This message was self-deleted by its author

[Chichiri]

This message was self-deleted by its author (Chichiri) on Tue Mar 8, 2016, 11:27 PM. When the original post in a discussion thread is self-deleted, the entire discussion thread is automatically locked so new replies cannot be posted.

[Chichiri]

2. Why do you say that? nt

[Chichiri]

7. And why do you say THAT? nt

[Chichiri]

10. So that's just based on foreign election models? No intra-primary evidence? nt

[leftynyc]

13. Why are you even bothering

with a newbie like that who question's Nate's ability to call American elections?

[Chichiri]

15. The Socratic thingy. nt

[leftynyc]

17. You have the patience of a saint

and my sincere admiration.

[Matariki]

25. He's often wrong

Oklahoma was wrong.

[Chichiri]

26. By one percent.

He had Hillary at a 51% chance to win. Isn't that what some people around here call a "virtual tie?"

No such uncertainty exists in any of today's contests.

[Matariki]

33. Whatever. For all his accuracy he might has well be pulling names out of a hat.

[Chichiri]

35. *facepalm* nt

[Deny and Shred]

53. Such uncertainy DOES exist.

Nate had Clinton winning Michigan by 22.6% - Clinton 60.1% vs Sanders 37.5%.

As I type, it is Sanders 51.0% and Clinton 47.1%

Nate Silver is as fallible as the rest. He has a formula, and it is only so accurate.

You were so convinced. How do you explain this one?

[DemocratSinceBirth]

16. A lot of people got the UK elections wrong; the shy Tory theory ?

I'd like to see where he got the Canada election wrong and the Scottish referendum wrong...

Thank you in advance.

[MineralMan]

18. You're using foreign election results?

He's done awfully well in this years primary election forecasts. Have you looked at those.

I think he's going to be proven right over and over again this year in US elections.

[SidDithers]

21. A toast to Bernie for winning 3 states one the weekend...

Sid

[zappaman]

24. Nailed it!

[PyaarRevolution]

6. He was wrong about KS.

For one and used the excuse we're bigger Liberals here. Give me a break. I saw TYT cover it but apparently most on the DU board ignored it.

[Chichiri]

9. 538 didn't make a prediction for Kansas.

I don't know if Nate made an off the cuff remark or something, but there simply wasn't enough data for Kansas to be called. So far, every state that has been called has been called correctly.

[Beacool]

23. The 538 didn't make a prediction about the last three caucus states (NE, KS & ME).

[SidDithers]

19. Welcome to DU...

Sid

[Thinkingabout]

4. Thanks for the post, every primary and state is important.

[Ellipsis]

5. Likely voters?

[Chichiri]

11. It's not a poll, it's a projection.

It's obtained by taking all the relevant data, including all reputable polls, and applying their own model. They haven't predicted every state, but so far all their predictions have been reasonably accurate or better.

[Ellipsis]

12. Ah... I think it will be closer. Thanks. We'll find out soon enough.

[Chichiri]

14. It could be closer.

It could also be wider.

(Those two words have no business being antonyms . . .)

[Lucinda]

20. Thank you!

[auntpurl]

22. Oh, those are very good numbers.

I can see it's going to be another late night for me in the UK, waiting for results! Damn my addiction to electoral politics!

[NurseJackie]

27. Great numbers! It makes me feel like ...

[Attorney in Texas]

28. Hopefully Nate will be as accurate as his forecasts that Clinton and Trump would both win Oklahoma,

his forecast that Trump would win Iowa, his forecast that Trump would win Kansas, his forecast that Trump would win Maine, etc.

Nate is on a hot streak! He's as "infallible" just like Hillary is "inevitable"!

[Chichiri]

29. Hillary was at a 51% chance to win Oklahoma. Isn't that a "virtual tie?"

And I have no idea what he's doing on the GOP side, but I know that races with more than 2 candidates are MUCH harder to model.

[Attorney in Texas]

30. It would have been if she didn't lose by over 10%

[Chichiri]

32. Rather than spend time arguing the nuances of Oklahoma, I'll just ask . . .

Do you have any specific reason to believe that Hillary and/or Bernie will fall outside the confidence intervals in tonight's states? Or are you pinning your hopes on a general principle of he's-been-wrong-before?

[ismnotwasm]

31. Fingers crossed!

[SheenaR]

34. I'll play along

My models which rely on my opinion say there is a 0% chance that this is the final result.

Keep patting yourselves on the back though.

I would be monumentally stunned, having spend the last few weeks on the phone with Michiganders if a blowout of this proportion happened.

Good luck this evening

[bobthedrummer]

36. Welcome to DU SheenaR!

[Bjorn Against]

37. I thought this thread needed a kick so we can compare 538's prediction to the results

Looking at the returns so far I get the feeling that Nate Silver is not going to help his credibility with this prediction.

[ebayfool]

40. TY - I was looking for this! n/t

[Cheese Sandwich]

38. Wow 538 is good at math

I should have learned good in school like 538

[Avalux]

39. Was 538 drunk when they did their calculation? n/t

[cherokeeprogressive]

41. n/t

[Chichiri]

42. Message from Nate Silver

If Sanders winds up winning in Michigan, in fact, it will count as among the greatest polling errors in primary history. Clinton led by 21.3 percentage points in our final Michigan polling average. Previously, the candidate with the largest lead to lose a state in our database of well-polled primaries and caucuses was Walter Mondale, who led in New Hampshire by 17.1 percentage points but lost to Gary Hart in 1984.

From http://fivethirtyeight.com/live-blog/michigan-mississippi-idaho-hawaii-primaries-presidential-election-2016/

Whatever's happening in Michigan tonight, it's literally a once-in-a-generation event. It's also systemic -- everyone is wrong, and probably for the same reason. To those who question 538's competence, though, I'd simply point out that they called Mississippi pretty much dead-on, and with a lot less data.

[artislife]

49. Heck, I could have called Mississippi with a lot less data. nt

[Kalidurga]

52. You mean like no data except knowing that it's a deep red state?

[Ken Burch]

43. ...

http://www.democraticunderground.com/12511448469

[jillan]

44. Nom Nom Nom Nate!

[DemRace]

45. When was this projection made?

[Chichiri]

47. It was finalized only a few hours ago.

[DemRace]

51. Thanks!

[tabasco]

46. LMAO!

It's funny to watch people tap dance to explain their gigantic fuck ups.

[cherokeeprogressive]

48. n/t

[Motown_Johnny]

50. Bwahahahaha !!!!! n/t

[bunnies]

54. I can't wait to hear your story tomorrow.

Really. I mean it.

[bunnies]

55. Kick and motherfucking rec! Bernie wins Michigan!

Fuck yeah!

Please tell me about your confidence intervals again. It's riveting.

[SMC22307]

59. Now, now, don't taunt.

On second thought, do!

Another kick and motherfucking rec!

[bunnies]

61. Hell yeah!

[SMC22307]

63. "Optics problem," my tight white ass!

[John Poet]

56. Big Fat OOPS!

Michigan doesn't like being told how they will vote.

[ozone_man]

57. That is unacceptable for any statistician

His credibility is gone.

[coyote]

58. This just in from the Hillary camp:

https://m.

My first snark of the evening and I feel so much better.

[SMC22307]

60. It's healthier to let it out.

Snark away!

[kstewart33]

62. Congrats Bernie supporters.

Geez did Silver miss that one. Along with just about every other pollster.

[SMC22307]

65. Thanks. And yes, he, and they did. (n/t)

[phleshdef]

64. Bernie rolled 20.

[TSIAS]

66. Nelson Muntz says

http://photobucket.com/images/nelson%20muntz%20ha%20ha