Yeah Nate has god awful politics and political instincts, but his statistical predictions are generally good. He was one of the few to give Trump more than 2% chance of winning in 2016 and was the only person that came close to predicting Bernie's downfall in the primaries. I just think this election is impossible to predict due to covid and the chaos that it brings.
They recently did an article on the "shy Trump voter" and concluded they don't really exist, even in 2016. Not in large enough numbers anyway, and it made sense to me, given that Trump overperformed polling in red states and underperformed his polling in blue ones, which is the opposite from what you'd expect if the shy voters did exist. They said the swing state polling errors in 2016 were largely due to failing to adjust for education, which many polls now do.
Not unlike the "Bernie-to-Trump voter" myth that Hillary stans love to parrot. The phenomenon was nowhere near as widespread as the Blue MAGA crowd would lead you to believe, and it was mostly confined to open primary states where those voters weren't Democratic loyalists to begin with.
I agree that polls should be getting better not worse, but 2016 did have a lot more undecided voters, especially compared to an election like 2012 where everyone basically had their mind made up already. On August 12th, 2016 the polls had Hillary beating Trump 45% to 38%, so 17% of voters were undecided or voting third party. Even on election night the number was still 12%. Compare that to this year where there's less than 9% third party/undecided voters at the moment.
Trump won largely because he got all the Gary Johnson defects at the very last second, there's no similar big block of voters to count on this year in the same way.
deleted by creator
Nate has awful political takes due to galaxy pundit brain, but does know statistics, and people dunking on him get confused between the two.
Yeah Nate has god awful politics and political instincts, but his statistical predictions are generally good. He was one of the few to give Trump more than 2% chance of winning in 2016 and was the only person that came close to predicting Bernie's downfall in the primaries. I just think this election is impossible to predict due to covid and the chaos that it brings.
deleted by creator
They recently did an article on the "shy Trump voter" and concluded they don't really exist, even in 2016. Not in large enough numbers anyway, and it made sense to me, given that Trump overperformed polling in red states and underperformed his polling in blue ones, which is the opposite from what you'd expect if the shy voters did exist. They said the swing state polling errors in 2016 were largely due to failing to adjust for education, which many polls now do.
Not unlike the "Bernie-to-Trump voter" myth that Hillary stans love to parrot. The phenomenon was nowhere near as widespread as the Blue MAGA crowd would lead you to believe, and it was mostly confined to open primary states where those voters weren't Democratic loyalists to begin with.
deleted by creator
I agree that polls should be getting better not worse, but 2016 did have a lot more undecided voters, especially compared to an election like 2012 where everyone basically had their mind made up already. On August 12th, 2016 the polls had Hillary beating Trump 45% to 38%, so 17% of voters were undecided or voting third party. Even on election night the number was still 12%. Compare that to this year where there's less than 9% third party/undecided voters at the moment.
Trump won largely because he got all the Gary Johnson defects at the very last second, there's no similar big block of voters to count on this year in the same way.
what is his statistical knowledge worth when it doubles as a get out of jail free card when his statistical knowledge fails?
What?
I'm cursed with interest in nerd shit
deleted by creator