Pre-Convention, four predictive models on November

[Godhumor]

Princeton Election Consortium (Sam Wang): Clinton 80%
New York Times (Nate Cohn): Clinton 74%
PredictWise (David Rothschild): Clinton 70%
538 (Nate Silver): Clinton 59%

Quite the range, but all have Clinton ahead, 3 by good margins, after FBI week and the RNC. Not bad at all.

Numbers from https://twitter.com/williamjordann/status/756694528627736576

[Doctor Jack]

1. I'm not sure whats up with 538

It seems like she keeps going down, despite what the polls say. A poll will come out with Clinton +7, up from +5, and at 538 she slips from 65% to 63%. Then no polls the following day, Clinton slips from 63% to 61%. Then a poll showing her up by 6 in Florida, and she slips to 59%. It makes no sense to me where these numbers are coming from. Clinton does great in a poll, she falls at 538, she is the same week to week in a poll, she slips, she drops in a poll, she falls at 538, absolutely no polls come out, she falls. I'm not saying they are being dishonest or trying to tip the scale. I think Nate Silver is an honorable and neutral analyst. Its just very confusing and I think they need to spend some time to tell us what is going on over there because it seems that she can do no right on 538.

[ProudToBeBlueInRhody]

3. To be fair, he has maintained Hillary is right now around or ahead of where Obama....

....was in both 2008 and 2012.

I'm not sure the last few days, however....

[Doctor Jack]

5. I guess we will see where things are in a week or two

I predict she will take off like a rocket in his forecast. The RNC was a disaster, the DNC will be amazing, and we have Big Daddy Kaine on our ticket now. The future is looking bright.

[ProudToBeBlueInRhody]

6. Also, we are still too far out from the election

He points out that mild swings one way or another mean more right now than they will later as people settle into their pick.

I think the debates is where the polls start to solidify our way. And that will be good for us.

[Doctor Jack]

7. I think the polls will solidify after this upcoming week

The dems just need one massive push for unity and positivity and this election will be over. Hillary, Sanders, Obama, Kaine, Biden, Warren, and Bill all coming out in a united front will be too powerful to ignore.

[Hortensis]

17. Yes, as usual in most years after typical midsummer

numbers weirdnesses. But all those predictions are already obsolete.

The story of the many connections between Trump and members of his campaign and Putin should be a big change factor.

Then there's the little issue of whether Putin was able to manipulate the Republican Party into inserting a Russia-helpful plank in its platform.

[OnDoutside]

9. I agree. I think she will pull away, and win well. Trump doesn't have a ground game and it will show

as we go on. My biggest fear is that the RNC will orchestrate a previously unseen campaign of voter obstruction.

[CrowCityDem]

10. It looks like Nate is taking away extra points from her because of some nebulous 'trend'.

That's what I get from looking at his Florida assessment right now. He takes the numbers where she's ahead, and for some reason assumes that the contracting from Trump's convention is going to continue as a trend, so he sucks away another couple of points to tie things up.

He's supposed to be about analyzing the data, but now he's creating his own. It's bullshit, and deserves to be called out as such.

[ProudToBeBlueInRhody]

13. Nate tries to tie news events into polls to show where it's trending

Dr Wang has always criticized him for doing so, believing they are "baked into the cake". It's not a new phenomenon for Silver.

[CrowCityDem]

16. I'm criticizing him because he's guessing what the effect is, with no evidence.

Look at FL. In his analysis, despite Hillary being ahead in every metric, he uses the cop-out "trend line" to declare it a tie. There is absolutely no evidence to show Trump tied, other than an assumption that the slightly tighter polls the last week or so will continue moving Trump's way.

[John Poet]

14. Nonsense, Nate is BRILLIANT, he's like THE GOD of poll analysis!

You folks told us that over and over for the past year. So is he going under the bus now, because he doesn't show Hillary's chances quite as high as some others do?


LMAO!!!

[CrowCityDem]

15. He's great when all he does is analyze data. He's now creating it.

[MohRokTah]

8. Silver's model is only predictive in the moment. It's a snapshot if you will.

The others are not and are predictive of November. Something pretty extreme will have to happen to move them very much.

[sweetloukillbot]

11. I wonder if he's overstating Trump after failing miserably with him in the primaries

[StevieM]

2. But we have never before had to deal with dishonesty at the level of what James Comey did.

FBI Director Comey should have resigned after he held that press conference. No American had ever before been subjected to that and it turned out that some of the things he said--important things--were dead wrong. He should be out.

[csziggy]

12. I have to disagree - Comey should have resigned BEFORE that press conference

If he wanted to make statements outside his scope he should have quit and done it as a private citizen.

[Dawson Leery]

4. kick

[anotherproletariat]

18. I'm curious how much they take newly registered voters into account.

One of the reasons the Sanders campaign was often surprised at the outcome of certain primaries, is because they mistook new voter registration as support for them, when many times it was for Trump.