geonuc wrote:
Sorry, I can't agree. Your premise and especially your conclusions are, in my mind, flawed. Statistical analysis is not necessarily 'one of the best ways' to identify election rigging and you certainly can't conclude there are 'two possibilities for errors', particularly those two. I know enough about statistics from the work I do to know that laypeople like you and me and everyone else on FWIS who aren't trained in a statistical analysis can't draw conclusions from such data.
Which is why we rely on statisticians who around the world use these techniques for checking elections and they use it because it is a proven technique.
I can conclude that they are the most likely possibilities for reasons I'll explain below. You however are allowed and encouraged to disagree.
I'll even help you:
https://www.nytimes.com/2016/06/28/upsh ... nders.htmlThis author states he has no idea why people think exit polls are considered a valid fraud detection technique and thinks they are not accurate. Here's the problem:
http://electiondefensealliance.org/freq ... exit_pollsAround the world, exit polls have been used to verify the integrity of elections. The United States has funded exit polls in Eastern Europe to detect fraud. Discrepancies between exit polls and the official vote count have been used to successfully overturn election results in Ukraine, Serbia, and Georgia.
Yes this paper was peer reviewed
Have your papers been peer reviewed?
Yes. There is no formal mechanism for papers like this (nor is there any good forum in which to publish them), but when I leave a "t" uncrossed in these papers, people write to the dean and demand my dismissal (actually, they do that anyway). The conclusions of the initial paper, in fact, have been accepted, and the "debate" has moved on.
The US Count Votes paper which I co-authored with 11 mathematicians, statisticians, and other social scientists was extensively peer reviewed.
I will note here that this paper is different from the one I linked to. It uses exit poll data and has been peer reviewed and the conclusions accepted. I put it in to debunk the above link that claims exit poll data is not valid. Clearly it is and is widely accepted. That means the reporter and self acclaimed numerical analyst (I checked his bio and credentials) did not do his due dilligence at all considering he is claiming to have experience in the field of polling data analysis. Or he is playing spin doctor. Take your pick.
Here's the link to that study:
http://electiondefensealliance.org/freq ... z4Xp3h8UNFOf course you can choose what you want to accept.
geonuc wrote:For example, one other possible reason for the EP/VC discrepancy is what some people have been saying since the election (where Clinton was heavily favored to win over Trump) is that polls - even exit polls - are becoming less reliable because people simply lie about their voting preferences more than they used to, for whatever reasons.
And there could be other reasons for the data looking the way it does. I don't know.
The problem is that the same interviewers at the same locations on the same day also took data where there were simultaneous republican elections and the Republican data is well within error and has none of the anomalies.
When people lie at exit polls it won't be limited to just the Democratic voters. It also would require Sanders voters to claim the voted for Hillary or Hillary voters to claim they voted for Sanders but always with a leaning such that it came out that Hillary appears to get more counted votes than the exit polls show. Now the article I linked above (that contradicts my views) provides alternative explanations which seem legitimate. I just don't agree with that because the data corrected for demographics and because the problem below.
The second problem is that as the electoral population went up, so did the discrepancy. That means that the more people that were sampled the greater the percentage of people lying. Again, only in one direction and again only in the case of Democrats. This one makes no logical sense whatsoever statistically from everything I know and I do know enough about statistics to understand that the more sampling we use the better the result should be. In fact I use this very same property of statistics in engineering to get more accurate results than should be possible. It works, it's proven. it's reliable and is the cornerstone of much of the technology we use today. It never gets less accurate. This data however goes completely against it. This aspect of the data was left out of the article I linked above. It was not addressed at all. To me this is the strangest part of the data and I suspect any statistician would be equally interested in what is going on. The author missed or chose to ignore this aspect and if he did choose to ignore it, I can understand why. It's not explainable in any way other than fiddling the data.
Highly noted and regarded statisticians have raised questions about the issues. The report I linked was written by lawyers and statisticians and they have concerns. They are not laymen. If they are raising questions, maybe we should listen.
Sure it's not proof. It is however enough to say things aren't as simple and clear cut as they appear. Given the rest of the evidence they also collected I believe that's a given.
There is evidence it has happened before (the other link I gave).
I have shown evidence of someone forging peoples signatures to alter their registrations. This was not a one off and certainly not just one state. It was systematic. No investigation has been initiated by the DNC and they would want one if someone was hacking their database.
I have provided links and sources to leading statisticians, mathematicians and peer reviewed papers whose conclusions were accepted that used exit poll data to detect fraud. I have shown that exit poll data is used and has been used to successfully detect fraud in elections by the US government overseas. There is no reason to think that it works elsewhere but not here.
Lastly, you mentioned that people lie more and more at exit polls. The correct thing to say there is that there is a discrepancy between exit polls and counted votes that is growing. Lying is one possible explanation. So is more systematic rigging of the results. I call Occam's Razor here. People all around the country lying more and more as an explanation, or count tampering on the increase.