|
|
Liberals should applaud election season madness. When you think about it, all that's happening is millions of dollars get redistributed from wealthy donors like Sheldon Adelson to lower and middle class working folks who put together ads, make phone calls, and all that other crap.
|
On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio.
Good poll for the president but I'd be cautious about reading too much into one poll.
Interesting that this poll and Gallup have such a wide disparity in how they are seeing the results among people who have already voted. We'd expect that to be much easier to measure.
|
On October 31 2012 15:53 ticklishmusic wrote:
i had a friend who had a professor who said "i don't care what version of the book you get, the history of the roman empire hasnt changed much in the last 1500 years."
best. professor. ever.
Also fairly false quote :p
I get what he means, but there have been quite a few changes in what we know about the history of the roman empire in the last 50 years alone 
But yes, most of my history texts are between 10 and 25 years old (excluding "Zeitgeschichte", which usually has a relatively brand new book).
Frankly I am always shocked how much my friends pay for some of their books (law books are between 60 and 80€ each, for that amount I buy the books for my entire semester).
|
On October 31 2012 22:51 Agathon wrote:Show nested quote +On October 31 2012 15:21 Sanctimonius wrote:On October 31 2012 14:54 Lmui wrote:Political spending in Canada is something I think is pretty fair compared to the states. http://en.wikipedia.org/wiki/Federal_political_financing_in_CanadaThere is money that comes directly from the federal government, a per-vote subsidy (For every vote you receive in an election, you get some taxpayer money). This was also coupled with the introduction of spending limits. The maximum any individual can contribute is $1,100 to the party and another $1100 to their representative. (There are loopholes but they aren't all too serious compared to what can be done by individuals in the states. Corporations also have no voice. There are cases where political spending from a large number of individuals from the same company/group have used the contribution limit to a single candidate but these (I feel) pale in comparison to the massive amount of money in US politics. Sounds a little similar to British political spending, and I wonder if they influenced one another, We have three major political parties. By law they have to have a certain amount of air time on the national TV networks, equal to each other. They can do what they will with this time, they cannot purchase more, and it is advertised ahead of time. People can tune in or ignore it as they wish. Donations are severely restricted, because we're of the opinion that if a person can give enough money to a man he will buy that man. We limit the amounts of money from donors to both politicians and parties, since it is in effect bribery. Politicians and parties should be aiming at doing what is best for the state and the citizens, not for whoever can pay them the most money. Spending on campaigns is a big one. The parties have a certain amount they are allowed to spend in each county in the UK, dictated by the population of the county. I forget how much it is per head, but it's a small amount, intentionally, to make sure the rampant increases in spending that is evident in US elections doesn't happen. Part of this may be a reflection of having irregular elections - Prime Ministers tend to stay in power until they are unpopular in their party, and a snap election can be called at any time - this means if David Cameron thought he would win, he could call for an election next month and the Brits would all march (well, just over half of them would march...) to the voting stations. Insanity, I know, since we miss out on politicians spending most of their time in charge campaigning for the next election. How does anyone get anything done in this country? I am curious, though. The amount of money spent in elections is mind-boggling and only on the increase. That's a huge amount of money spent on air-time, billboards, tv ads, newspaper ads, whatever. Would the Us economy suffer if campaign financing and lobbying was banned or tightly controlled? Same in France, public money per votes and limited gifts to 6900€ per election, and only for physical persons. Moral persons can't give any money neither services/goods etc.. Many candidates and parties got caught by the police for hidden gifts in cash, it's like corruption in my country. In addition, all parties, big or small, have the same limited time on TV to show their ideas and programs. The US presidential election looks anti democratic here, it's something impossible to imagine in France (or in Europe). Another US' specificity
Not US' specificity, it is called Americal Exceptionalism. It is the mission of USA to spread their election system to the rest of the world! ;-)
Then again, I still need to see the perfect election system. German model is good, France has some very good elements and ideas while other ideas fail, Scandinavia + Iceland has some decent systems (In some cases far inferior to the German system, in other aspects superior). I do not like the british system (Winnner takes all variant), the greek system (be the biggest party and get 50 extra parliamentarians!) and the spanish system (willfully creating a federalistic government with very autonomous regions.).
|
On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio.
Ehh, I'm pretty sure Florida is going to Romney lol.
|
On October 31 2012 23:41 radiatoren wrote:Show nested quote +On October 31 2012 22:51 Agathon wrote:On October 31 2012 15:21 Sanctimonius wrote:On October 31 2012 14:54 Lmui wrote:Political spending in Canada is something I think is pretty fair compared to the states. http://en.wikipedia.org/wiki/Federal_political_financing_in_CanadaThere is money that comes directly from the federal government, a per-vote subsidy (For every vote you receive in an election, you get some taxpayer money). This was also coupled with the introduction of spending limits. The maximum any individual can contribute is $1,100 to the party and another $1100 to their representative. (There are loopholes but they aren't all too serious compared to what can be done by individuals in the states. Corporations also have no voice. There are cases where political spending from a large number of individuals from the same company/group have used the contribution limit to a single candidate but these (I feel) pale in comparison to the massive amount of money in US politics. Sounds a little similar to British political spending, and I wonder if they influenced one another, We have three major political parties. By law they have to have a certain amount of air time on the national TV networks, equal to each other. They can do what they will with this time, they cannot purchase more, and it is advertised ahead of time. People can tune in or ignore it as they wish. Donations are severely restricted, because we're of the opinion that if a person can give enough money to a man he will buy that man. We limit the amounts of money from donors to both politicians and parties, since it is in effect bribery. Politicians and parties should be aiming at doing what is best for the state and the citizens, not for whoever can pay them the most money. Spending on campaigns is a big one. The parties have a certain amount they are allowed to spend in each county in the UK, dictated by the population of the county. I forget how much it is per head, but it's a small amount, intentionally, to make sure the rampant increases in spending that is evident in US elections doesn't happen. Part of this may be a reflection of having irregular elections - Prime Ministers tend to stay in power until they are unpopular in their party, and a snap election can be called at any time - this means if David Cameron thought he would win, he could call for an election next month and the Brits would all march (well, just over half of them would march...) to the voting stations. Insanity, I know, since we miss out on politicians spending most of their time in charge campaigning for the next election. How does anyone get anything done in this country? I am curious, though. The amount of money spent in elections is mind-boggling and only on the increase. That's a huge amount of money spent on air-time, billboards, tv ads, newspaper ads, whatever. Would the Us economy suffer if campaign financing and lobbying was banned or tightly controlled? Same in France, public money per votes and limited gifts to 6900€ per election, and only for physical persons. Moral persons can't give any money neither services/goods etc.. Many candidates and parties got caught by the police for hidden gifts in cash, it's like corruption in my country. In addition, all parties, big or small, have the same limited time on TV to show their ideas and programs. The US presidential election looks anti democratic here, it's something impossible to imagine in France (or in Europe). Another US' specificity Then again, I still need to see the perfect election system.
The perfect election system is a mathematical impossibility (Arrow's Impossibility Theorem: http://en.wikipedia.org/wiki/Arrow_theorem). In practice, one has to pick which properties are considered important and base an election system on that. So depending on which qualities you value in an election system, there can definitely be various levels of badness. In my view, the US electoral college system ranks pretty high on the "bad" list.
|
On October 31 2012 23:54 Risen wrote:Show nested quote +On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio. Ehh, I'm pretty sure Florida is going to Romney lol.
The disconnect and inconsistency between many of the polls is very amusing. Someone's going to write a book on this when it's all done.
|
On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio.
These CBS/NYT/Quinnipac polls are insane. They assume a Democratic turnout advantage at or ABOVE 2008 levels.
Obama won Ohio in 2008 by +4, you really expect that to go up? The entire country is moving away from Obama, even the states he's gonna win, and yet somehow Ohio is moving MORE towards it?
The fact that every poll they released show Obama like 4-5 points better than all the other polls for that state should let you know the polls have some sort of systematic problem.
Obama could easily win Ohio, but if he does it will be less than 4% margin he got last time.
|
On November 01 2012 00:13 Zaqwert wrote:Show nested quote +On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio. These CBS/NYT/Quinnipac polls are insane. They assume a Democratic turnout advantage at or ABOVE 2008 levels. Obama won Ohio in 2008 by +4, you really expect that to go up? The entire country is moving away from Obama, even the states he's gonna win, and yet somehow Ohio is moving MORE towards it? The fact that every poll they released show Obama like 4-5 points better than all the other polls for that state should let you know the polls have some sort of systematic problem. Obama could easily win Ohio, but if he does it will be less than 4% margin he got last time.
Yeah, I'd guess closer to a 1.5%-2% win at the most in OH. Though who knows, maybe Chris Christie helped him out enough to win over 2% of Ohio (I doubt that very much though).
Edit: It is important to note that the trends in inaccurate polls tend to be accurate themselves, so Mitt Romney's lead in FL probably has shrunk some (just hard to tell how much).
Edit2: 538 just did a pretty cool article about the current disparity between national polls and state polls (and established that there is a disparity), including altering the model for turnout in various years.
|
Yeah basically 538 takes the state polls and projects them up, rather than taking the national polls and projecting them down to the states.
Right now they are telling two different stories. If all you had was the state polls you'd say Obama is gonna win, if all you had were the national polls you'd say Romney was gonna win.
|
On November 01 2012 00:18 TheTenthDoc wrote:Show nested quote +On November 01 2012 00:13 Zaqwert wrote:On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio. These CBS/NYT/Quinnipac polls are insane. They assume a Democratic turnout advantage at or ABOVE 2008 levels. Obama won Ohio in 2008 by +4, you really expect that to go up? The entire country is moving away from Obama, even the states he's gonna win, and yet somehow Ohio is moving MORE towards it? The fact that every poll they released show Obama like 4-5 points better than all the other polls for that state should let you know the polls have some sort of systematic problem. Obama could easily win Ohio, but if he does it will be less than 4% margin he got last time. Yeah, I'd guess closer to a 1.5%-2% win at the most in OH. Though who knows, maybe Chris Christie helped him out enough to win over 2% of Ohio (I doubt that very much though). Edit: It is important to note that the trends in inaccurate polls tend to be accurate themselves, so Mitt Romney's lead in FL probably has shrunk some (just hard to tell how much). Edit2: 538 just did a pretty cool article about the current disparity between national polls and state polls (and established that there is a disparity), including altering the model for turnout in various years.
Nate Silver and Sean Trende both had commentary today about the differences between national and state polling. Bottom line, while a popular/electoral vote disagreement is possible, the current polls are overstating it. Given Obama's state polling numbers right now, he would win the popular vote by 2%. So one or the other must be wrong (or both wrong by a little)
http://fivethirtyeight.blogs.nytimes.com/2012/10/31/oct-30-what-state-polls-suggest-about-the-national-popular-vote/
http://www.realclearpolitics.com/articles/2012/10/31/whats_behind_the_state-national_poll_divergence_115979.html
|
|
|
On October 31 2012 23:56 xDaunt wrote:Show nested quote +On October 31 2012 23:54 Risen wrote:On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio. Ehh, I'm pretty sure Florida is going to Romney lol. The disconnect and inconsistency between many of the polls is very amusing. Someone's going to write a book on this when it's all done.
It's the same every election. I believe some mathmatician once proved that polls didn't do any better than random chance.
|
On November 01 2012 00:29 Signet wrote:Show nested quote +On November 01 2012 00:18 TheTenthDoc wrote:On November 01 2012 00:13 Zaqwert wrote:On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio. These CBS/NYT/Quinnipac polls are insane. They assume a Democratic turnout advantage at or ABOVE 2008 levels. Obama won Ohio in 2008 by +4, you really expect that to go up? The entire country is moving away from Obama, even the states he's gonna win, and yet somehow Ohio is moving MORE towards it? The fact that every poll they released show Obama like 4-5 points better than all the other polls for that state should let you know the polls have some sort of systematic problem. Obama could easily win Ohio, but if he does it will be less than 4% margin he got last time. Yeah, I'd guess closer to a 1.5%-2% win at the most in OH. Though who knows, maybe Chris Christie helped him out enough to win over 2% of Ohio (I doubt that very much though). Edit: It is important to note that the trends in inaccurate polls tend to be accurate themselves, so Mitt Romney's lead in FL probably has shrunk some (just hard to tell how much). Edit2: 538 just did a pretty cool article about the current disparity between national polls and state polls (and established that there is a disparity), including altering the model for turnout in various years. Nate Silver and Sean Trende both had commentary today about the differences between national and state polling. Bottom line, while a popular/electoral vote disagreement is possible, the current polls are overstating it. Given Obama's state polling numbers right now, he would win the popular vote by 2%. So one or the other must be wrong (or both wrong by a little) http://fivethirtyeight.blogs.nytimes.com/2012/10/31/oct-30-what-state-polls-suggest-about-the-national-popular-vote/http://www.realclearpolitics.com/articles/2012/10/31/whats_behind_the_state-national_poll_divergence_115979.html
I find the line at the end of the RCP article about not averaging state and national polls somewhat peculiar. Given the fact that all polls have a margin of error, trying to extrapolate by averaging combination of state and national polls seems no different to me than what they do by just averaging state polls.
|
On November 01 2012 00:39 TheTenthDoc wrote:Show nested quote +On November 01 2012 00:29 Signet wrote:On November 01 2012 00:18 TheTenthDoc wrote:On November 01 2012 00:13 Zaqwert wrote:On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio. These CBS/NYT/Quinnipac polls are insane. They assume a Democratic turnout advantage at or ABOVE 2008 levels. Obama won Ohio in 2008 by +4, you really expect that to go up? The entire country is moving away from Obama, even the states he's gonna win, and yet somehow Ohio is moving MORE towards it? The fact that every poll they released show Obama like 4-5 points better than all the other polls for that state should let you know the polls have some sort of systematic problem. Obama could easily win Ohio, but if he does it will be less than 4% margin he got last time. Yeah, I'd guess closer to a 1.5%-2% win at the most in OH. Though who knows, maybe Chris Christie helped him out enough to win over 2% of Ohio (I doubt that very much though). Edit: It is important to note that the trends in inaccurate polls tend to be accurate themselves, so Mitt Romney's lead in FL probably has shrunk some (just hard to tell how much). Edit2: 538 just did a pretty cool article about the current disparity between national polls and state polls (and established that there is a disparity), including altering the model for turnout in various years. Nate Silver and Sean Trende both had commentary today about the differences between national and state polling. Bottom line, while a popular/electoral vote disagreement is possible, the current polls are overstating it. Given Obama's state polling numbers right now, he would win the popular vote by 2%. So one or the other must be wrong (or both wrong by a little) http://fivethirtyeight.blogs.nytimes.com/2012/10/31/oct-30-what-state-polls-suggest-about-the-national-popular-vote/http://www.realclearpolitics.com/articles/2012/10/31/whats_behind_the_state-national_poll_divergence_115979.html I find the line at the end of the RCP article about not averaging state and national polls somewhat peculiar. Given the fact that all polls have a margin of error, trying to extrapolate by averaging combination of state and national polls seems no different to me than what they do by just averaging state polls.
The two averages (of state & national polls, respectively) differ by more than a conceivable margin of error, which indicates that they are not measuring the same thing, or that one (or both) systems have some systematic error. You don't just average that out and call it a day.
|
On November 01 2012 00:45 Ryalnos wrote:Show nested quote +On November 01 2012 00:39 TheTenthDoc wrote:On November 01 2012 00:29 Signet wrote:On November 01 2012 00:18 TheTenthDoc wrote:On November 01 2012 00:13 Zaqwert wrote:On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio. These CBS/NYT/Quinnipac polls are insane. They assume a Democratic turnout advantage at or ABOVE 2008 levels. Obama won Ohio in 2008 by +4, you really expect that to go up? The entire country is moving away from Obama, even the states he's gonna win, and yet somehow Ohio is moving MORE towards it? The fact that every poll they released show Obama like 4-5 points better than all the other polls for that state should let you know the polls have some sort of systematic problem. Obama could easily win Ohio, but if he does it will be less than 4% margin he got last time. Yeah, I'd guess closer to a 1.5%-2% win at the most in OH. Though who knows, maybe Chris Christie helped him out enough to win over 2% of Ohio (I doubt that very much though). Edit: It is important to note that the trends in inaccurate polls tend to be accurate themselves, so Mitt Romney's lead in FL probably has shrunk some (just hard to tell how much). Edit2: 538 just did a pretty cool article about the current disparity between national polls and state polls (and established that there is a disparity), including altering the model for turnout in various years. Nate Silver and Sean Trende both had commentary today about the differences between national and state polling. Bottom line, while a popular/electoral vote disagreement is possible, the current polls are overstating it. Given Obama's state polling numbers right now, he would win the popular vote by 2%. So one or the other must be wrong (or both wrong by a little) http://fivethirtyeight.blogs.nytimes.com/2012/10/31/oct-30-what-state-polls-suggest-about-the-national-popular-vote/http://www.realclearpolitics.com/articles/2012/10/31/whats_behind_the_state-national_poll_divergence_115979.html I find the line at the end of the RCP article about not averaging state and national polls somewhat peculiar. Given the fact that all polls have a margin of error, trying to extrapolate by averaging combination of state and national polls seems no different to me than what they do by just averaging state polls. The two averages (of state & national polls, respectively) differ by more than a conceivable margin of error, which indicates that they are not measuring the same thing, or that one (or both) systems have some systematic error. You don't just average that out and call it a day. It might not be as bad as averaging the locations of NYC and Washington DC, but still...
RCP doesn't even include a margin of error but if it did a difference of 3% of the two totals is within a 1.5% margin of error. Considering 538's margin of error for all the state polls is above 3%...I just think it's silly. He's still averaging the 5% Gallup poll with the other tied national polls, and that's much farther apart!
|
On November 01 2012 00:48 TheTenthDoc wrote:Show nested quote +On November 01 2012 00:45 Ryalnos wrote:On November 01 2012 00:39 TheTenthDoc wrote:On November 01 2012 00:29 Signet wrote:On November 01 2012 00:18 TheTenthDoc wrote:On November 01 2012 00:13 Zaqwert wrote:On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio. These CBS/NYT/Quinnipac polls are insane. They assume a Democratic turnout advantage at or ABOVE 2008 levels. Obama won Ohio in 2008 by +4, you really expect that to go up? The entire country is moving away from Obama, even the states he's gonna win, and yet somehow Ohio is moving MORE towards it? The fact that every poll they released show Obama like 4-5 points better than all the other polls for that state should let you know the polls have some sort of systematic problem. Obama could easily win Ohio, but if he does it will be less than 4% margin he got last time. Yeah, I'd guess closer to a 1.5%-2% win at the most in OH. Though who knows, maybe Chris Christie helped him out enough to win over 2% of Ohio (I doubt that very much though). Edit: It is important to note that the trends in inaccurate polls tend to be accurate themselves, so Mitt Romney's lead in FL probably has shrunk some (just hard to tell how much). Edit2: 538 just did a pretty cool article about the current disparity between national polls and state polls (and established that there is a disparity), including altering the model for turnout in various years. Nate Silver and Sean Trende both had commentary today about the differences between national and state polling. Bottom line, while a popular/electoral vote disagreement is possible, the current polls are overstating it. Given Obama's state polling numbers right now, he would win the popular vote by 2%. So one or the other must be wrong (or both wrong by a little) http://fivethirtyeight.blogs.nytimes.com/2012/10/31/oct-30-what-state-polls-suggest-about-the-national-popular-vote/http://www.realclearpolitics.com/articles/2012/10/31/whats_behind_the_state-national_poll_divergence_115979.html I find the line at the end of the RCP article about not averaging state and national polls somewhat peculiar. Given the fact that all polls have a margin of error, trying to extrapolate by averaging combination of state and national polls seems no different to me than what they do by just averaging state polls. The two averages (of state & national polls, respectively) differ by more than a conceivable margin of error, which indicates that they are not measuring the same thing, or that one (or both) systems have some systematic error. You don't just average that out and call it a day. It might not be as bad as averaging the locations of NYC and Washington DC, but still... RCP doesn't even include a margin of error but if it did a difference of 3% of the two totals is within a 1.5% margin of error. Considering 538's margin of error for all the state polls is above 3%...I just think it's silly. He's still averaging the 5% Gallup poll with the other tied national polls, and that's much farther apart!
538 has a fairly complex model that includes state and national polls as well as economic data and state "fundamentals". RCP is just an average. In context of the latter, if Trende really believes that the state and national polls are measuring different things, it doesn't make sense to average them. Whereas in Silver's model, they could be different factors that are combined in a specific way based on historical data and results.
Election Projection is another site that uses both state and national polling (and I think some other variables?) to predict state-by-state results. His electoral college prediction is the same ad RCP's and within a few EV's of 538's. So to some extent it doesn't matter whether you're averaging state and national data or not -- current polling gives a picture of Obama having a slim electoral lead, whatever specifics you're using to interpret it. Now if Romney ends up winning, that may be evidence that national polling is more accurate then we have to figure out what made it right and the state polls wrong.
|
On November 01 2012 00:35 Recognizable wrote:Show nested quote +On October 31 2012 23:56 xDaunt wrote:On October 31 2012 23:54 Risen wrote:On October 31 2012 22:38 nevermindthebollocks wrote:I admit it is always hard for me to image Romney getting more than 40% of the national vote (or even 20%) but I think this shows the key big swing states are Obama's" http://www.cbsnews.com/8301-250_162-57542715/poll-obama-holds-small-ohio-edge-fla-va-tight/?tag=categoryDoorLead;catDoorHeroMr. Obama now leads Romney 50 percent to 45 percent among likely voters in Ohio - exactly where the race stood on Oct. 22. His lead in Florida, however, has shrunk from nine points in September to just one point in the new survey, which shows Mr. Obama with 48 percent support and Romney with 47 percent. The president's lead in Virginia has shrunk from five points in early October to two points in the new survey, which shows him with a 49 percent to 47 percent advantage. I have a feeling there's still a chance for North Carolina too and the election will be all but over before the polls even close in Ohio. Ehh, I'm pretty sure Florida is going to Romney lol. The disconnect and inconsistency between many of the polls is very amusing. Someone's going to write a book on this when it's all done. It's the same every election. I believe some mathmatician once proved that polls didn't do any better than random chance.
Bullshit.
|
On October 31 2012 23:17 jdsowa wrote:
Liberals should applaud election season madness. When you think about it, all that's happening is millions of dollars get redistributed from wealthy donors like Sheldon Adelson to lower and middle class working folks who put together ads, make phone calls, and all that other crap.
No one benefits from the decline of political discourse which is ultimately caused by the lengthy election season in the US.
|
On October 31 2012 15:01 sam!zdat wrote:Show nested quote +On October 31 2012 14:57 Souma wrote:
And textbook companies are evil, christ. In Japan I'm not even sure I spent a fifth of what I usually pay here for textbooks. Yes, I work part time at a community college bookstore and it's ridiculous. They do everything in their power to quash used book markets. There's a lot of 200+ dollar textbooks that are essentially disposable one-use only because of "added content" with scratch-off access codes or things like that. The textbook reps lobby the profs pretty hard to get them to buy their books - I don't really know how they convince them, though, the profs must not give a shit about how expensive that stuff is. There's no reason you need to release a new algebra textbook, render all the old ones useless, and make everybody buy new hardbacks. Algebra didn't change. It's disgusting.
Oh, you don't know how that works? One word: kickbacks. There is a reason professors force all their students to buy the new edition of the same $100+ books over and over every semester, and it has nothing to do with the rising quality of the book. Of course they will deny they get a penny.
The newest scam are these "custom" textbooks they put out now. With those, they can bribe legally with "royalties."
|
|
|
|
|
|
EPT
