Bill Mitchell, conservative radio host and all-around Truth Teller, posed an interesting question to the Internet: If you didn't have polls, how else would you 'know' that Hillary was winning? The left (and some on the right, we're ashamed to say) attacked immediately. Nate Silver even made up a little song. Whatever gets you through the night, Nate. You know you're done after this election anyway.
Almost no one, however, took it seriously. Almost no one but us. We called around to the various Mainstream Media political explainers. Not surprisingly we didn't get very far--but we've got methods and we located one Eugene Euclid who was an intern at the Washington Post's "Monkey Cage" political explainer department. Even better, he had been "let go" (fired with prejudice) after trying examine some of their 'findings.' This is what we learned:
RealTrueNews: Can you tell us your background?
Eugene Euclid: I have a Masters in Political science and a Doctorate in Mathematics from Georgia State.
RTN: And you were working as an intern?
EE: It's rough out there.
RTN: What can you tell us about the state of the race--assuming there were no polls?
EE: That was what got me fired. We started looking at regression analysis of the 'shadow constellation' around polling-data. Correlations to things like free-media exposure, yard-signs, Internet chatter and followers, and, of course, rally size.
What we started finding was . . . contradictory.
RTN: Can you explain? Hopefully in simple terms?
EE: Well, sure. So let's say you have a candidate with a lot of momentum--political scientists don't like that term--but just bear with me--and you have one without that same level of intensity and interest. You see things like bumper-stickers, yard-signs, Internet searchers, and rally attendance for the more invigorating candidate. Now, the conventional wisdom has long held that that doesn't translate directly into votes. But we started looking deeper into that. What we were finding was, well, it was kind of disturbing.
RTN: Disturbing how? What did you find?
EE: Well, for one thing, for the GOP, the 'constellations' were highly correlated with the actual results. For example, bigger rallies, more internet searches? More votes. For the Democrats . . . It wasn't. It was kind of the reverse. Sanders got a LOT more traffic than Hillary. More yard signs. Huge rallies--but it didn't correlate to votes.
That was when we looked at un-adjusted exit polls.
RTN: What--wait. What does that mean?
EE: Well, the media conducts exit polling--they ask people coming out of the polling place who they voted for. They'll pick like one out of every five people. Something like that. Then, after they get the data, before they publish it, the 're-fit' it to match the demographics and stuff they think it should represent. This is a kind of black-box. It means that if you are looking at the final numbers--the ones you see on TV--they might be off the actual vote but they're not that far off.
RTN: You got the un-edited numbers?
EE: Sure. We worked for WaPo. It was easy--but then when we ran the analysis it was right there: the same correlation we saw with the GOP. It was clear as day: Sanders should have won if the exit polling was even half-right. Yard signs--for this election anyway--were much tighter. The correlation wasn't strong by any means--but it was trending right. Rallies though? And Internet? Small-dollar donations? Those were smack on. Sanders should've had a maybe 12 to 15 point lead. Easy.
RTN: What happened?
EE: Well, I tried to get John [ John Sides, head of the Monkey Cage blog at the Washington Post ] to do a story on it or even ask around and see if anyone was seeing the same things.
RTN: What did he say?
EE: He told me to shut up. He said it was toxic--nuclear-waste-level stuff. That I'd better keep my mouth shut. He took away all the raw data we'd gotten.
RTN: Did he say why?
EE: Not exactly--but it wasn't hard to figure out. Bezos [ Jeff Bezos, founder of Amazon.com, and new owner of the Washington Post ] has had issues with Trump and Sanders for a LONG time. Ever since the Mech-Turk fiasco, they've been dependent on all these visas and illegals that Trump's going to get rid of. We think that he probably set the rules for the adjustment in various place so--
RTN: Hold on--what was the Mech-Turk fiasco. Is that the Google thing?
EE: [laughs]. No. Bezos wanted to mechanize his warehouses. Put thousands of Americans out of work and replace them with robots. So he did what a lot of these guys do: he took the lowest bid on a robot solution from an Indian company. They delivered the robots as asked and started working. It looked pretty good until they discovered that the robots were actually--
EE: Well, they were really cheap. It turned out one of the Amazon guys went to the test warehouse and caught these illegal workers getting into the stocking robots. They were just little golf-carts with mechanical arms dressed up to look like robots. It was no wonder they were so cheap. But they were way, way cheaper than Americans so Bezos went with it.
But he needed to get legal so he wants visas and stuff. So: no Trump. No Sanders. He wants Hillary. Big donor too.
I think that's why Sides canned us. Probably.
What you see above is the data that we were able to get showing the rally size compared to the vote percentage for the Republicans and the Democrats. As you can see it matches almost perfectly for the Republicans. For the Democrats, Clinton would have lost to O'Malley.