the point data and not just smoothed out lines.

There is one sure fire way to see that analysis... and it doesn't involve dismissing someone else's work.

 

[Liberty1789]

Some illustration from Washoe County, Nevada.

Calculating correlation from the smallest precincts ordered by growing number of votes and using the first 20% votes cast (1337 of 6697), you end up with very steady straight lines, like in the SC original post. Linear correlation coefficients are 0.9992 for Gingrich, 0.9981 for Paul, 0.9995 for Romney and 0.9974 for Santorum.

Extrapolating those straight lines, you would expect Paul to get 1,628 votes and Romney 2,381 for a total of 4,009. In reality, they have got 4,004 combined, with Paul at 1,168 (460 below expected) and Romney at 2,836 (455 above expected). It is just an amazing repetition of Anderson County, SC...

 

[wgadget]

And don't forget to factor in the horrible weather that day in SC. With soaking rain and tornadoes that day, it's amazing that the turnout was so high, IMO.

 

[bcreps85]

Some illustration from Washoe County, Nevada.

Calculating correlation from the smallest precincts ordered by growing number of votes and using the first 20% votes cast (1337 of 6697), you end up with very steady straight lines, like in the SC original post. Linear correlation coefficients are 0.9992 for Gingrich, 0.9981 for Paul, 0.9995 for Romney and 0.9974 for Santorum.

Extrapolating those straight lines, you would expect Paul to get 1,628 votes and Romney 2,381 for a total of 4,009. In reality, they have got 4,004 combined, with Paul at 1,168 (460 below expected) and Romney at 2,836 (455 above expected). It is just an amazing repetition of Anderson County, SC...

My jaw just dropped when I read this. At first I was interested in this because I wouldn't put it past these people, but I had a suspicion that as we looked further into this we'd find some kind of pattern and start to come up with other explanations. The chances of Romney being the exact opposite of Paul's "target" though in multiple occasions can't be that high...

Keep up the good work. When there's enough, it needs to be documented in a pretty package and shouted on mountain tops. I'd start making offsite backups, giving copies to friends/family as you progress, etc etc. I'm going to go buy some tin foil now...

 

[puppetmaster]

we all need to get on your state gop committee and change them all to caucuses....stop this vote BS once and for all

 

[socal]

Some illustration from Washoe County, Nevada.

Calculating correlation from the smallest precincts ordered by growing number of votes and using the first 20% votes cast (1337 of 6697), you end up with very steady straight lines, like in the SC original post. Linear correlation coefficients are 0.9992 for Gingrich, 0.9981 for Paul, 0.9995 for Romney and 0.9974 for Santorum.

Extrapolating those straight lines, you would expect Paul to get 1,628 votes and Romney 2,381 for a total of 4,009. In reality, they have got 4,004 combined, with Paul at 1,168 (460 below expected) and Romney at 2,836 (455 above expected). It is just an amazing repetition of Anderson County, SC...

FYI I found a random number generator on the web and got 17 numbers, (40,5,54,25,46,25,12,97,77,21,27,24,4,10,73,58,91), and then got a correlation coeff of 0.2488 using another web program I found. Then when I added the random numbers together I got a coefficient of 0.988. To repeat, it would be interesting to see a plot of %vote vs precinct size for each of the 4 candidates.

 

[S.Shorland]

Maybe just inform Ben Swann and Rachel M of the existence of the Daily Paul thread and this one? Then they can start doing some research with their own contacts.If this can get out by Super Tuesday it might be a game changer.Interesting that Newt might be affected too.This is potentially as big a story as Watergate if provable.

 

[JJ2]

I have run the same analysis on Iowa (all precincts together). Similar patterns are visible and would indicate votes going from Paul, Bachman and Perry to Romney. Gingrich and Santorum unaffected by precinct size.

I have run Washoe County, NV. Same pattern, only from Paul to Romney. Gingrich and Santorum unaffected by precinct size.

Are you aware of the fact that the Washoe County chairman has publicly admitted that the numbers he gave the state GOP are different than the numbers they reported? He said he gave his numbers directly to the media, including the local paper and the AP, but they simply never published them. It was incredible. The state GOP even changed the numbers slightly for a number of hours, before switching them back.

After that, the county chairman just went silent about it.

Here's the Washoe County chairman's Twitter: https://twitter.com/#!/dbuell53