Nate Silver: For Democrats, Losing the House Is Not Inevitable. (Just Very Likely.)

[babylonsister]

For Democrats, Losing the House Is Not Inevitable. (Just Very Likely.)

http://fivethirtyeight.blogs.nytimes.com/2010/10/27/for-democrats-losing-the-house-is-not-inevitable-just-very-likely/?ref=politics

big snip//

Our projection says that Republicans are favorites in 231 House races, which would reflect a net gain of 52 seats.

But suppose that our forecast is biased against the Democrats by one point across the country as a whole, perhaps because pollsters are overestimating the enthusiasm gap very slightly. Just one point. Well, there are 6 seats in which we have the Republican candidate projected to win by less than 1 full point (it might be a very long election night, by the way). If Democrats hold those 6 seats, the projected Republican gains would be down to 46.

Now suppose that the forecast understates Democratic support by 2 points. There are 8 seats in which we project the Republican candidate to win by a margin of between 1 and 2 points; now these would also be wiped off the board. Now the Republican gains would be reduced to just 38 seats — and the Democrats would hold the House, 218-217!

Read that again: it means that if our forecasts turn out to be biased against Democrats by just 2 points overall, the party becomes about an even-money bet to hold the House.

more...

http://fivethirtyeight.blogs.nytimes.com/2010/10/27/for-democrats-losing-the-house-is-not-inevitable-just-very-likely/?ref=politics

[Beetwasher]

1. Garbage In Garbage Out

Rasmussen and Gallup who FLOOD the field with their shitty polls can have a significant impact in skewing the aggregates.

How much? I dunno, we'll see, but I'm inclined to believe there is a bias.

Nice to see Nate is hedging his bets this way a week out.

[karynnj]

5. Even without casting aspertions, we can say that the polls already

built in exceptional turnout for the Republicans and mediocre turnout for the Democrats. If the polling is otherwise perfect (and we know it can't be - for many reasons - especially because there is an ever increasing non response rate as more move to caller id and don't answer. Some also exclude cell phones - even though those calling both say the former is more Democratic than the later) in terms of turnout, we could easily beat expectations - and they might have trouble making their expectations.

What this all says is that GOTV is the best thing we can do - this could cause a higher turnout and invalidate the LV screen.

Note this is assuming the LV and basic model were perfect to begin with.

[kentuck]

2. Everybody trying to cover their asses before election...

Just in case...

[stopbush]

3. And if your projections are off by 8-10 points because

1. you don't poll people who no longer use a land line, and 2. your percentage of voters who you count as Rs in your models bear no relationship to reality, well, then you could be talking of an election with very little change at all.

[BlueIdaho]

4. WTF Nate?

It looks like even Nate doesn't believe Nate's predictions...

[gkhouston]

18. He's realized for some time that his data is crap. Week by week, the amount

of "weasel words" and uncertainty in his predictions has grown.

I had MSNBC on this morning, out of curiosity. Joe Scarborough and Chuck Todd are pushing "the R's will win" even harder than they were a week ago. It made me think of people clapping to bring back Tinkerbell.

The Dems are going to squeak out a victory, and they fear/know it.

[impik]

6. Take a day off, Nate. Seriously.

[kentuck]

7. What if the "enthusiasm gap" is more than 1 point??

What if it is 3 or 4 points? Does that mean the Democrats would keep the House?

[DarthDem]

8. Nate . . . .

Stop. Please, for your own sake, just stop.

[emulatorloo]

9. Is it me or did the New York Times fuck him up somehow after they hired him?

[Lebam in LA]

10. No

Its not just you.

[vaberella]

15. You aren't alone. n/t

[Ramulux]

11. Hes trying to salvage his career

He is going to be way off come election day and now a week before he is finally admitting that there is a "chance" pollsters overestimated the enthusiasm gap and that just a tiny difference can result in dems keeping the house.

Translation- "Dont blame me for being wrong, blame all the pollsters". The problem with this is that Nate has been one of the biggest apologists for the republican wave meme and has constantly defended all the crap polling over the past few months, so its a little late in the game to try to save your reputation and renounce everything you have said up to this point.

[Aramchek]

20. he'll have an article out the day after the election explaining why he was wrong

but by that point his Golden Boy image will be tarnished beyond repair.

[Tarheel_Dem]

12. Way to cover all the bases, Nate.

:eyes:

[struggle4progress]

13. gotv

[vaberella]

14. Nate...please, just take a damned nap. n/t

[DemocratSinceBirth]

16. I'm Bookmarking This Post

In five days we will know who the Nostradamuses and the Nostradumbasses were.

[Jennicut]

17. In other words, no one really knows how big the "enthusiasm gap" really is. And Nate and others made

their projections on House races where the polling is simply telling us that it is too close to call. One point? That does not leave me with confidence that projections can be made.

[zipplewrath]

21. They have no models

Their problem is that none of them have any models to adjust their data. And the results they are getting suggest that it's going to be a close race alot of places. As Nate is suggesting, very small changes in turn out percentage will have huge impacts in the end.

It's called Chaos theory and there isn't much anyone can do to predict the outcomes because the data with which they have to work isn't accurate enough.

[Aramchek]

19. c'mon, Nate, just fess up and admit that you don't know what's going to happen.

You're making yourself look like an idiot by pretending that you do.

[and-justice-for-all]

22. i think he will be eating those words next week..

and I will remain optimistic.

[pstokely]

23. Would you still distrust him if he was saying different?

either way, GOTV

[and-justice-for-all]

24. I call it like I see it...nt

[county worker]

25. All the pollsters are trying to manufacture a repub win by trying to

keep us demoralized. Yet they all back peddle and say they could be wrong if it doesn't work.

Which says to me nobody knows what's going to happen and the only ones capable of having an effect on the outcome is us! The repubs have peaked.