Does Selzer provide hope for the polling industry?

[AGA]

They were pretty close to the final Iowa margin. What methodology are they using and can other pollsters learn from them?

[Amenhotep Bakari-Sellers]

No, because IA, OH, TX, FL and GA are R states. Dems don't need OH and IA any longer to win Elections

[Buzz]

No.  The pollsters didn’t learn a thing from 2016 and actually got much worse.  They obviously just don’t know what the hell they’re doing.  Selzer is great but the other pollsters obviously will do their own thing.

[lfromnj]

Iowa is also a pretty uniform state so that does give Selzer an advantage, however most other Iowa pollsters sucked bad.

[AGA]

I know that most of the other pollsters were terrible. What I'm asking is does Selzer prove that good polling is still possible?

[woodley park]

Why were the polls so bad this time around?

[TiltsAreUnderrated]

Some, but with my doomer hat on, I've got to suggest that maybe their final poll was an outlier - that they just happened to get a Republican-leaning sample without actually reaching or accounting for the Trump voters affected by the non-response bias. The rest of Selzer's polls showed Democratic leads in line with other surveys and it's not as if there were many signs of a last-minute tightening. If the polling error was a result of this long-term non-response bias, the Selzer polls seem to have suffered just as badly.