Smaller story: survey research theory in Kuhnian "crisis". Same issues as ‘16, missing clown voters, almost same result as '16, a Trump win over an HRC predict. No blue tsunami, not even a wave: no Senate control; lost House seats. For second POTUS elec in row pro-clown pollster, Trafalgar, ('16: Rasmussen) most accurate. RCP's final forecasts:
RCP Electoral College prediction accurate as to winner, Biden, but damning with faintest praise. Nationally, RCP pop vote: Biden +7.2, for real so far Biden +2.4. A HUGE 4.8% miss.
FL was a huge miss and PA may turn out to be. RCP's average (supposed to do away with margin of error) FL: Trump +0.9. Actual FL vote Trump +3.4, a 4.5% error. If PA stays Trump and Biden wins NV Biden will become president with 270 EV's, the minimum.
It goes on and on. Ohio RCP final: Trump +1.0; real, OH Trump +8.1, a humongous 7.1% error. MI RCP final Biden +4.2; real MI vote Biden +2.7.
Iowa RCP Trump +2; real Trump +8.2. WI RCP Biden +6.7; real: Biden +0.7.
How to explain this? Both floridly pro-clown Trafalgar and arch-Lefty Michael Moore posit “lying clown voter” syndrome. Not “shy clown voter,” LYING clown voter. Won’t tell pollsters truth because pollsters part of “deep state”; wanna “make libs cry”. Trafalgar says conservatives lie more than libs (knew that); they smoke out lying clown voters by asking them how neighbors are going to vote in addition to how are you going to vote. Very tricky. Very trickily accurate. Big Boy pollsters scoff at lying respondents theory. They have to scoff, the "normal science" survey research theory depends on truthful respondents. Lying respondents throws theory into crisis. So what's normal survey research scientists' explanation going to be this time? One says if it's lying (won't concede), it's Trump-only phenom. That might be but since Trump came so close to ending democracy in America that's not a satisfying explanation.