Is Nate Silver a pollster one can trust?.. or has he jumped the shark

[Peacetrain]

and is permanently trying to promote Trump or Trumpism.. This is a question for those who understand polling and the process better than I do..

[Johnny2X2X]

1. He works for Peter Thiel

Trump Super Donor.

[Peacetrain]

4. Ohhh... I was wondering

what was going on with him

[bullimiami]

6. And he's not a pollster. He's a poll aggregator. Shit polls equal shit aggregates.

[Yavin4]

9. And Thiel is a straight up Nazi.

Straight up.

[ColinC]

2. He is not a pollster, he is a data wonk

I do not believe he has jumped the shark

[Peacetrain]

7. 58.2 prediction that Trump will win seems to me

selective data picking.. But I will be the first to admit.. polls confuse the hell out of me

[ColinC]

13. It's just a hobby that made him popular. None of it really means anything outside of entertainment

The underlying reasons and substantial analysis having nothing to do with predictions can shine a lot useful light on the electoral situation and demographics.

[kerry-is-my-prez]

16. Where are you seeing 58.2? It's Harris: 48.9% Trump:45.5%

Harris winning by +3.4 points. I’m not an expert, but I do dig into the polls and have taken a couple of Grad School classes on statistics and I think Silver is pretty accurate.

[kerry-is-my-prez]

17. Oh I see where he said that. Currently the momentum is moving in Trumps direction

But Harris is still ahead. Momentum can change. Especially with the debate coming up. All kinds of things can affect momentum. Although it seems that Trump isn’t hurt by anything. He can basically say or do anything without any repercussions.

[obamanut2012]

18. "Currently the momentum is moving in Trumps direction"

Nothing at all says that.

[kerry-is-my-prez]

38. Have you been looking at the polls? I look at all of them every day.

n/t

[Bucky]

22. Please keep in mind that the "chances of winning" stat is bullshit

It's not really based on probability science. It's based on non simultaneous state polls and treats fluctuating poll numbers like they are closed data sets. Only popularity polls are in constant motion and are subject to change with a long list of unknown future variables.

The thing is you can't give reliable probabilities of outcomes if you don't even know what the variables are--especially human variables--or even know how many variables there are.

State opinion polls can give you a sense of where the race might be going. The authoritative sounding "odds of winning" index can't.

[Matthew28]

3. use to be respected

about like 538 and rcp. I think he has jumped the shark. At least I hope so.

[getagrip_already]

5. Nate isnt, and never was, a pollster

He aggregates polls collected by others, massages them with "rankings", and uses pseudo science as though it was accurate to five 9's.

He was never reliable, and his rankings are suspect, allowing pollsters like rasmusen and other partisan pollsters to be included in results.

You are better off with a ouigee board.

[Peacetrain]

8. That is basically where I am at most of the times when it comes to polls..coin toss

and his (Silvers) latest 58.2 % prediction.. where the hell did he pull those numbers from?? they make no sense..

[getagrip_already]

11. Peter thiels bank account....

Nate is an analyst for hire. He lost his job at 538 and needs to make a living.

[Fiendish Thingy]

10. No, nope, noppity-NOPE

Silver got fired from 538 for allowing the averages and projection model to be openly and deliberately manipulated by a flood of shady polls in the 2022 midterms.

Just read the other day that Silver was gambling heavily (thousands of $ per month) during this time.

Now he works for an online gambling company owned partly by Peter Thiel, JD’s sugar daddy.

He should never again be used as an expert source on polling, but sadly, many MSM outlets continue to seek his opinion.

[Peacetrain]

12. Wow.. did not know that.. wow

[Fiendish Thingy]

14. Yup, and when confronted after the midterms about the manipulation,

Silver shrugged and said “well, the Dems could have done their own shitty polls but didn’t”

[SWBTATTReg]

27. Wow. Did not know this either. Talk about a way to ruin one's livelihood when one was on top of

the world.

[House of Roberts]

15. Silver has a book out

and is making the shows touting it.
Saw him on BBC News yesterday or the day before getting interviewed.
If he's taking Peter Thiel's money, he's not trustworthy.

[LAS14]

19. Thanks for launching a good thread. nt

[kerry-is-my-prez]

20. Her momentum has stopped going up and is going down.

538: Harris +3.1 9/4
538: Harris +3.3 9/2
538: Harris +3.2 8/30
538: Harris +3.3 8/29
538: Harris +3.6 8/25
538: Harris +3.7 8/23
538: Harris +3.6 8/22

+3.4 Harris - 9/3, 9/4
+3.5 Harris - 8/31
+3.3 Harris - 8/30. 8/23-8/25, 8/28
+3.8 Harris - 8/29
+4.5 Harris - 8/28
+3.8 Harris - 8/27
+4.8 Harris - 8/24-8/25
+4.0 Harris - 8/24
+4.3 Harris - 8/23
+2.5 Harris - 8/22

[Bucky]

24. Republicans have been hitting back.

They're still in the spaghetti throwing stage. They have yet to find a theme

Also there's still many debates, campaign events, news events, ad campaigns, gaffes, and endorsements left to roll out. Plus we don't know if Trump is gonna get a ground game pulled together by late October

[kansasobama]

34. Trump is betting all money on GA and PA

Yes, he will campaign elsewhere but money goes to only those two. He has assumed an NC win. So, Harris could catch up in AZ and NC. I am very angry at AZ where Gallego is running a solo campaign. I am not even sure he cares about a Harris win.

[kerry-is-my-prez]

21. Pennsylvania is concerning. She may have to rely on Georgia.

Pennsylvania:

From +1.3 to +.6:

270 to Win: Harris: +.6 - 8/29-9/4
270 to Win: Harris: +.4 - 8/19-8/31
270 to Win: Harris: +1.0 - 8/19-8/30
270 to Win: Harris: +1.3 - 8/9-8/19


From +1.8 to +1.0:

538: Harris +1.0 - 9/4
538: Harris +1.2 - 9/2, 9/3
538: Harris +1.3 - 8/29
538: Harris +1.8 - 8/23
538: Harris +1.8 - 8/22

[crimycarny]

30. Trump is not winning PA

First of all, those numbers reflect a tightening after a post-convention bump.

So many things still have yet to occur which include:

1) First Harris v Trump debate on 9/10
2) VP Debate

In addition, the news about the DOJ exposing Russian interference is going to have an impact as well.

Polls are getting more and more unreliable each year. The demographic most likely to respond to a poll are white older voters, the same type you see at a Trump rally.

Nate Silver gave Dems a 41% chance of retaining the Senate in 2022. He was also wrong in several of the House Seat winners, in one case he gave the winner, Marie Gluesenkamp Perez (Dem, W03) a 2% chance of winning.

Way too early to worry about polls, which are getting less reliable every election.

[kansasobama]

31. Yes PA is concerning

I hear it is all about Lehigh County. If she gets 43%, she wins. If not she loses PA. City of Philly has gotten redder. But suburbs are fine. If she wins, it will be closer than a Biden win.

Yes, she has to win GA and NV. Hope it is NC instead. GA is where they could steal.

[kerry-is-my-prez]

23. In Michigan the momentum is also going down

Michigan:

Going up and down:
270 to Win: Harris: +1.7 - 8/30-9/4
270 to Win: Harris: +1.6 - 8/29-8/31
270 to Win: Harris: +2.0 - 8/23-8/30
270 to Win: Harris: +1.8 - 7/26-8/19

From +3.4 to +2.3:
538: Harris +2.3 - 9/4
538: Harris +2.4 - 9/3
538: Harris +2.9 - 8/29
538: Harris +3.3 - 8/25
538: Harris +3.4 - 8/23

[Bucky]

25. Thank you for these updates

I'm naming my nightmares tonight after you as a show of gratitude

[kansasobama]

32. Kerry-Is-My-Prez is a truth teller

I don't want to always assume it is a slam dunk. Do not forget. This is a racist country and it is not a fair democracy.

[obamanut2012]

36. Lol kk

[samnsara]

26. he interprets statistics...

i used to rely on him but hes kind of all over the place lately. Hes not looking so nerdy tho..

[Indykatie]

28. Nate Silver is NOT a Pollster

All the poll data he uses for his various projections and analyses come from polls done by Pollsters. There are several pollsters who data is excluded because they don't meet Silver's requirements. Rassmussen is one of those pollsters because of their strong pro Republican bias.

[Bev54]

29. It is rumoured that he has a gambling problem, he alluded to it in his book

I would not trust a pollster with a gambling problem who needs money to feed that problem.

[kansasobama]

33. Nate Silver should join Mark Halperin

Mark Halperin uses all negative polls and comes up with his own justification for why Kamala will lose.

[Jarqui]

35. Right wing pollsters are part of the reeling in

Trafalgar, Rasmussen, etc.

They did a blast of polling to rein her surge in.

With the media seeming to have their thumb on the scale for Trump, it's hard to sort out where the race is precisely.

Ad money and some rumblings tell you that Trump seems to pulling out of NH, MN & VA.
That is confirmation that Kamala has made some plausibly permanent change to the contest.

Even Texas and Florida gaps have significantly dropped to just outside the margin of error (if you believe those polls ..).

Without doing the polls yourself, you can't be very sure.
But you can review their methods and crosstabs, etc and derive some feel for the accuracy of some of them.

My impression is she is doing well - better than Trump - but it remains pretty competitive and we have a long way to go.

I will never feel like there is a point where we can relax. She is up against some ruthless forces.

[notroot]

37. He's jumped the shark so hard he made the Fonz look cool.

I trashed Nate and 538 after 2016.

It's not just Nate and 538. ALL the polls are off by more than their own MOE.

Especially if reproductive rights are on the ballot.

[uponit7771]

39. +1, "more than their own MOE.", ... and their MOEs have been horribly high

[kerry-is-my-prez]

40. On the positive side, Georgia is doing well, we may not need PA

Georgia:


Trump goes from +1.0 to a tie, then to .1 ahead
RCP AVG: Trump: 1.0 - 7/22-8/14
RCP AVG: Tie - 7/24-8/28
RCP AVG: Tie - 7/24-8/28
RCP AVG: +.2 Tr - 7/26-8/31
RCP AVG: +.1 Tr - 7/26-8/31

Trump goes from a mere .8 to Harris being ahead, then a tie.
270 to Win: Trump +.8 - 7/22-8/19
270 to Win: Trump +.2 - 8/19-8/30
270 to Win: Harris +.6 -8/28-8/31
270 to Win: Harris +.7 - 8/29-9/4
270 to Win: -Tied- - 8/29-9/5

Trump goes from a mere +.05 to Harris being ahead by +0.4
538: Trump +0.5 - 8/22
538: Trump +0.4 - 8/23
538: Trump +0.5 - 8/25
538: Harris +0.5 - 8/29
538: Harris +0.4 - 9/3, 9/4, 9/5

I just hope the Reps don’t pull some shenanigans.we know they’re trying.