If Nate is wrong again he is a piece of toast....

[leftofthestorm]

He will just fade away..........

[vollehosen]

1. He wasn't wrong in 2016 nt

[JI7]

2. He wasn't wrong last time

[Tom Rivers]

3. What is Nate saying?

[Quixote1818]

4. Nope, he is just going by the information given him by pollsters. nt

[Turin_C3PO]

5. He wasn't really incorrect

in 2016. I believe the final numbers were something like a 65/35 chance of Clinton winning. So Trump had a very clear path. Not this time.

[dflprincess]

7. He also had Feingold favored to win in Wisconsin

which is one reason I've doubted the Wisconsin count in 2016.

[Spider Jerusalem]

6. If he's wrong again America is toast, we have bigger problems then than Nate Silver's career (n/t)

[MLAA]

8. So true

[Ponietz]

9. We're all toast if Nate's wrong leftofthestorm

[PTWB]

10. He wasn't wrong in 2016.

He gave Trump 1 in 3 odds in 2016.

He’s got columns on 538 RIGHT NOW about how Trump could win tomorrow.

No, he wasn’t wrong. People just don’t understand probability.

[GulfCoast66]

11. You are incorrect. He is not a pollster. He aggregates polls.

He was not wrong in 16. He clearly started there was a 30 odd percent chance trump would win based on polls other people do. And the majority of polls show a Biden win. But there is still a chance trump could win. Nate has even talked about how early voting has made polls and aggregating less accurate this year. And how things like unexpected turnout for either side can make polls and aggravating way less effective.

If trump wins then the majority of polls were again off. Nate keeps up slimy polls like trefalger in the mix to try and get a good poll average.

Have you listened to any of his podcasts? It is explained extensively. You can listen to him explain it.

Nate does not predict winners. He explains what the current polling data shows. And currently it shows that trump has a chance.

[kurtcagle]

12. Trump CAN win

However, it would require a polling error double the size of the 2016 error in a race that has been remarkably stable for it to happen.

More to the OP's point, it would also require that some of the best pollsters in the country would also have to be consistently off in a race that likely is overestimating Trump's final numbers - not just 538, but the Economist, Marist, Charlie Cook and others. Again it's possible, but as someone who runs a data analytics community site, I would be more inclined to say that Trump and the GOP would have managed to completely subvert the electoral system at that point than to think that the polls are all consistently off in the same direction by the same amounts.

[DrToast]

13. *sigh* Four years later...

And people still don't understand probabilities.

[GulfCoast66]

16. Right? But people want certainly. Chance is a dicey thing.

[Dark n Stormy Knight]

18. 😅 And, as you imply,

dice are chancy things.

[greenjar_01]

14. Polls have Biden up 10+ nationally and over 50%

If they're all wrong, then the whole enterprise of polling is finished, quite frankly. I mean, what kind of a moron would you have to be to ever take a poll seriously again after that.

It's wild that nobody expects Trump to win the popular vote, and most polls (given turnout projections) have him losing it by more than McCain lost to Obama! That's really fucking incredible. When some of us middle-to-oldies were kids, a candidate winning the electoral college but not the popular vote was a wild nineteenth century scenario. Now, for Trump to win tomorrow, it will have to happen three out of the last five elections, with increasing margins (Gore 500,000+, Clinton 2.8 million, Biden 7 million+ at least). That's a nightmare scenario because it also suggests complete minority rule. I'll also note that everytime we've seated a president who lost the popular vote, it has been disastrous, calamitous. The people actually do know. Democracy is not just ethical. It works.

[Buckeye_Democrat]

15. It didn't seem to end Wang at Princeton Consortium.

His probability of a Clinton victory was insanely high, as I only vaguely recall because I dismissed it right away.

[Fiendish Thingy]

17. For Nate to be wrong would mean most major pollsters were more wrong than in 2016

The pollsters would be the ones to fade away.

[Mariana]

19. It's a right wing talking point that Nate or the polls had Hillary as a sure win.

Like every other right-wing talking point, it's a lie.

[BannonsLiver]

20. He wasn't wrong in 2016

His model gave Dump a 30 percent chance of winning on Election Day. His ball was called in the big bingo game. The end.

[jcgoldie]

21. He wasn't wrong last time

Evidently you don't even know what he does for a living.