Surveys of multiple states are meaningless.
Any survey is meaningless if it has a distorted sample -- like reaching only people with landlines.
Or like reaching only people that use internet, huh.
According to 538 both pollster have B rating.
Yougov is D+1.6.
Emerson is R+1.3
I admit, that Yougov is very likely better than Emerson, but still.
Methodologies:
Emerson — only landslide. 40% have cellphones only, so it is indeed a very stupid methodology, but those groups that have higher turnout: older or/and high-income voters, probably are OK. Younger though must be highly underestimated.
Yougov — internet only. 15% don't use internet, but who are they? According to Pew are those older mostly white people often with low income, often with low education, often rural. Pretty much Trumpish, no?
So yeah, Yougov is better. But it also has some problem. That's why I average. 538, Upshot, Betting, you chose. But don't cherry pick...
