So, does this election finally get us to stop calling crappy polls junk?

[#TheShadowyAbyss]

Freaking LOLras, Emerson and Trafalgar were pretty good and for the most part accurate with regards to the final result. So can we stop jumping on the OMG X AND Y POLLS ARE JUNK!!!!!

[Ronnie]

I don't even know what to think about polls anymore.  It could be that Trump being who he is made this one race more unpredictable and difficult to evaluate with data than it should have been.  Or, it could be that polls are fundamentally inaccurate now.  Who knows?  This election confused the hell out of me in so many ways.

[Ljube]

The forum is dominated by younger, university educated, or soon to be university educated people.
They naturally think they know all there is to know.

They also trust the polls, naturally.

However, polls have certain limitations. They highly depend on the quality of samples, but more importantly, on the quality of demographics stratification. If the pollsters don't get the strata right, the polling result will always be skewed.

In this particular election, it became clear to me, very early on (October or November last year) that the key factor would be crossover votes - reliably Democratic working class voting Trump.

The polls didn't pick that up. First time voters voting for Trump also didn't register. Likewise the deplorables (racists, bigots and such). Ten months ago I wrote that PA may eventually be decided by some 30000 or so deplorable voters.


Lessons learned:

1) Don't trust the polls, they are frequently wrong
2) Don’t trust polling of early voters - these polls are terrible
3) Focus on the state fundamentals
4) Adjust based on the early vote
5) Adjust based on the actual vote

[The Other Castro]

Of course not. We're all going to forget many of our mistakes very very quickly.

[Devout Centrist]

It's one cycle. 1980 didn't invalidate polling.

[🦀🎂🦀🎂]

nah just means we can call all polls in general junk

[vileplume]

No but people should stop taking them to be the be all and end all. If the methodology is junk then it is a junk poll regardless if it 'accidently' gets the end result correct.

But this election should serve as a lesson for people to be less dismissive of people with different opinions on the way the election is going. For example instead of dismissing someone who says Trump could win Pennsylvania with a 'lol no' it would be far more productive to discussion to ask them how they think a Trump win in that state would be possible (as it happens by getting huge margins out of the rural counties and counties like Luzerne sharply swinging his way) and assessing that on it's merits rather than just dismissing them out of hand

[kyc0705]

nah just means we can call all polls in general junk