[centurion316]

Monmouth results in 2016 make their polls credible. But, 2016 results is not necessarily a predictor of results in 2020. Looking at the crosstabs, I suspect that the turnout model that they used may not be correct. All of the registered Democrats are saying that they will vote, but historically many Registered Democrats don’t bother to vote. This year, with the Democrats scared witless over COVID-19, that number may be lower than in 2016. The polls, the narratives, and the historical data aren’t adding up for this election. I think that anyone who predicts the outcome is a fool or a partisan.

[OldGoatCPO]

I am lazy today and am not researching, but Manmouth was one poll that changed it's numbers a week before election day. Prior to that Manmouth had Hillary up by double digits. They used 500 registered voters, worthless in October. They only samples 900 total. Anything less then 1100 to 1500 likely voters is useless at this stage.

[ScottfromNJ]

“Monmouth results in 2016 make their polls credible.”

Clinton Maintains Lead; Senate Race All Tied Up

Tuesday, Oct. 4, 2016

West Long Branch, NJ – Pennsylvania is looking good for Hillary Clinton, driven by gains among white women voters, according to the Monmouth University Poll . She currently has a 10 point lead over Donald Trump

Monmouth.edu