I wrote about this Tuesday morning, when it became clear that Trump's supporters (and Trump himself) were going to be citing unscientific online surveys as evidence that Trump had triumphed. I noted that a CNN-ORC poll conducted using a scientific methodology with screened participants showed what most Americans who watched seem to have seen: a Hillary Clinton victory. But the Trump camp was more interested in nonscientific surveys hosted at Time or on Twitter or at the Drudge Report — polls that anyone could click on and which made it onto globally circulated lists of surveys that encourage frequent voting — because they showed a different result. It's the difference between making vegetable soup from a recipe and making vegetable soup by telling anyone in the world to send in whatever ingredients they want, resulting in a bathtub full of water and vodka in which three carrots, a potato and a dead pig are floating.
On MSNBC on Thursday, Trump spokesman Jason Miller was called out for referring to those online surveys as somehow being legitimate. And Miller, who has been around politics long enough to know that the surveys aren't actually indicative of anything — but who also gets occasional paychecks signed by Donald Trump — offered up a tepid defense.
"Why do you think — multiple polls, scientific polls have said Hillary Clinton won that debate by a 2-to-1 margin? I mean, it's not even been close," host Chuck Todd asked.
"All right, I've got to set you straight on that one," Miller replied. "The polls that happened that night, the night of the debate, the snap polls? The ones that happened online? Those all showed Mr. Trump winning."
"Wait. Wait. What scientific poll had Donald Trump winning?" Todd interjected. "Give me one scientific poll. Everything else — those are fan polls, man! Those are polls that, like, computer programmers can mess with! Those aren't real!"
"You get immediate reactions," Miller replied. "Those are a snapshot of what people are thinking who actually watched the debate."
No, man. No. No, they aren't. That's exactly not what they are. There's no screening process to ensure anyone watched the debate, not even a second of it. There's no way to ensure respondents are voters, much less 18, much less American, much less not a pony who's tapping around on an iPad with its nose. It's total junk that can be — and, it seems, was — manipulated to give one participant an overwhelming margin of victory, which is what happened. The CNN-ORC poll and an NBC-SurveyMonkey one (which Miller tried to undermine in his conversation with Todd) showed the same margin of victory for Clinton, about 2-to-1. The online trash surveys were all over the place: Trump at 80 percent or so at Drudge, Trump losing in one hosted on Twitter by Alex Jones. The results are all over because there's no limit on who participates! Trump fans passed around images of all the different results showing Trump winning the garbage surveys, apparently indifferent to the fact that such massive deviations in results suggest that there's a huge problem with how the numbers were collected.
"People that were watching thought that Mr. Trump won the debate," Miller said. His evidence? Junk surveys.
Todd, frustrated, pointed out that Miller's assertion simply isn't true. "You've been doing this a while. You know those are bogus! You know these are bogus! They're beyond nonscientific," Todd said about the surveys. Miller smiled.
On Tuesday morning, Kellyanne Conway, Trump's campaign manager, tweeted about the same garbage surveys as evidence that Trump had prevailed the night before. Conway is a pollster by trade — and definitely knows that the surveys show no such thing. In short order she deleted the tweet.
That poll also shows that Trump lost the debate.