<img src='http://unskewedpolls.com/map_unskewed_projection_10_16_2012.gif'> Landslide victory for Romney?
Can always count on Yahoo's homepage to carry the Obama torch- (Reuters) - U.S. President Barack Obama gained ground on Republican rival Mitt Romney for the third straight day, leading 46 percent to 43 percent, according to the Reuters/Ipsos daily tracking poll released on Tuesday. http://news.yahoo.com/obama-extends-slim-lead-over-romney-white-house-165738747.html
Romney's got to keep the momentum after tonight's debate. Then we have to see if Obama's drone attack in Libya gives him any sort of a bump. That may not happen until the last week in Ocober.
Unskewed is unrealistic. You can't simply increase the GOP voter turnout to equal DEM turnout in every state. Regardless of GOP enthusiasm, PA, OH, WI, NM, & NV have all had very real demographic changes during the past decade or more. There is almost no chance (<10%) that Romney wins NM, PA, or WI. The rest will take a fight. Lucky for him I think the Mormon things offsets the minority shift in NV some. He will really have to fight to get the auto workers in OH to back him, because there are about 4X as many auto workers as there are coal workers that he is playing to so far.
That's just your assumption. If you even toss out the Democrat and Republican vote and you look just at the Independent likely voters you reach basically the same result. Even in PA of the two polls released the last two days one shows Obama ahead with independents by a few points and the other show Romney ahead by about 10 with them. It seems PA is in play and moving toward Romney.
That's the point. You can't just toss out the partisan voters. Let's take NM for example. Romney leads by about 8 points with IND voters. That still isn't enough to offset the fact that the number of DEM voters there is about 18 points higher than the number of GOP voters. 1 in 3 voters are Hispanic, and they support Obama 2:1. Romney will not win NM unless Obama fully self-destructs and confesses to the nation that he talks to a pink unicorn every night for advice. Any study that claims he will lose at this point loses all credibility in its ability to properly un-skew the results.
"In Play" doesn't mean a likely win. It means that where there used to be almost zero chance of a Romney victory, there is now a mathematical possibility of a Romney win. Obama "only" has a 5-point edge there with just 3 weeks left. So team Romney just scored two touchdowns in the last 4 minutes of the 4th because team Obama went into prevent defense. They are now only down by 6 with 30 seconds left and they are setting up for an onside kick. Do they have a chance? Yes. Is it a good chance? No. Winning the debate tonight would be like recovering the onside kick. Winning the next one too would be like getting the necessary touchdown. Similarly, that is the only chance of a Romney victory in PA.
In short I believe all polls should be filtered using the previous election year and off year as a template. So a properly balanced poll would show results using the 2008 template and the 2010 turnout template. Then if they wished to make 3rd pass they should call their result projected and they should explain their projection.