How you know it's over...liberals already blaming others

[DC Tom]

not true they also blame the homosexual agenda, feminazi agenda, the Hollywood agenda, political correctness, victim mentality and China

 

You forgot the main stream media.

[dayman]

You forgot the main stream media.

 

and no unions blamed?

Edited October 30, 2012 by TheNewBills

[DC Tom]

and no unions blamed?

 

And don't forget to blame the Sikhs.

 

Or Muslims. Same difference.

[OCinBuffalo]

The real question is will certain unnamed posters...in the event of an Obama victory which some places like 538 predict is very likely (pure propaganda if you ask any hardline GOP thumper)...say to themselves "Hmmm...you know when I look back at the things I read and what I thought would happen I realize now what happened...I read what I wanted to hear" or will they just somehow say it was a fraud election...or will they just B word and not think about it too much...what will they do if Obama wins? They can't continue to have faith in their sources that shape and back up their wishes can they?

I don't care if you name me...I am naming me. So...WTF? The fundamental difference is: if I'm wrong, I will go looking for the answer as to why. If you/Nate Silver is wrong...it will be everybody else's fault, and you will just move on. (Moveon.org is good. NoIntrospection.com is better)

 

If Obama wins, it will be because the story he and his team have been pushing about D turnout being so high, that it overcomes both energized R and I +7-10 turnout for Romney, is accurate. It's possible, but not probable. EDIT: So far? Obama is short 250k votes in Ohio from this point in time in 2008. F'ing about with poll weighting...does little against that. Like I said: not probable.

 

But really? The difference between my "sources" and yours? I don't have any.

 

3 months ago, when we were seeing polls of Obama leading by 7-10, I looked at the "internals", and what I now know are called "crosstabs" and said: Horsecrap.

 

I said that, because BI is part of what I do for a living, and these polls are clear examples of what I have learned to avoid, and trained people not to do. Clients often fail, because they go looking to prove a pattern, rather than letting the patterns come to them. I am speaking from wisdom - the result of experience and knowledge over time - here, not partisan buffoonery.

 

Nate Silver smells, and most certainly writes, like a client VP trying to get over on another client VP. I've seen this before. I've seen them spend, literally, $3 million on their own private "I'm Right" system, and have had to decommission those systems more than once.

 

It's not about "what I want to hear". It's about "I've seen people like Nate Silver before. I've seen them do this before. They are rarely right, and they almost always have an ulterior motive. It doesn't mean they are never right. It does mean that their system has little to do with it."

 

Example: How do you explain Silver giving more weight to a poll from 4 weeks ago, rather than one from yesterday, because the older poll is in line with an unsupported belief, while there is a lot of data that contradicts that belief?

 

Obama may win the election...but what about that excuses this bizarre approach to poll weighting?

Edited October 30, 2012 by OCinBuffalo

[B-Large]

I don't care if you name me...I am naming me. So...WTF? The fundamental difference is: if I'm wrong, I will go looking for the answer as to why. If you/Nate Silver is wrong...it will be everybody else's fault, and you will just move on. (Moveon.org is good. NoIntrospection.com is better)

 

If Obama wins, it will be because the story he and his team have been pushing about D turnout being so high, that it overcomes both energized R and I +7-10 turnout for Romney, is accurate. It's possible, but not probable. EDIT: So far? Obama is short 250k votes in Ohio from this point in time in 2008. F'ing about with poll weighting...does little against that. Like I said: not probable.

 

But really? The difference between my "sources" and yours? I don't have any.

 

3 months ago, when we were seeing polls of Obama leading by 7-10, I looked at the "internals", and what I now know are called "crosstabs" and said: Horsecrap.

 

I said that, because BI is part of what I do for a living, and these polls are clear examples of what I have learned to avoid, and trained people not to do. Clients often fail, because they go looking to prove a pattern, rather than letting the patterns come to them. I am speaking from wisdom - the result of experience and knowledge over time - here, not partisan buffoonery.

 

Nate Silver smells, and most certainly writes, like a client VP trying to get over on another client VP. I've seen this before. I've seen them spend, literally, $3 million on their own private "I'm Right" system, and have had to decommission those systems more than once.

 

It's not about "what I want to hear". It's about "I've seen people like Nate Silver before. I've seen them do this before. They are rarely right, and they almost always have an ulterior motive. It doesn't mean they are never right. It does mean that their system has little to do with it."

 

Example: How do you explain Silver giving more weight to a poll from 4 weeks ago, rather than one from yesterday, because the older poll is in line with an unsupported belief, while there is a lot of data that contradicts that belief?

 

Obama may win the election...but what about that excuses this bizarre approach to poll weighting?

 

Do you read is daily articles? He explains much of his methodology and tweaks to the system as the race progesses. I don't know if you read the articles, but perhaps in them you will find your answers.

[fjl2nd]

Do you read is daily articles? He explains much of his methodology and tweaks to the system as the race progesses. I don't know if you read the articles, but perhaps in them you will find your answers.

 

Clearly, he doesn't. Nate Silver goes through exactly how his model works and keeps the reader educated on it.

 

The hate for him now is actually a little funny though. Because they will all look very silly come election day.

[Buftex]

And don't forget to blame the Sikhs.

 

Or Muslims. Same difference.

 

Chicago?

[B-Large]

Clearly, he doesn't. Nate Silver goes through exactly how his model works and keeps the reader educated on it.

 

The hate for him now is actually a little funny though. Because they will all look very silly come election day.

 

I spend alot of time reading the articles, only because I want to learn/ know more about the polling and electoral process. Since OC seems to know quite a bit about polling and weighting and what not, I think he would find Silver's explanations compelling, even if he disagrees...

 

Silver's analysis has been pretty conistent, and has not been biased at least from my view. I do think he works for the Obama campaign as a polling consultant, but I don't see him being a shill for the left.

[IDBillzFan]

Their greatest concern is who to blame

 

And who can write up a quick "You're to blame" jingle that little kids can sing.

Edited October 30, 2012 by LABillzFan

[jjamie12]

I spend alot of time reading the articles, only because I want to learn/ know more about the polling and electoral process. Since OC seems to know quite a bit about polling and weighting and what not, I think he would find Silver's explanations compelling, even if he disagrees...

 

Silver's analysis has been pretty conistent, and has not been biased at least from my view. I do think he works for the Obama campaign as a polling consultant, but I don't see him being a shill for the left.

 

I find it as more of a 'smell-test' than anything else. He can write and reason and tell me all sorts of stuff -- Ultimately it comes down to: Will this election (I suppose mostly in the key battleground states) have a turnout that is D + 5 or 6 or 7 or whatever, OR will it have a much lower D+ turnout? It's hard to conceive that turnout for D's (as it relates to R's) will be as strong as it was in 2008. Most of the polling relies on that 2008 model, where turnout is D+ 6+. If it IS D + 6+ President Obama will win going away. If it isn't, well -- since those D+ 6+ polls show toss-ups -- you're probably looking at a Mitt Romney Presidency.

 

I don't think anyone (intelligent) is saying there's some vast conspiracy -- People (the intelligent ones, at least) are saying that D+6 or more doesn't feel like the right turnout numbers to use. In other words, expecting a 2008 style turnout even when we see more R engagement and less D engagement doesn't seem real compelling from a common sense view. I'm sure people can construct a very detailed story where that appears plausible and even likely. However. Stepping back and looking at this thing from 10,000 feet, it sure doesn't feel like we're going to see a D+6 or more turnout. It's very possible I'm wrong -- ask my wife; I'm wrong all the time -- but from a big picture perspective, it would seem D+6 polls are too high on the D side and, thus, inflate the current numbers for the President.

Edited October 30, 2012 by jjamie12

[dayman]

I find it as more of a 'smell-test' than anything else. He can write and reason and tell me all sorts of stuff -- Ultimately it comes down to: Will this election (I suppose mostly in the key battleground states) have a turnout that is D + 5 or 6 or 7 or whatever, OR will it have a much lower D+ turnout? It's hard to conceive that turnout for D's (as it relates to R's) will be as strong as it was in 2008. Most of the polling relies on that 2008 model, where turnout is D+ 6+. If it IS D + 6+ President Obama will win going away. If it isn't, well -- since those D+ 6+ polls show toss-ups -- you're probably looking at a Mitt Romney Presidency.

 

I don't think anyone (intelligent) is saying there's some vast conspiracy -- People (the intelligent ones, at least) are saying that D+6 or more doesn't feel like the right turnout numbers to use. In other words, expecting a 2008 style turnout even when we see more R engagement and less D engagement doesn't seem real compelling from a common sense view. I'm sure people can construct a very detailed story where that appears plausible and even likely. However. Stepping back and looking at this thing from 10,000 feet, it sure doesn't feel like we're going to see a D+6 or more turnout. It's very possible I'm wrong -- ask my wife; I'm wrong all the time -- but from a big picture perspective, it would seem D+6 polls are too high on the D side and, thus, inflate the current numbers for the President.

 

Well in any event he uses a million polls but ultimately the issue is simple and you don't really have to understand anything to get it.....there are a lot of close battle ground states and Obama needs to win less of them. While Ohio could go Romney, if most people had to bet their saving on it they would choose Obama, which would makes things even worse for Romney. That's it. That's all you need to know.

[OCinBuffalo]

Do you read is daily articles? He explains much of his methodology and tweaks to the system as the race progesses. I don't know if you read the articles, but perhaps in them you will find your answers.

Of course I have. Where do you think I got the the "weighting polls from 4 weeks ago heavier than ones from yesterday" from? Can you explain this?

Clearly, he doesn't. Nate Silver goes through exactly how his model works and keeps the reader educated on it.

 

The hate for him now is actually a little funny though. Because they will all look very silly come election day.

Perhaps you can explain the above?

 

As I said, my issue is with the method. I've pointed out flaws over and over, and every time I do: your response is that of a cheerleader whose team just threw a pick = "It's alright, it's OK, we'll still beat them anyway". Stop cheerleading, and explain why you think this will be a 2008 electorate.

 

Be aware that this Gallup of 9000+ voters, http://www.gallup.co...-like-2008.aspx, and shows current party ID as +1 for Republicans...up 11 frigging points from 2008, will severely hamper you and dopey Silver's explanation.

 

You and Silver are still F'ing about defending D+6-10 turnout polls.

 

The man is lying to himself, and you. Obama may still win, but there's no way that the electorate today = 2008, and that is the FUNDAMENTAL tenet of Silver's work. All I am asking is that we use proper methods, and then see where things stand.

I spend alot of time reading the articles, only because I want to learn/ know more about the polling and electoral process. Since OC seems to know quite a bit about polling and weighting and what not, I think he would find Silver's explanations compelling, even if he disagrees...

 

Silver's analysis has been pretty conistent, and has not been biased at least from my view. I do think he works for the Obama campaign as a polling consultant, but I don't see him being a shill for the left.

Silver has been consistent. Consistently, wrong. As I said, the worst kept secret is that Silver was getting his data from Obama internal polls in 2008. I mean, come on, this board called the race for Obama as soon as McCain did his goofy campaign suspension thing. We could have spent our time working out models to prove ourselves right, but why? The fact is that Sliver got the states picked right(all but one), but how hard is that in a D+8 election?

 

This time...we'll see how accurate he is...when things are much more granular. Given Gallup's poll it's possible that Silver is blown out of the water. As I have said 100 times: the numbers have shown that this is a D+2 election at best. I won't believe it is R+1...until I see more data that backs this up.

 

Nate Silver has been consistent: he has consistently ignored these #s. I wonder what he will say if it is in fact R+1?

I find it as more of a 'smell-test' than anything else. He can write and reason and tell me all sorts of stuff -- Ultimately it comes down to: Will this election (I suppose mostly in the key battleground states) have a turnout that is D + 5 or 6 or 7 or whatever, OR will it have a much lower D+ turnout? It's hard to conceive that turnout for D's (as it relates to R's) will be as strong as it was in 2008. Most of the polling relies on that 2008 model, where turnout is D+ 6+. If it IS D + 6+ President Obama will win going away. If it isn't, well -- since those D+ 6+ polls show toss-ups -- you're probably looking at a Mitt Romney Presidency.

 

I don't think anyone (intelligent) is saying there's some vast conspiracy -- People (the intelligent ones, at least) are saying that D+6 or more doesn't feel like the right turnout numbers to use. In other words, expecting a 2008 style turnout even when we see more R engagement and less D engagement doesn't seem real compelling from a common sense view. I'm sure people can construct a very detailed story where that appears plausible and even likely. However. Stepping back and looking at this thing from 10,000 feet, it sure doesn't feel like we're going to see a D+6 or more turnout. It's very possible I'm wrong -- ask my wife; I'm wrong all the time -- but from a big picture perspective, it would seem D+6 polls are too high on the D side and, thus, inflate the current numbers for the President.

Be careful. You are going to get told that you don't like Nate Silver because he's a Democrat, works for the NYT, and is schill....rather than merely pointing out that there's no way this is a 2008 electorate.

 

Look...in a lot of ways, this is no different than Libya: the Ds have convinced themselves that something is true, in this case that the white turnout will be 68%, not the 74% from 2004/2010, and that Latinos will make up that 6% difference and will vote 80% for Obama. Nothing can now shake that belief, and they will do/say whatever to try and support it.

 

Every fact that contradicts that belief, even 9000+ sample polls, is an "outlier", and thus is weighted down by Silver accordingly. This belief stems from a single book that tells them Demographics == Republicans lose. Even though events have overtaken this book, they are still pushing it as Holy writ.

 

The belief is that Latinos are going to increase 6 pts in the total electorate, when every poll shows them less likely to support Obama than in 2008 , and, that even more white voters will stay home this time, than did for McCain?. Ludicrous.

 

This belief is a matter of faith, not reason. We can't argue against faith, no matter how right we are.

 

This is also a matter of common sense. Ask yourself, when talking about liberals, how often do we talk in terms of common sense?

[OCinBuffalo]

Well in any event he uses a million polls but ultimately the issue is simple and you don't really have to understand anything to get it.....there are a lot of close battle ground states and Obama needs to win less of them. While Ohio could go Romney, if most people had to bet their saving on it they would choose Obama, which would makes things even worse for Romney. That's it. That's all you need to know.

Not anymore. What's it going to take for you to realize that this is, and always was, the "narrative"...and representative of "the data"?

 

This is the story, the belief, that Silver has started with, and THEN found facts to support it. He has not started with an objective model, and then seen what story emerges. He has weighted polls, that have already been weighted...incorrectly.

 

I could see if we were talking about party ID shifts of 3-4% here. In fact I would have dismissed them the same way Silver has. I could see it if the polls had come back in September with Obama leading by 2-3 in the swing states, with a party ID of D+4. But that's not what we saw, was it?

[dayman]

Not anymore. What's it going to take for you to realize that this is, and always was, the "narrative"...and representative of "the data"?

 

This is the story, the belief, that Silver has started with, and THEN found facts to support it. He has not started with an objective model, and then seen what story emerges. He has weighted polls, that have already been weighted...incorrectly.

 

I could see if we were talking about party ID shifts of 3-4% here. In fact I would have dismissed them the same way Silver has. I could see it if the polls had come back in September with Obama leading by 2-3 in the swing states, with a party ID of D+4. But that's not what we saw, was it?

 

Close election, Obama needs to win less swing states. That's it. When you look at 538 that is basically all that is presented. And when you look at the electoral college you see it is true. Also, we'll know in a week. So who cares.

[jjamie12]

Close election, Obama needs to win less swing states. That's it. When you look at 538 that is basically all that is presented. And when you look at the electoral college you see it is true. Also, we'll know in a week. So who cares.

Well in any event he uses a million polls but ultimately the issue is simple and you don't really have to understand anything to get it.....there are a lot of close battle ground states and Obama needs to win less of them. While Ohio could go Romney, if most people had to bet their saving on it they would choose Obama, which would makes things even worse for Romney. That's it. That's all you need to know.

That's not it, and that's NOT all you 'need to know'. 538 aggregates polls and polling data then does simulations based off of that data to try to arrive at the 'most likely' course of events. No one really disputes that. What people have been trying to point out (apparently to no effect) is that the *underlying data* for 538 (the polls it uses to aggregate into its simulator) might be wrong. The reason that the might be wrong is because almost all of those underlying polls are using turnout models (different models, to be sure, but generally triangulating in the same direction) that look remarkably like the actual turnout of the 2008 General Election. In other words, D+ 6 or more. Some people are pointing out that this doesn't really 'feel' right.

Is turnout going to be like it was in '08? That's "it". *That's* "all you need to know".

 

What about you? Do you think that turnout in 2012 is going to be roughly the same as turnout was in 2008?

[OCinBuffalo]

Close election, Obama needs to win less swing states. That's it. When you look at 538 that is basically all that is presented. And when you look at the electoral college you see it is true. Also, we'll know in a week. So who cares.

The only thing I have ever cared about: using bad math to support a story. The story is that Obama can't lose no matter what. Remember the "Blue Wall"?

 

Or, that, like you are saying, he has a structural probability of winning. The undeniable implication that goes along with the story = "so don't bother voting, because Obama is probably going to win anyway."

 

That is the story, the narrative, that has been consistently informing Silver's work, and, it's just a coincidence that the same story, "we will win because of our turnout", is what we have been getting from OFA for the last year? Come on.

 

Do you honestly believe that..."latinos are going to vote in massive, +6, #s this year" and "hey, look Obama is winning due to projected massive D turnout...just look at these (August) polls!"...are merely coincidental narratives?

 

Come on, you are a lawyer. None of this smells foul to you?

 

I have a narrative for you. You may not like it, but it's just as likely to be true: "This race was in fact never close. Obama had a high probability of losing. The only way Obama ever had any chance of winning, was by depressing white voter turnout <= 2008. So every story, poll and analyst that could be gotten to...was."

 

Which story, given the #s we have today, is more likely?

[dayman]

The point i'm making is not that the data is all perfect. It's simply that nobody except Dick Morris will run around talking about a landslide in all the swing states. It's that simple. Almost nobody denies it is close and Obama does have an easier route. Romney still can win btw, it's just he needs a bit more b/c he has less in his pocket. That's why Obama has a better shot. You can complain about basically every poll and pretend you know exactly what the turnout will be, but you don't. Hell sirius radio just talked about new polls including a right leaning poll showing NC is close now! Does that mean obama will win there? No. It means what every other poll means, nobody knows for sure, it seems close, obama has more in his pocket

[OCinBuffalo]

The point i'm making is not that the data is all perfect. It's simply that nobody except Dick Morris will run around talking about a landslide in all the swing states. It's that simple. Almost nobody denies it is close and Obama does have an easier route. Romney still can win btw, it's just he needs a bit more b/c he has less in his pocket. That's why Obama has a better shot. You can complain about basically every poll and pretend you know exactly what the turnout will be, but you don't. Hell sirius radio just talked about new polls including a right leaning poll showing NC is close now! Does that mean obama will win there? No. It means what every other poll means, nobody knows for sure, it seems close, obama has more in his pocket

The key difference between August, when I started talking about polls, and now?

 

You are saying this is a close race now. In August, perhaps not you, but certainly Nate Silver, was saying Romney had no chance.

 

The data has not changed. It was saying this was a close race in August. In fact, it was saying that there's no way Obama will get a D+7-8 turnout, like he did in 2008. But, that wasn't "the narrative" in August, was it?

 

No. I KNOW the turnout will be D+2-3 at best. I know that, because the data is clear....and I've cited plenty of examples. As I've said, I'm being conservative in that estimate. There are people out there, like Dick Morris, who get paid to do this, and who are saying this will be an R+1 election. The difference is: Dick Morris has historical data to back up that claim.

 

Where is your historical data, that shows that an incumbent, who can't reach 50 pts in most polls, state or national, wins the election? That's where we currently stand.

 

Why isn't that, or any of the fundamentals that are used by reasonable models...such as the U of Colorado http://www.colorado.edu/news/releases/2012/10/04/updated-election-forecasting-model-still-points-romney-win-university, included in Nate Silver's "analysis"?

 

What is the most likely explanation for that?