(07-16-2020 10:18 AM)WalkThePlank Wrote: True...but a national polling average of more than 5% likely makes the Electoral College a moot point. 77,000 votes separated 2016 over less than 3% popular vote advantage for Clinton.
Not following. Makes it moot to what? When would the EC vote EVER be moot as it decides the White House?
California and NY (and a few others) could have gone for Hillary 100% and the other states and territories gone 50.1/49.9... giving Hillary a 20% popular advantage, and she still would have lost the election. Am I missing/unaware of something?
(07-16-2020 10:22 AM)Redwingtom Wrote: Seems there were quite a few national polls right up until election day in 2016.
https://www.realclearpolitics.com/epolls...-5952.html
You clearly misunderstand what I'm saying.
You can obviously create a national poll by aggregating state polls, but that's not the same as conducting a national poll. If your state polls are flawed, they will not be corrected by a national poll. The fact that they 'got the national number right' is not because they accurately predicted the outcome, but ACTUALLY because they were (in my example) wrong about swing states and wrong about turnout in western states after 'difference making' surprises in the east.
Again... If you're already surveying 50 states and making predictions and estimates about turnout in each of those states, leading to a number that matters... the electoral college votes....
WHY would you then conduct a seperate 'all state' poll and make predictions about turnout nationally, leading to a number that really didn't matter?
You (almost assuredly) wouldn't. you would simply aggregate the polls that you already had... that were meaningful and detailed... and in this case, WRONG.
It might be different if Trump needed to swing one state (versus the projections) to win, but he swung something like 10.... meaning they misjudged voter opinions in 10 states... and THEN, also misjudged (though in the other direction) voter opinions in a few more.
Imagine we were playing 3 sets of Tennis... and that Vegas predicted that you would win the first set 6-4, that I'd win the second 7-6 in a tie breaker and you'd win the third 6-2. They could easily then say that you'd win 18 games and I'd win 13... Now let's say what really happened was I was more aggressive in set 1 and won the first 7-6, that I strategically decided to coast through set 2 which you won 6-0 and I fought hard again while you were tired and won the third 6-4. You won 17 games, I won 13... almost exactly what they predicted, except I won the match. They were wrong about every single set/state as well as wrong about the outcome/ec vote... I won 2-1 as opposed to you winning 2-1... but they were almost spot on in the number of games we each would win.... now consider that almost EVERYBODY had that same 2-1 prediction 'your way'.
Would you say that they were 'right' with their estimates? Or would you say that they were absolutely wrong... missing how many sets we would each win and thus who would win the match... and that the ultimate number of games won being 'right' was merely chance?