2016 Postmortem
Related: About this forumGimme a break: Tweety says he never saw this last minute shift to Obama coming
Yeah, that's because it wasn't a last minute shift. There was a steady flow of support from Romney to Obama from the moment of the Biden debate and it was almost linear. It didn't accelerate at the end.
This is just more beltway conventional wisdom to try to reconcile their ridiculous narratives.
It wasn't "neck and neck". It never was. If the election would have been held the day after the first debate, maybe that would have been neck-and-neck, but that is the only point in the entire campaign that it was close.
That guy is only about 20% as clever as he thinks he is.
OhZone
(3,212 posts)Obama was pretty much ahead by a point or two for months.
Oh well.
NRaleighLiberal
(60,024 posts)seabeyond
(110,159 posts)he never lost it. that was created by the repugs, media and unskewed polls and just flat out bad polls.
obama never lost the vote.
OldDem2012
(3,526 posts)Anthony McCarthy
(507 posts)he was an ass all during the Clinton administration and the Bush II regime. He's a silly ass.
TheZug
(966 posts)He's always been something of a clown, but was occasionally insightful. He fell off the wagon or something.
NewJeffCT
(56,829 posts)If he hears during the day that there was a last minute shift to Obama, that's what he'll say on his show that night... If tomorrow, somebody says the opposite, he'll then say that.
zipplewrath
(16,646 posts)He should have Nate on his Sunday Morning show to explain to him how to track election polls.
BlueStreak
(8,377 posts)Now, I think Nate is being overly modest because he has built up a rather elaborate model. But the concept is not complicated.
1) Use as many legitimate polls as you can because that gives the most meaningful sample size and negates biases related to particular question or questioner.
2) Average them together, but weight the polls based on proven historical reliability of the pollster
3) Integrate some additional information that correlates well with election outcomes (consumer confidence, unemployment rate, etc)
4) Calculate probabilities of winning individual states
5) Finally use those state-level probabilities to calculate overall probabilities of an electoral win by considering all possible paths to 270.
The concept is straightforward. The execution is more difficult, but 538 and Princeton have both done that.
Of course, this all assumes that one understands the concept of the electoral college, and it is manifestly clear that few of the on-air personalities actually understand that because down the stretch they kept obsessing about how close the national totals were. And indeed, the popular vote was pretty close -- almost exactly what Silver predicted. Close but completely irrelevant.
darkangel218
(13,985 posts)He can really get his audience fired up. And I loved his GOTV last night.
Adenoid_Hynkel
(14,093 posts)I think you saw Dems get worried and enthusiasm dropped, but I don't think anyone really turned on the president - which is why it was so easy for him to make up for it in debates 2 and 3.
Tarheel_Dem
(31,243 posts)ladjf
(17,320 posts)with what he has. It might be time for MSNBC to find a replacement. nt
budkin
(6,721 posts)He didn't support Romney at all. I don't understand
BlueStreak
(8,377 posts)ashling
(25,771 posts)I thought it was for his tirade about how disappointed he was that Obama didn't effusively thank/praise Bill Clinton. He went on and on about that and Rachel even tried to calm him down and finally had to interrupt him to go to something else.
It was like he was still pissed about the first debate.