After President Trump curbstomped the brains out of Felonia's ears, they STFU.
???? If Independents are more pro-Trump than thse pollsters assume, then shouldn’t oversampling Indies provide pro-Trump results? What is he saying?
If we thought the polls in 2016 were ridiculous, the next three months will likely make those appear unbiased.
Reminds me of the recent past when all raging libs (e.g., Bill Maher) CLAIMED they were independents and religiously carried the liberal water, thus “proving” that even the “independents” were against conservatism in this (and every) case.
Bill Mitchell is a dim bulb. A ‘trust sessions’ siren.
Why people like and retweet his crap is mystifying.
Trump voters lied to or hung up on the pollsters in 2016. That’s why almost all of the polling was wrong.
It’s happening again.
Theyre going to slant the polls again, so that when they dont take either house of Congress, they can scream some more about Russian meddling. They have to pretend that when they dont win, its always because of some skullduggery.
Dont people understand, this is, or would be, a one-party state if not for fraud of some sort. Thats their story and theyre sticking to it.
Amazing the lengths the Left goes to defy nature.
Twitter is just so much BS.
Every time I go there I am wait listed or Twitter is taking to load. Try again later
Idiotic communications platform
Sorry, Twitter is taking too long to load
Again...
There hasn’t been a fair poll in over 40 years since the democrats learned how to manipulate them. Everyone should do themselves, and everyone else a favor and ignore polls and pollsters.
Bill Mitchell was grounded, focused and accurate, something much needed during the 2016 campaign. He kept me sane during the dark days as the corrupt media tried to depress those of us opposed to the thought of a Hillary presidency. Mr. Mitchell knows his stuff. He’s a twitter must follow.
Why polls are so often wrong:
When pollsters call and get no answer, or a hangup, they just move on. Instead, and for statistical accuracy, they MUST treat these calls as part of the sample with the replies as “unknown”. This would of course force them to use MUCH larger samples to obtain the same margin of error (statistical confidence). But they just skip them as if they hadn’t occurred.
If they included the actual sample size, and counted the hangups / no answers among the sample, their margins of error would be more like +/- 10% instead of +/- 3%