U.S. 2016 Unadjusted Exit Poll Discrepancies Fit Political, not Statistical, Patterns

[Coyotl]

Last edited Fri Dec 9, 2016, 07:49 PM - Edit history (2)

U.S. 2016 Unadjusted Exit Poll Discrepancies Fit Political, not Statistical, Patterns
By Ron Baiman Ph.D., Chicago Political Economy Group and Benedictine University

1) Introduction
As I write this in late November 2016 press reports indicate that Wisconsin has agreed to conduct recounts based on petitions filed by the Stein Green Party, and De La Fuente independent, Presidential campaigns, and the Stein campaign has raised almost $5.7 million for this effort and for additional recounts in Michigan and Pennsylvania. If voting irregularities are discovered in these three states sufficient to overturn Trump’s exceeding small victory margins (Michigan, 27,200 in Wisconsin, and 68,000 in Pennsylvania), Clinton who has an over 2 million popular vote lead over Trump, will win 276 electoral college votes and become the next President of the U.S. Already three Wisconsin precincts have been found to have given Trump more votes than he received. As will be shown below this is consistent with 2016 analysis that shows a pattern of highly significant unexplained increases in Trump’s state vote counts relative to unadjusted exit polls in battleground and deep red states. Politically, but not statistically, consistent patterns of UEP discrepancy have also been apparent in earlier U.S. elections.
After a short introduction (Section 1) this paper will include an analysis of Presidential UEPs (Section 2), Senate Race UEPs (Section 3), and a short Conclusion (Section 4). Figures illustrating the analysis, provided courtesy of Greg Kilcup and Peter Peckarsky, will be presented for: Clinton in PA (Figure 3, p. 8), Trump in WI (Figure 5, p. 11), Trump in NC (Figure 6, p. 12), Trump in FL (Figure 7, p. 13), and Dem Senate Candidates: Kander in MO (Figure 11, p. 17), Feingold in WI (Figure 12, p. 18), and McGinty in PA (Figure 13, p. 19).
a) Unadjusted Exit Polls
If you google U.S. Presidential Election exit polls you will find multiple reports and analysis that, unlike pre-election “polls,” purport to provide analysis of the demographics and voting preferences of actual voters. However it is important to understand that these “exit polls” are adjusted versions of actual exit poll data that approximate real exit polls only to the extent that official vote counts are accurate and that the adjustments made are good approximations of what would have resulted from unadjusted exit polls that roughly matched the official vote count without adjustment. None of this is “conspiracy theory” but rather has been repeatedly confirmed by executives of the polling company Edison Research that conducts the exit polls for the mainstream media consortium in the U.S. For example ...........

U.S. 2016 Unadjusted Exit Poll Discrepancies Fit Chronic Republican Vote-Count Rigging, not Statistical, Patterns
https://docs.google.com/viewer

[Madam45for2923]

1. Bookmarking! Thanks!

Looks interesting! Shoot!

[underpants]

2. Bookmarking

[jmg257]

3. Still don't get this - at 8pm Nov 8 in PA, there were 2613 respondents, and HRC is up.

At 3:35pm Nov 9th, with 2935 respondents, trump is up 1%.


So they made up 320 respondents in order to change the results??

Is "adjusting" the polls actually making up responses??


““We will know shortly after the polls close,” Lenski said. “We’ll have individual precinct results from all
the locations where we conducted interviews, so we’ll know how much understatement or
overstatement for the candidates we have. Our calls are based on all the information we have at the
time – exit polls, returns from sample precincts and county results from AP – and we may re-weight the
exit poll results later in the evening to match the vote estimates by geographic region.”

[Pacifist Patriot]

4. I wish I could understand this stuff.

My brain just does not work that way. I need someone who will assume my number crunching ability is stuck in grammar school to explain it to me.

[Takket]

10. he's showing a probability dome

the peak of the dome is the most likely outcome, which is the actual vote matching the exit polling data. it is also likely that the exit polling is off by .1% or .2%. Right? Math tells us (by analyzing margins of error in the exit polling data) what a reasonable deviation from the exit polling data would be. If you look at figure 3, the author has made this + or - 2.5%. Meaning that if exit polls said Clinton 50.5%, it is reasonable to assume her actual total will be between 48% and 53%. He calls this the 95% confidence interval. This means that if the election were held 100 times, that 95% of the time Hillary's total would be somewhere in that interval (48% to 53%).

Her actual total of 47.6% falls outside this interval, meaning it is highly unlikely that the exit polling data was that far off.

That being said, there are 50 states, so the concept of a state being outside the confidence interval is not only believable, it is probable. The problem though is that several states are falling outside this confidence interval, meaning that "the odds of this happening are 5%" event is having many times, which is highly improbable. imagine you have 5 white balls and 95 red balls in a bin, and one at a time, you keep drawing white balls but no red ones. pretty unlikely right? The data is basically showing that white balls keep being drawn.

[Pacifist Patriot]

14. You lost me at "probability dome"

Just kidding. I appreciate the summary and understand. Thanks!

[triron]

15. good analogy

[uponit7771]

19. Are those states falling outside of confidence level Voter Suppression states? tia

[Coyotl]

23. That hardly matters since the exit polls survey only voters.

There could be a bias if provisional ballots are thrown out later. I doubt the exit polls asked if the voter voted provisional.

[Coyotl]

22. "This means that if the election were held 100 times" should be if the poll were conducted 100x

Then the pollster would be confident of matching the election result to within the margin of error 95% of the time.

When the poll and the vote count are outside the margin of error it means there is almost no chance at all that randomness can explain the deviation. There has to be another explanation. One possibility is the vote result is not true. The other possibility is to polling is defective. When you compare 2004 and 2016, you have to ask why would the same polling methods suddenly produce double the error.



Same states with the same Senate races and the same pollster using the same methods, but in 2012 they are off by twice as much plus the greatest error is in Republican-dominated states. In states with Democrats in the Secretary of State position, the election results and polling match much better. In states where the republicans were losing by the widest margins in the polls, the exit polling shows the largest red shift. The red shift even matches how much the election had to swing for Republicans to win.

In the graph above, if the trendlines were flat and matched zero, the exit polls would all be averaging precise and red/blue states would not bias the result. What we have instead is a pronounced red shift in almost every state. The odds of that happening are astronomical.

The battleground states had the most robust polling. This is seen in the match to the pre-election polling. What are the odds all those many polls are also off by the same amount?

A wise gambler would bet the election was rigged if they knew the truth could be shown. A wise politician knows that once an election is successfully stolen, there's about zero chance of proving it, we are just left with these exit polls for solace.

[triron]

5. Looks like presidency + 2 or 3 senate seats stolen.

[Madam45for2923]

13. Yes! This!

[Coyotl]

24. Exactly, just enough to alter the balance of power, and in just the right places too.

Which is how we end up with this graph, surgically precise electoral college theft does not alter the national trend:

[triron]

6. knr

[Madam45for2923]

7. KnR!

[Coyotl]

8. K & R

[Takket]

9. why are only 28 states shown in figure 1?

[triron]

11. I believe that is all the states

where exit polls were done.

[Coyotl]

17. That's how many unadjusted exit polls were shown election night

Last edited Thu Dec 8, 2016, 11:28 AM - Edit history (2)

There were exit polls in only 29 states, I believe. All the battleground states and states with Senate races. This is a change from past elections, when all states were polled.

The number of respondents varies by state, the battleground is reliable and the less contested states have far fewer respondents, both in exit polls and pre-election pooling. Note in the comparative graph above how close the polls are to each other in the battleground states. Again, what are the odds that dozens of averaged polls and the exit polls are all wrong and the election results are true? Astronomical. Billions to one this is not the result of randomness.

[triron]

12. The evidence that might show up would probably be in the voting machines

or in the final tallies. But seems machines can't be examined.

[flamingdem]

16. Enough of the circular firing squad

This election was stolen

[uponit7771]

20. +1, Comey, Voter suppression, Russia...

[triron]

21. you got it right

[triron]

18. knr

[Coyotl]

25. Edit: "... Unadjusted Exit Poll Discrepancies Fit Chronic Republican Vote-Count Rigging ..."

Last edited Fri Dec 9, 2016, 07:49 PM - Edit history (1)

https://groups.google.com/forum/#!topic/electionintegrity/Qa-HFOwVRrk

https://docs.google.com/viewer?a=v&pid=forums&srcid=MTI5Njc1MzU0MDY5ODE2MjQ5OTQBMTIyNzQ4MDc1MTgxNDQwOTYxNzIBaHItcmsxNHlCd0FKATAuMQEBdjI

Friends,

This is a slightly edited title and version of my earlier post ..... share/post this version of the paper going forward.

Thank you!

Ron

[triron]

27. knr

Thanks for the update

[Achilleaze]

26. "unexplained increases" Republican chicanery against America

Republicans better finally get around to reading that book - The Bible - that they like to claim is their guide in ethics.