2016 Postmortem
Related: About this forumNate Silver Dissappoints
http://fivethirtyeight.blogs.nytimes.com/2012/10/05/oct-5-day-after-debate-strong-swing-state-polls-for-romney/
I think Nate is amazing and have the greatest respect for him. That being said, I have to take issue with this analysis. It comes across as almost gushing over the reaction to Romney's performance during the debate and the respective bounce in polling.
A bit of historical perspective shows us that incumbent Presidents rarely perform well in the first debate and media chatter and the bounce in poll numbers post-debate are rarely sustained or relevant to the outcome on election day.
I don't understand why Nate would drag on for weeks after the conventions to aknowledge the bounce that the President got for the DNC as being a "real" gain in the polls and not a temporary bounce but is more than willing to adjust the numbers instantly based on flash polls and knee-jerk polls one day post debate.
How is it that what applied to one case, doesn't apply to the other. Both are instances where emotions and initial reactions could be coloring the outcome more than an actual change is voter tendencies. Would it not be better to gain the perspective of the jobs report, VP debate and maybe even the second Presidential Debate before inferring there has been an real and actual shift by voters toward Romney?
brooklynite
(94,597 posts)Either you accept Silver's methodology or you don't. Complaining about his analysis only when it doesn't conform to your own is ridiculous.
angrychair
(8,702 posts)with accepting or not accepting it. The point of my OP is to understand the method and how 1 whole day of polls moves the Nov 6th outcome and the NowCast and motivates him to infer a shift in voters when it took weeks to determine the same outcome from the DNC. I don't question his honesty only looking for opinions of what his motives might be for such a strong and sudden shift toward Romney two days post-debate.
oswaldactedalone
(3,491 posts)kicked ass in the debate, whether he lied or not, and is getting a bounce. He showed himself to be a formidable foe. Obama needs to be serious about debate prep this time or it might be too late.
brush
(53,787 posts)Seems many pundits fell for the pre-debate meme that Romney would win and they swooned all over themselves jumping on the Romney bandwagon to fulfill the meme that this would be a tight race. And of course high ratings and ad dollars coming from a close race had nothing to do with it. Many, not all, completely lost their objectivity as they blasted the President and gushed at Romney's shape-shifting performance. And that's all it was, a performance. They had to know, as I and the people that I watched the debate with did, that Romney was lying and bullying his ass off all night, yet there was not a word of critique towards the aggressive, rude, moderator bully. If that shallow, crude mendacity is what constitutes a debate victory, we're in sorry shape in the country and our journalist/punditry class. I think The next day when the fact checkers rolled out their info on the multiple lies Romney knowing told, it seems the gushers are now too embarrassed to admit their non-professionalism. I watched it on MSNBC and aside from McDonnell, Sharpton and Bashir, they were all in lockstep with their FOX and CNN counterparts. And that never happens. What's up with that?
Doctor Jack
(3,072 posts)I don't think Nate Silver is intentionally trying to play up Romney and that he doesn't actually believe what he is saying.
However, I don't always like Nate Silver's political commentary. He is very good at statistical analysis but I don't think that necessarily makes him an infallible political god. I remember last summer he wrote a series of articles about how if Rick Perry decided to run in the primaries, he would cruise to victory and be a tough opponent for Obama. Of course nothing close to that happened.
He has pretty good insight but he is wrong too. I don't know if he is wrong now. My gut says he is overstating Romney's position but then again I am not infallible either. Time will tell.
regnaD kciN
(26,044 posts)Generally, 538.com comes out with its info-dump early each evening. Then, the next morning or afternoon, Nate contributes a blog post explaining and interpreting the previous eveing's information. This time, he posted twice in one day -- his usual daily post (concerning the job numbers), and then, after the info-dump, he seemingly rushed out with a new entry which didn't so much explain the day's results as say he didn't think they were reliable, and that Romney was doing better than they indicated.
From the way he wrote about the Ipsos poll, it appears he had the same read on it as I did earlier -- that it reflected what would be an unexpectedly-strong debate bounce for Romney. Previously, Nate had been predicting a 2.2% bounce. I'm wondering if he still thinks that, because the whole tone of his conclusions suggests to me, without coming out and saying it, that he's now sensing the bounce could well be more in the 4%-6% range, which would clearly be enough to thoroughly shake up the previous narrative, and turn the race back into one that either candidate could win.
Laura PourMeADrink
(42,770 posts)he is always super cautious. But, I also can not imagine him skewing anything.
So, must be true. Obama didn't show up, and it's showing in the numbers....just like many of us thought. (that were accused of hand wringing)
regnaD kciN
(26,044 posts)One think noted, that I commented on before, was that, in the separate "insta-poll" taken after the debate, respondents still staked Obama to a 5-point lead. However, by today's tracking result (which was taken starting mid-day the day after the debate), that had shrunk to 2 points. For me, it lays open the question of what caused the numbers to move: the debate itself, or what the media reported afterwards concerning the debate?
bigdarryl
(13,190 posts)angrychair
(8,702 posts)bigdarryl
(13,190 posts)abumbyanyothername
(2,711 posts)Mondale v. Reagan.
Mondale scorched Ronnie in the first debate. And Ronnie did not do much more than battle to a tie thereafter.
Proud Liberal Dem
(24,414 posts)is about the quality of the polls. I'd be much more worried about polling from someplace like PPP, WAPO, etc. My understanding is that these other pollsters lean Republican. Curious that these polls are being touted as something significant when they have traditionally tended to lean Republican during the entire election season.
Laura PourMeADrink
(42,770 posts)back to reality. O blew it? I hope not. And, i hope this is all completely wrong.
bigdarryl
(13,190 posts)fugop
(1,828 posts)Really a coinkydink that you could just about run this exact story but sub in Obama for Bush and Romney for Kerry. A real coinkydink indeed. You'd almost think in reading this that the media knew going in almost exactly what narrative they asre going with, eh???
(Thanks for posting this! It's soothing in a way to see how similar the two elections are playing out. Gobama!
Drunken Irishman
(34,857 posts)Are you saying Obama may have lost the election because of one debate? That might be the most remarkably laughable statement anyone could make. There is absolutely no way Obama is losing this election because of his debate performance. Candidates don't lose or win elections on superficial moments like that. If they did, we'd have a President Kerry, President Gore, President Dukakis and President Mondale.
The debates have historically done little to predict the outcome. People need to chill out. Obama isn't losing this election because of the debate. It's not happening. Even if the unthinkable happens and he does lose ... it won't be because of last Wednesday and the loss would've happened regardless of the debates.
Laura PourMeADrink
(42,770 posts)They needed someone to say "Hang in there - we are doing great with x, y, and z. They didn't hear that.
Think it bothers me that someone has to tell him what to say and how to act.
No worries, all will be well in a couple days (I hope)
Just let us emotional people play it out.
regnaD kciN
(26,044 posts)There is absolutely no way Obama is losing this election because of his debate performance. Candidates don't lose or win elections on superficial moments like that.
Richard Nixon did in 1960.
Drunken Irishman
(34,857 posts)I think that 1960 debate is overstated. People forget there were three other debates and the pre-debate polling put Nixon & Kennedy in a statistical tie ... not the same as Obama-Romney.
Doctor Jack
(3,072 posts)angrychair
(8,702 posts)It also didn't help him either. The jobs report was a gift that keeps on giving though. It blew Rmoney right off the news cycle and left him to fight for space with only positive numbers and news for the President on the front page and lips of every news organization in the country. The wingnut "jobers" didn't do Rmoney any favors either. The Teapublicans come off looking like the batshit crazy glue huffers they are....
hrmjustin
(71,265 posts)courseofhistory
(801 posts)caveats to his numbers an also says that it will take a few more days to see if this holds or dissipates, pretty similar to what he said about the convention bounce.
Lex
(34,108 posts)Nope.
angrychair
(8,702 posts)My point is what would lead Nate to adjust those numbers from north of 90% down to 85%? Doing it based on one whole day of flash and knee-jerk polling that, IMHO, would not actually reflect the real opinion of a voter on Nov 6th. I HAVE NO DOUBT that the President will win on Nov 6th, I am only seeking opinions of what my fellow DU poll watchers think of Nate's shift in his numbers and why he did it.
Drunken Irishman
(34,857 posts)He seems to be pushing a narrative that hasn't unfolded yet. Instead of waiting for things to develop, he's creating the story. The media has a history of doing this ... by taking one or two polls and using it as proof Romney is surging. I've seen it on Yahoo! already today - their top story all day (after the unemployment numbers) was about how Ohio, Florida and Virginia have already shifted to Romney.
Okay. No. First, that We Ask America poll is really flawed. It states it's completely done after the debate ... which means, for it to be released this morning, had to have only taken samples from Thursday. The problem here is that single-day polling is rarely accurate and why most legitimate polling organizations collect data over a two or three day window (some even longer).
The other is Rasmussen, a right-leaning pollster who's had questionable results the entire campaign.
courseofhistory
(801 posts)The highest I saw it was day before yesterday at around 87%.
angrychair
(8,702 posts)I was warping NowCast and Nov6 numbers together in my head. The President's drop for Nov6th
is somewhere around 2.2 points (it was 87.1% on Oct. 4 and 84.9 on Oct.5 that appears to be the biggest one day drop since Aug 20th to Aug.21st when it dropped 2.4%) not sure what was happening then....sorry for confusion
Response to angrychair (Reply #20)
Chan790 This message was self-deleted by its author.
grantcart
(53,061 posts)So its hard to distinguish a genuine shift toward Mr. Romney, from a real but potentially temporary shift based on changes in voter enthusiasm, from an artificial change caused by a bias toward heavy news consumers.
angrychair
(8,702 posts)stood out for me...It wasn't faux news gushing but it was a little more than you see from Nate in an analysis:
"My subjective view is that, despite the somewhat mixed messages that the polls gave about the magnitude of Mr. Romneys bounce, this is still too conservative. The forecast model is pretty smart about distinguishing random movements in the polls from real ones, and so can be fairly conservative in interpreting the data. However, it does not have the advantage of knowing that the shifts may have come for a good reason in this case, Mr. Romneys strong performance in the debate."
grantcart
(53,061 posts)But the key is to understand that he thinks it is a bounce and that it will be temporary unless something else emerges.
Among his other conclusions is that incumbent presidents typically underpoll about 2-3 percent and Obama is probably at this level or a little more.
Even with a bounce it is unlikely that Romney will pick up any electoral votes, except for the possibility of North Carolina except that NC voter registration by Obama campaign has been so pronounced as to undermine the polls.
Tomorrow expect detailed analysis by Silver detailing how a drop in the unemployment rate below 8% has had an impact on Ge results.
angrychair
(8,702 posts)I do hope that is the case. I feel that any more positive feedback for Rmoney is unlikely to happen at this point. Faux will still campaign for him but most other news cycles have already moved on to the jobs report and the batshit crazy "jobers" claiming yet another conspiracy against the TPgop
I didn't know that about NC...that is great to hear
geek tragedy
(68,868 posts)1). Single day polls are questionable
2). No cell phones
3). Owned by Republican operatives
regnaD kciN
(26,044 posts)I think he's right about Ipsos showing a strong move toward Romney. I also think the electoral odds are too optimistic for Obama -- instead of them basically being 85-15 in his favor, I'd place it more at around 75-25 or, at worst, 70-30. Which is still better than they were around mid-summer.
The one thing that might be controversial is Nate saying he'd bet on Romney "given the odds." That doesn't mean he thinks Romney is more likely to win; it means that, given 7:1 odds (which is what 85-15 means), a Romney win would be a decent longshot bet.
fearnobush
(3,960 posts)I perceive it to mean betting on Romney moving the polls in his favor rather than his model displaying statistical variance. Yours makes better sense.
Doctor Jack
(3,072 posts)When i read that I thought he meant that he would bet on Romney winning the election at this point.
I clearly am not someone that gambles. That is a load off.
Denzil_DC
(7,242 posts)I'm not even clear what that statement meant in context. It seemed not to relate to the rest of the column very well.
Was he saying that he wouldn't bet against Romney getting a significant bounce come next Tuesday, or that he wouldn't bet against him winning the election? Given how wildly the numbers would have to swing from the current position for his model to predict that, I find it hard to believe he was saying the latter, but who knows?
Nate needs an editor. If he already has one, he needs a better one.
fearnobush
(3,960 posts)What Nate said and proved with network cross comparison is that the debate result in and of its self really doesn't mean crap. It's how the MSM and all their talking heads proselytize their perceived winner, not the publics perception. If the media actually focused on the issues and who said what, Friday's poll numbers would look better. But, since the MSM only cares about who looked better and who was more convincing at lying his ass off, they ended up hurting our President much more so than the actual debate. If anyone on this board has any influence on what the media dictates, please for the love of humanity and country, get them off their love affair with Rmoney.
Regarding the state pollsters, unfortunately Nate is very correct to say read these polls with an open mind. Last week those same pollsters, republican or not, showed a six point difference in Florida and Virgina. Six points! That's the number that has Nate scratching his head. He also said the republican respondents may have been more likely to do the surveys rather than ignore them due to their increase of enthusiasm, just the same as a dem ignoring the survey because his President failed to deliver resulting in an over all poll bias. Also, none of these polls will have any meaningful sample of today's jobs report. But again, I stress, it's not what the numbers say, it's how the MSM presents them. And based off of what I saw and read today, it was pretty abysmal regarding their immediate assumption that something is wrong with those numbers. That being said, there is still time for that narrative to change. And it likely will because you know Obama will reinforce the narrative of good and steady declines in the unemployment rate. Let's hope that perception occurs fast. Otherwise, the MSM is going to continue to carry Rmoney's water. As far as those state pollster, Ohio was still at least tied. Ras and WeaskAmerica both had Ohio Showing about a four pt loss. Again, like Nate said, we won't know until next week, until the better pollsters with better resources do their surveys. God willing, they show a better result. Obama only needs an average lead next week. Even if it's fractional, no candidate or incumbent has won the electoral college without a national lead exactly one week after the first debate in the last sixty years.
budkin
(6,703 posts)He's fascinated by what makes them go... can't blame him for being so into the ups and downs of the race. But for God's sake look at his predictions. Obama 85% chance of winning. And you're upset?
angrychair
(8,702 posts)but this is the first post I've had with over a 1000 views (as far as I can tell)....pretty awesome. Thanks for stopping and looking and commenting!!
MFM008
(19,816 posts)hes got to keep the gopers reading the NYT. Notice Obamas still killing Romnuts in the numbers.
Godless in Seattle
(120 posts)cemaphonic
(4,138 posts)Basically, he's anticipating a fair amount of volatility, and is overall adopting a wait-and-see approach until he gets a better picture of how the polls shake out over the next week. And he has written some about the historical impact of debates on poll movements.
Cosmocat
(14,565 posts)His numbers are his numbers.
I think he clearly can be trusted to run his numbers in the most accurate way he knows, and that they are going to be pretty darn solid.
But, he DOES, sadly, like most "media" or "analysts" fall into this crape of mitigating positive news for dems and doing what he can to prop up Rs.
I just don't get it, I really don't.
I guess as someone who makes money off of it, he has felt the wrath of conservatives acting like raging three year olds when they get the news they don't like to hear, so he feels like he has to walk on glass with it.
But, his framing absolutely is like most of the rest of the "liberal media."
Won't flat out say the wind is at Obamas back as it clearly was starting at the conventions, like pulling teeth to have him flat say BO is in a good spot, but one overly hyped debate and its GOOD NEWS for ROMNEY.
It is the framing, it is tiresome, and it is something this country will never get away from.
Denzil_DC
(7,242 posts)In 2012, he's part of the MSM and has daily space to fill.
Give Nate Silver enough data and on past record he can land it on a dime. As a political analyst and predictor outside that statistical field, his track record is just so-so, and not immune to peer pressure from other pundits' opinions and conventional wisdom.
His statistical model is among the best we currently have. His speculation about what's going on and how things will pan out at this stage, I take with a pinch of salt, not least because he himself warned us about jumping to the very conclusions his most recent columns have hinted at.
Jeff In Milwaukee
(13,992 posts)Romney's numbers in the Now Cast went from about 1.5% pre-debate to 4.2% as of this (10/6) morning. Yikes! His number almost tripled -- but what it means is that he went from getting his ass kicked and his house burned to the ground and his fields sown with salt to merely getting his ass kicked and his house burned down.
He's the most inept candidate for national office in our generation. His base doesn't like him, and any move he makes to shore up his base alienates the moderate voters he needs to win. Women hate him, Hispanics hate him, Blacks hate him, Students hate him, Seniors hate him.
The Obama Campaign still has work to do, but we've got this. Just GOTV and bring along all your friends.