Would love to hear from Nate Silver and his ilk...

[LAS14]

... about why we shouldn't fear a Columbia or brexit phenomenon here. Why should we trust our polls to accurately predict the turnout of the angry contingent?

[BainsBane]

1. Silver is a statistician

He creates statistical models and talks about the numbers. He's not a sage of the future. What you ask isn't in his purview.

[LAS14]

5. But his job is to understand how to make...

... polls accurately reflect voter sentiment and commitment to vote. He writes all the time about why polls differ and which are the most reliable.

[BainsBane]

8. Yes, and he adjusts his model to make it more accurate

but polls don't foretell the future. They are a snapshot of opinion in a moment in time.

[OhZone]

2. I dunno -

Did they really poll UK people as much as we get polled? How many polls were there?

[LAS14]

4. Dunno about UK, but a big part of the Columbia...

... reporting was how no one saw it coming. Polls didn't suggest this result. Don't know how energetic polling is in Columbia. But do understand it was a big surprise.

[alcibiades_mystery]

7. Colombia

Colombia
Colombia
Colombia
Colombia
Colombia
Colombia
Colombia
Colombia
Colombia
Colombia

[remaineruk]

3. Brexit is often cited but was different

Until the brutul murder of MP Jo Cox the leave vote was polling very well. It softened and then reverted back on voting

Also we didn't have a) trump b) nukes c) the government to vote on. And it was a binary vote. But we have shot our selves in the proverbial.

[beachbumbob]

6. If democrats stay home and not vote trump wins...what else do you

Need to know?

[obamanut2012]

9. lolz

[alcibiades_mystery]

11. Keeping hope alive

[pnwmom]

10. Well, for one thing they'd have to be registering. n/t

[Adrahil]

12. I don't think he will tell you that, but.....

They did just do an analysis that the registration of new voters seems to slightly favor Clinton in swing states. However, if a large number of white voters who typically don't vote turn out? All bets are off. Turn out models reflect some assumptions. That's why the polls disagree... sometimes by a lot. Or in the case of the USC/Dornsife poll, some really weird methodology.

[Imperialism Inc.]

13. Three reasons.

1) The polls actually showed close decision in Brexit so it wasn't a major upset. It was the prediction markets that failed not polls.

2) Republicans are losing the registration battle as far as we can tell

3) 80% of the people that voted for Trump in the primaries had also voted in the 2012 presidential election, so there is no reason to suspect he is actually drawing in new voters.