HomeLatest ThreadsGreatest ThreadsForums & GroupsMy SubscriptionsMy Posts
DU Home » Latest Threads » Forums & Groups » Main » General Discussion (Forum) » Nate Silver has some comp...

Sat Oct 10, 2020, 05:00 PM

Nate Silver has some competition

And I am nervous


?s=20

41 replies, 2301 views

Reply to this thread

Back to top Alert abuse

Always highlight: 10 newest replies | Replies posted after I mark a forum
Replies to this discussion thread
Arrow 41 replies Author Time Post
Reply Nate Silver has some competition (Original post)
Proud liberal 80 Oct 2020 OP
Sherman A1 Oct 2020 #1
jorgevlorgan Oct 2020 #2
grantcart Oct 2020 #11
jorgevlorgan Oct 2020 #14
LeftInTX Oct 2020 #28
Shrike47 Oct 2020 #3
Turin_C3PO Oct 2020 #4
sweetloukillbot Oct 2020 #15
Bernardo de La Paz Oct 2020 #35
Shrike47 Oct 2020 #5
NRaleighLiberal Oct 2020 #6
Under The Radar Oct 2020 #7
Proud liberal 80 Oct 2020 #10
still_one Oct 2020 #8
Cattledog Oct 2020 #9
BootinUp Oct 2020 #12
judesedit Oct 2020 #13
Thekaspervote Oct 2020 #18
judesedit Oct 2020 #41
Tribetime Oct 2020 #16
ProfessorGAC Oct 2020 #19
Roland99 Oct 2020 #21
ProfessorGAC Oct 2020 #33
PoindexterOglethorpe Oct 2020 #39
Thekaspervote Oct 2020 #17
Awsi Dooger Oct 2020 #22
Thekaspervote Oct 2020 #25
JDC Oct 2020 #20
Roland99 Oct 2020 #23
ProfessorGAC Oct 2020 #24
Thekaspervote Oct 2020 #26
nsd Oct 2020 #27
LineReply .
Hassin Bin Sober Oct 2020 #29
Marius25 Oct 2020 #30
LiberalFighter Oct 2020 #31
ProfessorGAC Oct 2020 #34
Just_Vote_Dem Oct 2020 #32
Quixote1818 Oct 2020 #36
BGBD Oct 2020 #37
Sunsky Oct 2020 #38
Xolodno Oct 2020 #40

Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:01 PM

1. Never heard of them

Are they even real?

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Sherman A1 (Reply #1)

Sat Oct 10, 2020, 05:05 PM

2. It's a bell weather model.

Based on simply who received a higher percentage of the vote in their primary elections. It is about as reliable as who wins the Redskins game. When they get this one wrong they will still have a 23/27 success rate.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to jorgevlorgan (Reply #2)

Sat Oct 10, 2020, 05:14 PM

11. It's like betting on the Redskins even though they no longer exist

Reply to this post

Back to top Alert abuse Link here Permalink


Response to grantcart (Reply #11)

Sat Oct 10, 2020, 05:22 PM

14. Pretty much

Reply to this post

Back to top Alert abuse Link here Permalink


Response to jorgevlorgan (Reply #2)

Sat Oct 10, 2020, 06:03 PM

28. How accurate!



I know people who actually believe primary results as general election results.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Sherman A1 (Reply #1)

Sat Oct 10, 2020, 05:05 PM

3. Yes, I looked him up on line. If he's not real, he has a very active group posting about him.

You know, that might be kind of fun, asking up a person and posting in various venues about him...

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:08 PM

4. Lol they have NM going red.

Hahaha good one!

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Turin_C3PO (Reply #4)

Sat Oct 10, 2020, 05:22 PM

15. And NY!

Guy seems to be a right wing British pundit of some sort

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Turin_C3PO (Reply #4)

Sat Oct 10, 2020, 06:53 PM

35. From the Wikipedia article


Norpoth developed[5] the Primary Model, a statistical model of United States presidential elections based on data going back to 1912. The Primary Model is based on two factors: whether the party that has been in power for a long time seems to be about to lose it, and whether a given candidate did better in the primaries than his or her opponent.[6]


I think the first factor outweighs the second this year. In a sense the Republicons have been running their con game for forty years, which is a long time and they look like they may fall hard this year.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:09 PM

5. For the model to be correct, Biden has to lose every battleground state plus states like New York.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:09 PM

6. it has NY going red!!!

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:09 PM

7. Has Putin approved the distribution of this report?

To me it appears that Trump himself conducted this poll. Not only is it completely opposite than what all other polling data is showing, but it has pushed its data far beyond the possibility of reason....like trump did with his hurricane map, his healthcare plan. If you are going to lie man at least make it somewhat believable

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Under The Radar (Reply #7)

Sat Oct 10, 2020, 05:14 PM

10. Exactly

For the lie to believable itís has to be somewhat realistic, and giving Hawaii and NY to trump isnít.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:10 PM

8. That map has NY as red? Give me a break

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:12 PM

9. Yep trump's gonna win NY



I love this tweet: " Also while the Helmut Norpoth model predicted Trump would win in 2016, his projected map didn't look anything like what really happened. He had Trump winning CO, NH, NM, NV, OR, VA, none of which happened."

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:16 PM

12. Clap if you believe!

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:17 PM

13. Just like last time, if this happens it is a stolen election due to Russian

and rightwing operatives hacking of e-voting machines We have been warned of this problem for decades now and it doesn't seem to be taken seriously. A child can flip votes and was shown doing it on national television. Why do you think Ivanka Dump bought electronic voting equipment companies from China? Please correct me if I'm wrong. Helmut may be a buddy of Pooty-poot and have inside info on the intended goings on.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to judesedit (Reply #13)

Sat Oct 10, 2020, 05:35 PM

18. Links please!!

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Thekaspervote (Reply #18)

Sun Oct 11, 2020, 02:11 AM

41. blackboxvoting.org is a link regarding the easily hacked electronic voting machines

Diebold and Sequoia if I recall correctly

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:28 PM

16. Hawaii and New York red ????

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Tribetime (Reply #16)

Sat Oct 10, 2020, 05:41 PM

19. And He's Got...

...Pennsylvania red, despite almost all other polls, and Colorado blue with a question mark.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to ProfessorGAC (Reply #19)

Sat Oct 10, 2020, 05:44 PM

21. And NM red?

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Roland99 (Reply #21)

Sat Oct 10, 2020, 06:14 PM

33. Someone Above Pointed That Out

That's why I left that one alone.
CO is a state we watch because one of my wife's best friends, since grade school, lives there.
I saw it was blue, then a question mark. Either he thinks it's a toss up or forgot to look up how many EC votes they have.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to ProfessorGAC (Reply #33)

Sat Oct 10, 2020, 07:24 PM

39. NM has a whopping 5 EC votes.

And it is absolutely not going Red. I live in NM, I know.

We will also have a House delegation that is not only all female, but all Hispanic and/or Native American.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:34 PM

17. If you do a search for his accuracy it's pretty weak sauce!

He bases his predictions on how a candidate did in the primaries, actual votes. Okay, but this guy is only looking at the first 3 democratic primaries in which Biden lost- IA, NH and NV which he didnít really loose NV, just wasnít in the first 2 slots.

What about the other 48 primary races where Biden blew the doors off?

Prof. norpoth, did you take into consideration that especially in NH and NV that 6-12% of the vote were gop crossover votes?? Voters who never intended on voting for the dem candidate in the general?

What a bunch of malarkey!!

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Thekaspervote (Reply #17)

Sat Oct 10, 2020, 05:45 PM

22. It's a logical model that ran into outliers in 2020

 

Thank goodness for those outliers. The Sanders/Biden situation in the primaries is something a model like this cannot account for, and basically doesn't respect, and likewise coronavirus impact in subsequent months is not contemplated by the model.

The model relies on normalcy and competence after the early primary season. That is going to hold up the vast majority of the time.

Donald Trump is the enemy and the asterisk of any systematic approach. He had massive advantage as incumbent and threw it away. That basically is what the model is indicating. In politics it stands out only because of the spotlight and the every 4 years aspect. In sports I have dozens of great systematic models than run into occasional outliers. Big deal. I understand the abnormality and hang onto the system. Likewise it would be silly to dismiss this model as no longer meaningful going forward. It makes tons of sense.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Awsi Dooger (Reply #22)

Sat Oct 10, 2020, 05:53 PM

25. It makes sense yes if he had sampled all the primaries..he only looked at IA, NH and NV

That makes no sense.

As a model, if all the primaries were sampled yes.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:43 PM

20. NM, HI AND NY equals quackery

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:46 PM

23. The Hoarse Whisperer brings up the obvious point!!



24 out of 26 presidential elections?

So, since 1916?

Pretty impressive since Helmut Norpoth wasnít born until decades later.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 05:49 PM

24. Tautological Nonsense

Big deal! He found a correlation between someone getting a lot of votes in a primary getting lots of votes in the EC.
Don't need a model, or even math to find that correlation.
But, this does, not at all, establish causation.
The factors that lead to a big primary total aren't necessarily the same that lead to an EC majority. This is a general consensus amount political experts. It's why candidates typically run to the middle for the EC.
It's contingent upon him to demonstrate a causative link, which is not done in the model.
And, for all know, he weighted some thing to assure that 24/26. After all, he claimed to accurately predict elections from over 90 years ago where the result was a known. Easy to tweak a model to get the correct prediction when you already know the result.
There's no valid math behind this, and it's a sloppy & simplistic model.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to ProfessorGAC (Reply #24)

Sat Oct 10, 2020, 05:55 PM

26. This!

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 06:00 PM

27. Norpath isn't a nut.

He's a (tenured full) professor in the political science department at Stony Brook.

But his model revolves around the importance of vote totals in early primary states, as a proxy for an incumbent's standing in his own party and for the enthusiasm the opposition party has for the challenger.

As a thought experiment, it's fine and it's not any worse than a lot of what passes for scholarship in political science departments. But it is also very easy to criticize (e.g., Norpath's model doesn't consider polling data) and its claim to fame (successfully predicting the winner since 1996) is not particularly impressive. Other than 2016, I think I could have done that too -- no model required!

Trump supporters have latched onto Norpath because his is the only model predicting a Trump victory, but there is nothing special about it. I don't think it deserves much attention. A poll-based model (like Silver's or that of the Economist) is much easier to take seriously.

Reply to this post

Back to top Alert abuse Link here Permalink



Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 06:07 PM

30. That is the dumbest forecast I've ever seen

They have Trump losing New Hampshire, which is a fairly purple state, although it leans blue, but they have Trump winning New York, one of the bluest states in the country.

And they have Trump winning Hawaii - which has zero Republican political officials.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 06:10 PM

31. He has been doing this for the last 104 years.

So he must know what he is doing.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to LiberalFighter (Reply #31)

Sat Oct 10, 2020, 06:18 PM

34. Well, You Made Me Laugh!

So, there's that!

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 06:14 PM

32. LOL-Trump ain't getting 300 EV's this time n/t

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 06:57 PM

36. Probably designed to get a few stupid Trump supporters to bet big on Trump

Then they can take their money.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 07:03 PM

37. Maybe you should

Reconsider your model that predicts Trump to win New York.

The model is overfit to that data and its reacting very badly this cycle.

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 07:09 PM

38. They'll soon be 24/27 nt

Reply to this post

Back to top Alert abuse Link here Permalink


Response to Proud liberal 80 (Original post)

Sat Oct 10, 2020, 08:32 PM

40. Allow me to throw in another from a nut on facebook.

[link:
|

You know, you would think there wood be links behind that to support the track recoord.

Reply to this post

Back to top Alert abuse Link here Permalink

Reply to this thread