Some polls now have Romney ahead.

[Balt]

It must be pretty unbearable to be a TV viewer in Ohio, considering how much hinges on it.

The oversaturation has probably made people tune out at this point anyway. Money and effort spent on 'get out and vote' door-to-door type efforts will probably still pay off, though.

 

[Doppel]

Man I would just love it if Obama can somehow hang on and win Florida. Put us all to bed early.

I can't imagine staying up late anyway. There's such a slim chance of Romney winning I fully expect to wake up to his idiotic speech he'll give when he loses.

 

[buckshot24]

I can't imagine staying up late anyway. There's such a slim chance of Romney winning I fully expect to wake up to his idiotic speech he'll give when he loses.

Define slim. Even if your God Nate Silver is right there is a 30% chance that Romney wins. He was way off in 2010 so you shouldn't be too confident in him for 2012.

 

[Theb]

Define slim. Even if your God Nate Silver is right there is a 30% chance that Romney wins. He was way off in 2010 so you shouldn't be too confident in him for 2012.

Bill Clinton is our god. Obama is our messiah. Nate Silver is a minor prophet.

I don't know whether FDR or JFK is the holy ghost As soon as my Obamaphone is charged I'll ask and get back to you.

As to his odds, Obama will win with over 600 EV once all 57 states have reported.

 

[buckshot24]

Bill Clinton is our god. Obama is our messiah. Nate Silver is a minor prophet.

I don't know whether FDR or JFK is the holy ghost As soon as my Obamaphone is charged I'll ask and get back to you.

As to his odds, Obama will win with over 600 EV once all 57 states have reported.

HAHA

Thanks that was good.

 

[Farang]

Define slim. Even if your God Nate Silver is right there is a 30% chance that Romney wins. He was way off in 2010 so you shouldn't be too confident in him for 2012.

He was not way off in 2010. I shouldn't have to explain why, also, it's much different to predict dozens of tight races than it is one giant one. It's apples to oranges.

 

[buckshot24]

He was not way off in 2010. I shouldn't have to explain why, also, it's much different to predict dozens of tight races than it is one giant one. It's apples to oranges.

He predicted 15 fewer seats for the Republicans than they got. It isn't exactly close. Hell, he even had Rasmussen's shitty polls favoring Republicans and he was still off.

 

[uclaLabrat]

Hey, screw you, pal!

yeah, you fuckers don't even put maurice's on your pig!

 

[Farang]

He predicted 15 fewer seats for the Republicans than they got. It isn't exactly close. Hell, he even had Rasmussen's shitty polls favoring Republicans and he was still off.

Maybe I do have to explain.

Predicting dozens of House races with sporadic polling is much different than predicting one major race with mounds of polling.

Second, you're simply wrong. He predicted a 54 seat gain and the Republicans gained 63. Given the data, Silver himself would probably say his accuracy was partly due to luck--because of the lack of data, the range of probable outcomes ranged as far as 21 Republican pickups to 83. Again because you're dealing within dozens of local races as opposed to one large one.

If your point is to prove that Silver has proven himself able to accurately predict a wide variety of elections, from parliamentary to national, then well done!

 

[woolfe9999]

He predicted 15 fewer seats for the Republicans than they got. It isn't exactly close. Hell, he even had Rasmussen's shitty polls favoring Republicans and he was still off.

Where did you get this from?

He predicted 54 seats and the GOP got 63. For the Senate, he predicted 7 seats for the GOP and they got 6.

http://en.wikipedia.org/wiki/Fivethirtyeight#2010_U.S._mid-term_elections

Edit: Farang already said this.

 

[buckshot24]

Maybe I do have to explain.

Predicting dozens of House races with sporadic polling is much different than predicting one major race with mounds of polling.

No shit. I didn't say otherwise.

Second, you're simply wrong. He predicted a 54 seat gain and the Republicans gained 63. Given the data, Silver himself would probably say his accuracy was partly due to luck--because of the lack of data, the range of probable outcomes ranged as far as 21 Republican pickups to 83. Again because you're dealing within dozens of local races as opposed to one large one.

I watched him predict 45-50 in an interview on PBS. The video was from 10-18-2010 if he had a different forecast later then good for him.

edit: You're right it wasn't 15 it was 9 for his final forecast.

 

[buckshot24]

Oh Charles, you thought the presidents approval rating which was basically polar opposite between the likely voter poll was inexplicable. The approval rating is taken from Adults only which makes the disparity a little easier to understand.

 

[randomrogue]

I'm calling it now. Romney lost. I don't see any way he can win other than a complete failure on just about every poll out there.

 

[jackstar7]

edit: You're right it wasn't 15 it was 9 for his final forecast.

So who did better? Who should we be paying more attention to?

Since you want to discredit Silver, do you have a better model to watch?

 

[Charles Kozierok]

Oh Charles, you thought the presidents approval rating which was basically polar opposite between the likely voter poll was inexplicable. The approval rating is taken from Adults only which makes the disparity a little easier to understand.

Thanks. I thought it was RV, which still would have explained the difference.

I think Gallup's LV model is bleh.

 

[Todd33]

Define slim. Even if your God Nate Silver is right there is a 30% chance that Romney wins. He was way off in 2010 so you shouldn't be too confident in him for 2012.

You even linked to it, presidential races are much more accurate than off years congressional ones. You seems to cherry pick the data to give yourself false hope. The path to victory for Romney is very hard and getting worse. Unless there is some swing in momentum Obama is going to get nearly 300EV.

 

[TraumaRN]

After looking at some of the initial polls that were released today, I just think one of the take home polling lessons to be learned from this election is that if you are not polling cell phone voters your results are gonna look....funky. That's my prediction anyways, because honestly I can't make heads or tails from some of these results even the ones with cell phones included.

But safe to say after this election I think you'll see a bigger effort to contact cell phone only voters as they are really very quickly going to become the majority. Landlines are going the way of the dodo.

 

[buckshot24]

You even linked to it, presidential races are much more accurate than off years congressional ones. You seems to cherry pick the data to give yourself false hope. The path to victory for Romney is very hard and getting worse. Unless there is some swing in momentum Obama is going to get nearly 300EV.

Who's cherry picking? You guys wanted to start comparing 2010 vs 2012. My point was it was harder to poll those congressional races accurately all along. You guys were blasting Rasmussen using 2010 when I didn't think you could make it stick in 2012. All i did was the same thing with Silver and then you complained.

Besides there are quite a few polls that are in agreement with Rasmussen's national poll. It's harder to make the bias charge stick.

Also I feel better about Gallup's poll now with a 3 point lead vs a 7 point advantage. We can't really call it an "outlier" now.

 

[buckshot24]

Then just like that Gallup has Romney up 5 again.

 

[Charles Kozierok]

Gallup also shows Obama's approval number at +1. That's a drop of 10 full points in two days?

I'm starting to think Gallup has just jumped the shark entirely.

Regardless of these stupid tracking polls, Obama remains firmly in command in the states.

 

[Todd33]

Then just like that Gallup has Romney up 5 again.

You love them outliers that agree with your bias. funny how that works.

 

[TraumaRN]

Gallup also shows Obama's approval number at +1. That's a drop of 10 full points in two days?

I'm starting to think Gallup has just jumped the shark entirely.

Regardless of these stupid tracking polls, Obama remains firmly in command in the states.

Not to mention when you look at just registered voters under Gallup it is tied, 48-48....I find that much more likely given the circumstances of a tight race.

 

[woolfe9999]

Not to mention when you look at just registered voters under Gallup it is tied, 48-48....I find that much more likely given the circumstances of a tight race.

I have to think that in tight swing states the discrepancy between registered and likely is going to be quite small. It's one thing to not bother voting if you're in say, California or Utah. Another if you're in Ohio or Virginia. I have one friend who lives in Ohio who favors Obama but is relatively apolitical and usually doesn't vote. I've e-mailed him encouraging him to vote because of where he is and he recognizes this and will vote this time. If polled, he would probably rate as unlikely. Another friend now lives in Toronto but is still entitled to vote by absentee in Ohio. Though he hasn't voted in a US election since 2004, he's going to vote this time because of how tight the race is and how pivotal Ohio is.

 

[buckshot24]

You love them outliers that agree with your bias. funny how that works.

Read my previous post. I just got done saying I feel better that Gallup was more inline with other polls then they went out of line.

 

[TraumaRN]

I have to think that in tight swing states the discrepancy between registered and likely is going to be quite small. It's one thing to not bother voting if you're in say, California or Utah. Another if you're in Ohio or Virginia. I have one friend who lives in Ohio who favors Obama but is relatively apolitical and usually doesn't vote. I've e-mailed him encouraging him to vote because of where he is and he recognizes this and will vote this time. If polled, he would probably rate as unlikely. Another friend now lives in Toronto but is still entitled to vote by absentee in Ohio. Though he hasn't voted in a US election since 2004, he's going to vote this time because of how tight the race is and how pivotal Ohio is.

Right. I agree.

And sorry to sound like a broken record but I think in some states you'll see better margins for Obama because of lack of effective polling of cell phone only households. It is to me at least one reason so many polls are all over the place. And its almost like some place pick an arbitrary number. I saw one Republican leaning pollster use 15% cell phone only in North Carolina when the number is more like 25-30% of actual people have only cell phones. Ditto I saw a Democratic poll using 38% for cell phone only.

Not to mention some polling firms either don't show you their complete internals, hide it behind a pay wall or explain their entire polling methodology in 50 words or less. Just poor frigging polling in general.