2012 Obama 53% and Romney 47% for 2008 Obama 51.5% and Romney 48.5% 2004 Obama 48% and Romney 52%
I started looking at a site called http://unskewedpolls.com and really had a hard time verifying all their facts, but I seemed to find a little underlying truths, and allot of raw data so started all polling companies should be a little clearer with their assumptions, or to be blatantly honest just come out and say it. We beleive that there will be Dem+5 turnout in this election, in 2008 it was Dem+3, in 2004 it was Rep+4 and in 2000 it was Dem+2 and the the independents nudged the final numbers a few points in either direction in each election.
HERE is some VERY SIMPLE MATH to try and explainThe sampling average for 2012 has been plus +6% for the Democrats, because that's what the current polling companies believe is how the Presidential voting will actually happen in the election. To keep it simple lets assume Independents are tied, and the vote was 50 for Republican presidential candidate and 50 for the Democratic presidential candidate. Now there would be Dem 56 votes and Rep 50 votes or a total of 106 votes. (50/106= 47%) (56/106-53%) This results in a poll that is honestly published as Obama 53% and Romney 47% with an over sampling of democrats of 6%
The sampling average for 2008 was reported as plus +3% for the Democrats, because that's what the polling companies reported the Presidential voting actually happened in the election. To keep it simple lets assume Independents were tied (Which they weren't), and the vote was 50 for Republican presidential candidate and 50 for the Democratic presidential candidate. Now there would be Dem 53 votes and Rep 50 votes or a total of 103 votes. (50/103= 48.5%) (56/103=51.5%) This results in a poll that is honestly published as Obama 51.5% and Romney 48.5% with an over sampling of democrats of 3%
The sampling average for 2004 was reported as plus +4% for the Republicans, because that's what the polling companies reported the Presidential voting actually happened in the election. To keep it simple lets assume Independents were tied (Which they weren't), and the vote was 50 for Republican presidential candidate and 50 for the Democratic presidential candidate. Now there would be Dem 50 votes and Rep 54 votes or a total of 104 votes. (50/104= 48%) (54/104=52%) This results in a poll that is honestly published as Obama 48% and Romney 52% with an over sampling of repiblicans of 4%
I'll let you do 2000 for yourself, but you probably get the point. A polling company that comes out and says they have weighted a poll with Dem+8 and shows Independents as tied and reports Obama 54% and Romney 46% is projecting that on voting day 8% more Democrats will vote than Republicans and if they do the will be absolutely correct, but in the last 10 years viting turnout has ranged between Rep+4 and Dem+3 so a plus at poll should also be shown with a 2008, 2004 and 2000 represntation of Likely voters in each years weighting to be completely transparent...
Just my opinion, I hope this helps... Here is a funny pic I found, hope it makes you smile
