Democratic Underground Latest Greatest Lobby Journals Search Options Help Login
Google

MAJOR METHODOLOGY CHANGE at 538 shows Obama could be leading by HUGE margin

Printer-friendly format Printer-friendly format
Printer-friendly format Email this thread to a friend
Printer-friendly format Bookmark this thread
This topic is archived.
Home » Discuss » Archives » General Discussion: Presidential (Through Nov 2009) Donate to DU
 
OneAmerica Donating Member (98 posts) Send PM | Profile | Ignore Sat Jun-14-08 10:58 PM
Original message
MAJOR METHODOLOGY CHANGE at 538 shows Obama could be leading by HUGE margin
Edited on Sat Jun-14-08 11:03 PM by OneAmerica
Nate Silver at FiveThirtyEight.com has altered his methodology to account for errors in national polling he wasn't factoring in, from the mid-1980's to before 2004. There is apparently another alteration to his methodology that he has implemented but not yet explained, but the changes are already reflected in his weighted poll averages, and thus the Electoral College map. Here's what has happened.

The map USED to be this:



It is NOW this:


Basically, this election could end up being very close, but it could also end up being a blowout, and if it is a blowout, then it's almost certainly going to be in favor of Obama if there are no drastic changes to the electoral landscape (like a terrorist attack). Obama was projected to have just under a 55% chance of winning the election, the updated formula now gives him a 64.7% chance.

http://www.fivethirtyeight.com/
Printer Friendly | Permalink |  | Top
skooooo Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Jun-14-08 10:58 PM
Response to Original message
1. graphics aren't working....
Printer Friendly | Permalink |  | Top
 
OneAmerica Donating Member (98 posts) Send PM | Profile | Ignore Sat Jun-14-08 11:03 PM
Response to Reply #1
3. Did that fix it?
Printer Friendly | Permalink |  | Top
 
skooooo Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Jun-14-08 11:04 PM
Response to Reply #3
4. yep thanks
Printer Friendly | Permalink |  | Top
 
cliffordu Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Jun-14-08 10:59 PM
Response to Original message
2. Busted links....
Printer Friendly | Permalink |  | Top
 
RichardRay Donating Member (1000+ posts) Send PM | Profile | Ignore Sat Jun-14-08 11:08 PM
Response to Original message
5. He's getting a lot of grief from fellow statisticians, though....
There is a fair bit of discussion over there on whether he's getting more information out of the data, or whether he's creating information out of noise. The jury is still out...
Printer Friendly | Permalink |  | Top
 
Peace Patriot Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Jun-15-08 02:19 AM
Response to Original message
6. I don't see how they can factor in two stolen elections ('00, and '04),
and two rather weird ones ('02, and '06) with the weirdness pointing to stolen seats in the House and Senate, and factor in the increasing Bushite lawlessness around elections over this period (voter "caging" lists, voter roll purges, use of the DoJ for political purposes), and the increasing non-transparency of U.S. vote counting--to the point of TOTAL non-transparency in many states, and near total non-transparency in the othersand expect to come up with a reliably weighted Democrat vs Republican prediction--for, say, an Obama vs McCain race, such as this one.

In '02, Max Cleland's loss in Georgia--in a totally non-transparent vote counting system--was clearly a wrong outcome. In 2004, the exit polls said Kerry won; the corporate news monopolies' exit pollster then DOCTORED the exit polls--in ridiculous ways--to force them to fit the results of Diebold, ES&S and other totally non-transparent vote counting formulas. So, not even the demographic stats are accurate. They cannot reliably tell you how many women, or how many Catholics, or how many African-Americans voted for Kerry or Bush. This, and other facts, point strongly to a stolen presidential election in 2004. And, in '06, while SEVENTY PERCENT of the American people said that the war was THE issue for them--they oppose it and want it ended-- we somehow produced a Congress that is the OPPOSITE of the American people (60% to 70% FOR the war)--a Congress that immediately proceeded to ESCALATE the war, and to lard Bush/Cheney with billions more of our tax dollars to keep killing, torturing and "occupying" Iraqis, and a Congress that now has a worse approval rating than Bush!

So how can you base anything on stats from these elections? 1) who knows who won these elections? The vote counting is so non-transparent that nobody knows for sure. 2) strong inferential evidence points to seriously wrong outcomes.

The only reliable stats, it seems to me, would be past and current opinion polls--IF we can trust those polls, and I think we can if we look at a range of polls, corporate and independent (which have shown some rather amazing things about the American people from 2003 through about 2005--including early and BIG opposition to the invasion of Iraq--55% to 60% opposed (Feb. '03, all polls)--and major disagreement with virtually every Bush policy, foreign and domestic, ranging from 60% to 90%--2004, 2005.)

What I am saying is that I have no doubt at all that this will be a blowout victory for Obama--but we may never know how big, because his mandate will surely be significantly shaved by the egregiously non-transparent voting systems, fast-tracked into place during the 2002 to 2004 period, run on 'TRADE SECRET,' PROPRIETARY programming code, owned and controlled by RIGHTWING BUSHITE corporations, with virtually no audit/recount controls.

I don't think they will steal it from Obama (--although they have the easy capability to do so), because they would likely have an insurrection on their hands if they do. But they can and will seriously reduce his mandate for change, and hamper him in other ways--for instance, with another Congress like this one, that may have a "D" in front of its name, but that votes and conspires against the interests of the American people.

Nobody--NOBODY!--holding public office today can prove that they were actually elected. So, how can anybody rely on election statistics to predict how people will vote? We simply don't know--and can't know, and have ZERO proof of--how they have voted.
Printer Friendly | Permalink |  | Top
 
Zynx Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Jun-15-08 02:48 AM
Response to Reply #6
7. Look at the polling data from '04 without the TIA bullshit analysis.
http://pollingreport.com/2004.htm

There it is in the raw. There was no large discrepency between polling data and the results. Period. End of story.
Printer Friendly | Permalink |  | Top
 
Yukari Yakumo Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Jun-15-08 02:58 AM
Response to Reply #7
8. Only one word to add to this...
Printer Friendly | Permalink |  | Top
 
Peace Patriot Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Jun-15-08 09:17 PM
Response to Reply #7
9. You may think that that's a "Period. End of story." --but I don't.
1. Do you support voting systems that are run on 'TRADE SECRET,' PROPRIETARY programming code, owned and controlled by RIGHTWING BUSHITE corporations, with virtually no audit/recount controls?

Yes or no?

And if you do support the current vote counting system, then I don't care what your opinion is, or what facts you might support it with. You don't believe in democracy. I can't trust the opinion of someone who doesn't understand that non-transparent vote counting, in rightwing corporate control, IS fascism. It is a violation of the most fundamental principle of democracy. And that it has been virtually black-holed by the Republican and Democratic Party leaderships, and by the corporate 'news' monopolies, is the scariest thing I know about the United States of America.

And if you do NOT support this new vote counting system, and if you understand the implications of it, then I ask you: How do you know who won the 2004 election? And, how do you know who won the 2006 elections? Say you wanted to know for sure. What proof would you require? And is that proof available to you?

2. The page of stats you sent me to proves nothing--except that the election was close enough to steal--without raising the suspicions of the majority of people, who were kept completely ignorant of the radical change in the voting system--from transparent to non-transparent--that had been fast-tracked into place during the 2002 to 2004 period (with a $3.9 billion e-voting boondoggle, voted by the Anthrax Congress in the same month as the Iraq War Resolution--Oct 02). I frankly don't think it was that close, when you factor in items like the one million African-American voters who were purged from the voting rolls, prior to the election.

3. The stats you point to are not the issue/approval polls that I was discussing, that show 55% to 60% opposition to the invasion of Iraq, just before the invasion (Feb 03--all polls), and polls on torturing prisoners, Social Security, the deficit and a host of other issues, that show amazing disagreement with Bush (ranging from 60% to 90%) during the 2004-2005 period.

4. Given the fascist/corporate control of the voting machines and of virtually all news and opinion, and the complicity of the corporate 'news' monopolies and of both the Republican and Democratic Party leaderships, on the voting system, on the war, on torture, on spying and on other fascist policy, what makes you think that some polls are not being fiddled, and others conducted honestly, for different purposes? (--say, fiddled polls to support a false 'news' narrative about the election, around election time; and more honest polls when they aren't needed to support false election narratives, and/or to provide the corporate rulers with the info they need, between elections, in order to plan the defeat of the country's anti-war, progressive majority). The polls, the voting machines, the corrupt office holders and the corporate media and their imagery and narratives work together to confuse, disempower and disenfranchise the American people.

I think that it is extremely naive to trust voting machines run on 'TRADE SECRET' code--let alone code controlled by Bushite corporations (--corps with CEOs that were Bush-Cheney campaign chairs and major fundraisers; corps with financing from rightwing billionaires who also support fascist 'christian' foundations with views that would raise the hair on the back of your neck--witch-burning types), and to trust a national political establishment that would TOLERATE such a radical, fascist vote counting system, and who supported aggressive war by the U.S.A., torture and other hideous and illegal actions.

Give me some reason that I should trust them. Tell where, in this country, there are ballots that were actually counted by human beings, and not by 'TRADE SECRET' code machines, in 2004 and 2006. Tell me how, with SEVENTY PERCENT of the American people opposed to the Iraq War and wanting it ended, and saying that this was THE major issue of concern to them in the 2006 elections, we ended up with a Congress that ESCALATED the war and gave Bush/Cheney everything they wanted--billions and billions more of our tax dollars to keep killing, torturing and "occupying" Iraqis.

Prove to me how I can trust the information I read in the New York Times and the Washington Post--after all the egregious lies and disinformation they printed about Iraq.

I think you are being too trusting. And I urge you not to be. Because the consequences of trusting the fuckheads who are running this country could be very dire, indeed--including a nuclear holocaust in the Middle East that could finish off planet earth, oil wars against Iran and against South American democracies, Great Depression II, and poverty and suffering such as we have never seen in this country, even in the first Great Depression.

They are playing with fire--literally, as to war--and economically, with every major indicator pointing to extremely grave problems in all systems (banking, investment, housing, food, loss of jobs to outsourcing, emergency services, the cost of gas, bankruptcy of the airlines, bankruptcy in state budgets, a $10 trillion federal deficit over the decade, and on and on--not to mention the impacts of global warming). "Organized money" (as FDR called it*) has squeezed us dry, and has now become a sort of trans-border, floating country all to itself, with loyalty to no one. We have been sold out. We have been betrayed. We have been fucked over. We are being turned into the biggest "Banana Republic" on earth. And you think that you are being told the truth about who has won our elections? If they throw a lot of shit at you, some of it might stick. That's how much you can trust what you hear/read in our so-called 'news' media. Some of the crap they are throwing sticks in peoples' brains. That it does.

------------

*("Organized money hates me--and I welcome their hatred." - Franklin Delano Roosevelt)



Printer Friendly | Permalink |  | Top
 
Liberal_Stalwart71 Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Jun-15-08 09:24 PM
Response to Reply #9
11. YES!! YES!! YES!!! MASSIVE TURNOUT = OBAMA WIN!! PERIOD!!
I don't care what any of these polls and so-called pundits say!! They cannot predict turnout. And they are so in love with McSame that they purposefully discount his problems. But apart from that, we NEED to turn out in MASSIVE numbers. The reason why they were able to successful steal Elections 2000 and 2004 without the average American blinking an eye is because it was TOO CLOSE!!! I don't trust the machines and I don't trust the Repugs. But if turnout is so high that they couldn't possibly steal it, then we win. And that's just the bottom line. I have full faith in Greg Palast's analysis. I think he is spot on. If we turn out, we win!!
Printer Friendly | Permalink |  | Top
 
Liberal_Stalwart71 Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Jun-15-08 09:20 PM
Response to Reply #6
10. AMEN, AMEN, AMEN!!! You win the post of the day! Analysis is SPOT ON!!!!! n/t
Printer Friendly | Permalink |  | Top
 
phrigndumass Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Jun-15-08 09:55 PM
Response to Original message
12. I caught this on their website this morning ... they changed two methodologies
and the electoral vote projection for Obama jumped from 278 to 308, up 30. Their electoral vote projections are more in line with mine now, as well as EV.com. I'm also showing 277 electoral votes for Obama beyond the margin of error.

I'm happy they made these changes. Their model was a bit too conservative for the general election imho, but dead-on for the primaries.

Thanks for posting this, 1Am! :hi:
Printer Friendly | Permalink |  | Top
 
DU AdBot (1000+ posts) Click to send private message to this author Click to view 
this author's profile Click to add 
this author to your buddy list Click to add 
this author to your Ignore list Thu Apr 25th 2024, 06:36 AM
Response to Original message
Advertisements [?]
 Top

Home » Discuss » Archives » General Discussion: Presidential (Through Nov 2009) Donate to DU

Powered by DCForum+ Version 1.1 Copyright 1997-2002 DCScripts.com
Software has been extensively modified by the DU administrators


Important Notices: By participating on this discussion board, visitors agree to abide by the rules outlined on our Rules page. Messages posted on the Democratic Underground Discussion Forums are the opinions of the individuals who post them, and do not necessarily represent the opinions of Democratic Underground, LLC.

Home  |  Discussion Forums  |  Journals |  Store  |  Donate

About DU  |  Contact Us  |  Privacy Policy

Got a message for Democratic Underground? Click here to send us a message.

© 2001 - 2011 Democratic Underground, LLC