Nate Silver on Rachel Maddow

[mgcgulfcoast]

right now he says its time for trepidation for democrats. i cant remember his exact words but he said if the next round of swing state polls show the romney surge is continuing then we are looking at a tied race.

[TroyD]

1. That's what I've been saying all day

There's been far too much overconfidence lately.

This is not over.

Too many people have been saying "this is just a Romney blip", "he's going down" etc.

Stop taking things for granted. Obama has gone down in support. That is the reality. let's stop pretending it's not true and do something to change that.

[mgcgulfcoast]

2. you make alot of sense

the debate was a monumental disaster. a "self inflicted" wound.

[courseofhistory]

4. It wasn't

a self inflicted wound, except that a lot of people believe that unfortunately. This country has a definite proclivity for jumping on band wagons based on the lousy media hyperventilating about something that didn't happen. But it may as well have based on what people gobble up as fact.

[blueknight]

35. you are right

it wasnt a self inflicted would, it was a self inflicted attempted suicide. the pres could not have done worse, at a worse time. to think it wasnt disastrous, is to bury your head in the sand.

[adigal]

29. I have trouble wrapping my head around the fact that 90 minuTes

Of rude, hyper behavior changed so many minds. Do people not remember the last republican presidency?

[Mutiny In Heaven]

34. Recalls 2010.

Polls were dismissed, people tried to put a slant on everything and - shock! - turned out that, by and large, the polls were pretty much right. As it stands, a big problem for Obama is the fact that the polls showing Romney take the lead are getting much more coverage. It's a simple fact that millions of people will change vote according to the narrative; that's momentum, and a winning candidate invariably possesses it.

Although Obama rebounded in the days following the jobs report, the fact that the polls so far this week are showing Romney breaking through is highly significant. It makes him look like a winner, the man, the alpha candidate riding the crest of a wave. In a direct comparison, Obama looked incredibly meek, weak and timid, and the polls only serve to re-enforce that in the eyes of the flimsy, floppy (and weak!) "undecided" voters.

The "fight-back" has been weak, and that's what concerns me most. Still no side-by-side, split screen ads of Romney saying one thing then the other, instead relying on on-screen text and talking heads. I think there's one, with the now faded "47%" thing, but that's not even an explicit contradiction.

[flamingdem]

44. One thing that changes "that" is PERCEPTION .. the pundits and Nate aren't helping

They need to give the POTUS a damn break

[cilla4progress]

51. How did this change so fast?

It's unreal ... like a nightmare.

One week ago tonight it was in the bag in a big way.

I knew we were getting overconfident, but this?

[TroyD]

52. 'How did this change so fast?'

Well, according to what Nate wrote, "One night in Denver undid most of the advantage Obama appeared to gain in Sept".

Btw, Nate is just doing his job. We shouldn't attack him just because we don't like the fact that he is pointing out the poll drop for Obama. We will end up like the right-wing.

To give you an idea of what Nate deals with, he is being attacked as an "Obama Propagandist" on Brietbart right now for saying Obama still has a 70-30 advantage. I won't link to a site like that, but I saw the headline when I was searching for Nate's name and it came up.

[regnaD kciN]

3. So, less than a day after saying people shouldn't stress out over polls...

...and consult his site for a less volatile view, he's on the verge of calling the race a tie???

The world has truly gone mad...

[TroyD]

5. No, it's just being sensible

Some of us have been saying for the past week not to get overconfident, and because we did, we got slammed by Romney.

Now we have to fight back or things could decline.

[regnaD kciN]

11. If it's "sensible" to say the race is almost tied...

...then how could it have been sensible to write, as he did just past midnight today, that the fundamentals still favored the President, and people should stop obsessing about polls for the time being? It doesn't seem like any dramatic developments over the last day should have changed that.

[mgcgulfcoast]

12. im just saying what he said on her show.

he was very specific about the next round of NBC/marist swing state polls coming out. if the romney surge doesnt stop he said we are looking at a tied race going into the last few weeks of the race.

[fugop]

15. Because yesterday it was on his blog

Today he's on tv, so RATINGS! Must have RATINGS!!

[fugop]

19. RATINGS!

Gotta have 'em. But yes, it makes no sense. We're getting polls on both sides (Obama down in Colo; no, up in Colo! He's down by 1 in Ohio; no, he's up by 4 in Ohio!)

The whole polling thing is out of control. Shame on Nate for ignoring his own stinking advice to ignore the polls until they settle down.

[flamingdem]

45. I call BS on Silver, he's tempted by power just like Ed, Rachel, Tweety

more than by giving Potus a break

70% vs. 30 is not tied btw

[GreenPartyVoter]

55. The popular vote is much, much closer. Whatever that is worth.

[ailsagirl]

6. I'm getting more confused all the time

I thought that the President had an 85% chance of winning (or somewhere around that), now suddenly it's time for trepidation??

How can NS be right when there are threads like this??

http://www.democraticunderground.com/1251128777

[mgcgulfcoast]

7. nate says its 70-30 right now

and if the next round of swing state polls show the romney surge is continuing it will be a tied race.

[regnaD kciN]

8. 71.2% now...

But I fail to see how 71.2% is one round of swing-state polls away from 50%.

[mgcgulfcoast]

9. nate may be positioning himself to bring this to 50/50

lets hope the polls go the other way fast.

[ailsagirl]

10. I didn't realize how "volatile" his predictions are

I really need to pick and choose my sources for the "state of the race." The eclectic approach hasn't been working so well.

[gkhouston]

13. They aren't. He's saying that *if* the state polls change, that would indicate a genuine shift

in the race. He's not stating that it will happen, merely that it's possible.

[ailsagirl]

16. I guess I don't understand statistics very well...

But I think, for the sake of my sanity (!), I might well avoid the NS stuff. At least for now.

[gkhouston]

26. Imagine you're at a party, trapped in the corner of the kitchen by a nerd.

You say something innocuous like, "How's it going?" You're expecting a response like, "Fine." or "Lousy. My car's in the shop again." or even "What?" if it's a noisy party. Instead, you get a detailed, possibly rambling response that encompasses not only how the nerd is at this moment, but how he/she would be if several events occurred in the near future. Some of them are fairly likely (like you suddenly spying someone in the living room you simply must talk to right away) while others are less likely (you saying, "That's fascinating! Tell me more.&quot and still others are somewhere in between (the party breaking up around 1 after the stereo has blown a fuse for the second time). That's what Silver's statement was about: he was just trying to cover all the bases.

Since it's very odd for the polls to shift so markedly at this point in the race, I'd say it's possible but not likely that the recent polls represent a true shift in the state of the race. Have a beer and stay out of the kitchen.

[ailsagirl]

28. Time will tell--four weeks from tonight

[RobinA]

57. It's Not Statistics

It's probablity based on past performance. Past performance is an OK indicator until it's not. The probability quoted is the probablity AT THAT MOMENT.

Plus, ya gotta realize, with a race this far out and fairly close, little things are going to cause big changes in numbers. Especially when swings occur based on the last big event. Remember when Poppy "won" the Gulf War? He was a shoo-in for reelection until he wasn't. Obama killed Osama? Guaranteed win in the election. Anybody remember that now? Nope, he laid an egg in the last debate. If he leaps buildings in a single bound in the next debate he will again be golden...until he mumbles through the third debate, at which point he will again be sweating a drop in polls.

[FBaggins]

14. Because his percentages are really driven by a handful of state races.

If Ohio and Virginia start to lean toward Romney, the race does too.

[mgcgulfcoast]

17. i wonder if nate has inside info on these upcoming polls.

[DCBob]

18. Here are his "exact words" from his blog..

"The forecast model is not quite ready to jump on board with the notion that the race has become a literal toss-up; Mr. Romney will need to maintain his bounce for a few more days, or extend it into high-quality polls of swing states, before we can be surer about that."

http://fivethirtyeight.blogs.nytimes.com/

Dont fret it.. Romney wont maintain his bounce for a "few more days".

[TardisBlue]

25. Exactly! It's ONE cycle of polling.

We need less hand-wringing over the polls and a lot more focus on the issues.

[wisteria]

30. You have been right before, I hope you are right now.

I will do all I can to get President Obama relected no matter what the polls say, but it is a shame that all the good he has done and the lead in the polls has all been undone with on debate. But one mistep is all it takes.

[TroyD]

31. 'Romney wont maintain his bounce for a "few more days" '

And you know this because ?

[demgrrrll]

50. Sadly, Troy he will not.

Just messing with ya kid...

[DCBob]

59. Check the latest polls...

and get back to me.

[jenmito]

20. No, he didn't say it's time for trepidation for democrats. He said IF the next round of swing

state polls show Romney's surge continuing, THEN it's time for Dems. to start fearing it's a tied race.

[helpisontheway]

21. Now Nate is starting to piss me off

If it is a tied race then put the numbers at 50/50 and be done with it. He kept saying don't freak out. Romney's numbers are back to per-debate numbers. Now he is saying that Obama is likely to lose due to no enthusiasm among democrats. He did not say that yesterday . Guess I will get used to President Romney. I hope the Dems that stay home or those that turned their backs on Obama are the ones that suffer the most under Romney policies.

[wisteria]

32. Don't give up so easily. Democrats are enthusiastic.

You just aren't hearing about that from the media.

[SingleSeatBiggerMeat]

22. Play With Any Map

And if Obama can win Ohio (where he has been up consistently for a month) - he will win, period.

Depending on how the next debate goes, he might be moving there soon until November 6th.

[Tutonic]

23. Not What Nate said

Is DU synonymous for digging underground now? Go to MSNBC TV and listen carefully. He does not think the trajectory of the race has changed. Over the next few days polls may tighten. Romney got a bounce similar to Obama's Convention bounce. Race is 70/30. When this race concludes Rachel and Ed need to be taken to the wood shed--and left there. MSNBC gold has turned them into fools. And I do recall that Rachel cast aspersions on Obama during the 2008 race. Recommend that you look at the Princeton Election site for honest information. FORWARD, FORWARD, FORWARD.

[geek tragedy]

24. Silver has said his model can't capture event-driven changes adequately.

Quite clearly, what's happened to the race is a result of the debate. But, his model treats it as noise until it's quite sustained and it's become obvious that the race has been turned upside down.

[ShadowLiberal]

37. Not quite true, some events will have polling delays, others will effect it right away

The 'Now Cast' of Nates focuses on ONLY polls.

The normal forecast takes other factors into thought, like the economy, to try to determine how polls will move by election day. The economic data and such gets weaker and weaker and has less of an effect on the numbers the closer to the election day. Right now Obama still does somewhat better in the Now Cast because of the bad economic numbers.

While Nate's method is only as fast as polls coming in to react to stuff like debates, big changes to the economy (like the collapse of some big banks in 2008 in September) are factored in quick. According to Nate, McCain dropped 25 points in Nate's forecast within 48 hours of the bank collapses starting in 2008, because the model knew that McCain's goose was cooked from the suddenly horrible economy with McCain's party in the white house.

[aletier_v]

27. What good is a poll that changes faster than Romney's position?

I mean, damn, I could say "95% chance of Obama" win, then say "50/50 chance" tomorrow.

What's the point of the thing if it's not accurate?

[ProudToBeBlueInRhody]

33. You don't seem to grasp the concept of his model

All he does is show a snapshot of the election RIGHT NOW. The polls have been quickly going the wrong way since last Wednesday. That changes the picture.

[aletier_v]

38. The model is worthless if I could do better by guessing

Heck, who is to say that I have a "sekrit model" too?

Look, look, 98% chance of Obama winning.

Ooops, no, now my "sekrit model" says 98% of Romney winning.

Ooops, model now detecting 50/50 chance of either.


Nate is starting to sound like snakeoil.

[ProudToBeBlueInRhody]

39. The 98% cast was a week ago.

Exactly, I believe.

Would you agree the polls have changed in some way, shape or form since then?

[Drunken Irishman]

46. I think that's the point. What use is the model then if it's telling us what we already know?

We already knew, a week ago, Obama was the clear favorite. It didn't take Nate's scientific model to tell us how many states he was actually up in and it doesn't take Nate's model to tell us the polls have shifted toward Romney. This is obvious to anyone who can read a poll or RCP's averages.

So, if Nate is conceding there is literally no point to his numbers except to tell us what we already know, I guess I don't see why so many people put their faith in his numbers and what he says when what he says is almost no different than what the polls are saying. It just seems useless to me.

[ProudToBeBlueInRhody]

48. It puts a general percentage on it

It's an easy number to digest.

Look, I certainly heard no complaints when Obama got a huge bounce in the polls and it was up to 98% Now Cast/86% November 6th, why do you suddenly hate it now?

Never mind that 70-some percent is still pretty good....but obviously if suddenly swing state polls come in and Obama's behind, what do you expect?

[Drunken Irishman]

49. That's the point...

You're making my point. I've always been skeptical of Nate for this reason. His numbers essentially mirror the political mood. They're not concrete ... they're not a prediction ... they're nothing more than just telling us what we already know. A week ago, everyone knew Obama was winning. A week later, we all know he's not doing nearly as well. We don't need Nate to tell us this, right? We can see it for ourselves. All you have to do is look at RCP's average and do your own electoral map.

My point, and I'm assuming the other poster's, is that Silver's numbers are just a more wonkish look at what we know. Everyone knew Obama was leading in Ohio ... we didn't need Silver to tell us this. Just go read a poll and you'd see he was leading. Everyone knew, last week, Obama was leading in the polls ... we didn't need Silver to tell us this. So, what is his use except to tell us what we already know? I think some people are disillusioned by Silver and expect him to make a predictive model. He's not predicting shit. He's just telling us the mood of the country at the moment and we all know the mood. I guess on election day it will be helpful if the polls are tight to have a percentage to it ... but even that could set up to Silver proclaiming a 50-50 election, or such a narrow percentage that he outright states you shouldn't take any meaning with it.

Basically, everyone looks to Silver to tell us the state of the race as if he's inclined to have some information that we don't. Well, I think this proves his model is no different than, really, the polls. So, what's the point? If it's all solely defined by polls ... why not just look at polls?

[blueknight]

36. im old

have seen these elections for decades. the pres dropped the ball BIG TIME during the debate. and from what i see and feel, i dont like how this is heading

[Butterbean]

40. Okay, but what about how Reagan did so poorly in the first debate with Mondale

and Mondale still lost in a landslide? I mean, isn't it possible we've got a situation like that going on again? Call me an eternal optimist...

[ProudToBeBlueInRhody]

41. Reagan had a huge margin for error

A muffed punt means a lot less when you're up by three touchdowns instead of one.

[Butterbean]

42. Do you know what the poll figures were like back then? Just curious.

I've googled but can't find any stats, did they keep track of stuff like that so neurotically back then (I was 11, sorry...)?

[ProudToBeBlueInRhody]

43. I think the general wisdom was Reagan had a double digit lead cut to the upper single digits....

....after the first debate. He recovered the next one and the rout was on.

And I was 8. LOL.

[Butterbean]

47. Whippersnapper. ;)

[Cattledog]

53. Reagan had an 18pt lead going into the 1st 2004 debate.

He dropped to a 9 pt lead after.

[TroyD]

54. You mean 1980? n/t

[dennis4868]

56. He meant

1984

[Butterbean]

58. HOLY CRAP. That's a massive lead.