http://www.gallup.com/poll/9442/election-polls-accuracy-record-presidential-elections.aspx "Mitt Romney extended his lead Sunday over President Barack Obama in a much-discussed Gallup poll that shows the GOP challenger now leading by 7 points, 52-45, among likely voters. While most surveys show Obama and Romney locked in a virtual dead heat, Gallup alone finds that the Republican would win by a comfortable margin if the election were held today." Based on the fairly accurate track record of the gallup polling in past elections, it appears Mr. O is in trouble.
Pretty sure all of the major polls have track records with similar or often better accuracy comparing to Gallup. Looking at a polling avg is better than looking at one poll.
Gallup was pretty terrible in both 2008 (too far Dem) and 2010 (too far GOP), though it got the winner right - just overestimated the margins. That said, that doesn't mean it will be or won't be accurate this time around. I think the hesitancy with Gallup right now is that they are so far out of line with the other polls. They probably have a different likely voter model than the consensus, but that doesn't mean it's more or less correct - only time will tell on that. Right now, it's just one of a number of data points that we have.
http://www.rasmussenreports.com/pub...dministration/daily_swing_state_tracking_poll Sunday, October 21, 2012 The full Swing State tracking update offers Rasmussen Reader subscribers a combined view of the results from 11 key states won by President Obama in 2008 and thought to be competitive in 2012. The states collectively hold 146 Electoral College votes and include Colorado, Florida, Iowa, Michigan, Nevada, New Hampshire, North Carolina, Ohio, Pennsylvania, Virginia and Wisconsin. In the 11 swing states, Mitt Romney earns 50% of the vote to Obama’s 46%. One percent (1%) likes another candidate in the race, and three percent (3%) are undecided. This survey is conducted on a rolling seven-day basis, so today’s update is the first in which the majority of the responses follow Tuesday night’s presidential debate. This is the second day in a row that Romney has been at the 50% mark in the combined swing states and marks the widest gap between the candidates since Obama posted a five-point lead on October 5. In 2008, Obama won these states by a combined margin of 53% to 46%, virtually identical to his national margin. Nationally, the candidates remain nearly tied in the Rasmussen Reports daily Presidential Tracking Poll.
We don't elect the POTUS through a popular vote. National polls are fun to look at but are ultimately pointless. The crosstabs from one of the Wednesday Gallup poll (which I think was the first one showing a big Romney lead) showed that when you break down the poll by region, Obama had a 4-6 point lead in the East, West, and Midwest but was losing by 22 in the South. Apply that to the electoral college and maybe it points to a Romney win in Florida but it's not really a good sign for Ohio, Wisconsin, Nevada, Colorado, etc. On top of that, if you just took the straight result of the poll, it was 49-47 Romney. Gallup gave stronger weighting to the South to give a Romney +6 result. I'm not getting into the methodology, but giving extra weight to a region where Obama wasn't going to win much of anything anyway might give a nice national number for Romney but still doesn't affect the Electoral College that much. Point being that losing in the South (shocker) and winning everywhere else doesn't really indicate trouble for Obama.
It's the enthusiasm factor. Republicans are very enthusiastic to vote this time around, Just like democrats were last time. Democrats don't seem to be motivated to get out and vote. I don't think people realize how important this election is. Were in the midst of a recovery (thanks to Obama). But it has just begun. Which ever party wins this election gets to reside over the recovery and take credit for it. Which will affect many elections to come.
I don't understand how people have this mindset. The Obama campaign has broken records in small donors (over 10 million this cycle). Their ground game is out pacing Romney's (Obama leads Romney 60 to 40% in early voting). We are more than ready to re-elect the President.
This is why I question Gallup's likely voter model. As I said before I certainly think a Romney win is possible but I don't put much stock in Gallup given how far they are out from the other polls.
Democrats have contended Gallup' likely voter qualifying questions are too difficult. Such as 'Where is the closest voting location to you?'
Three keys to reading polls: 1. What is the party ID sample? D+7 was the national sample from the actual 2008 election -- a wave election for the Democrats due to the financial crisis. This was overall. State-by-state is much different, and in the battleground states, the party ID sampling should be pretty close to neutral. Looking at NBC/Quinnipiac/CBS/NYT polls with D+9 samples in Ohio and Iowa are basically worthless -- they are dramatic over-estimations of Democratic voters. Unskewed polls has done a lot of work on the sampling issue, as has Battlegroundwatch. 2. Voter enthusiasm. It takes a candidate higher than polling might suggest. The Republicans have it this year; the libs had it in 2008. 3. Trends -- Assuming consistency in the sample and a large enough sample, flawed polls can still provide valuable information by looking at how they are trending.
Of course, I suspect most of those "small donors" come from outside the U.S. The Obama campaign won't allow the back of the card code nuimbers to be used lest voter officials have access to from where those donations are coming.
Big enough and the Swing State polls suggest Romney won't secure enough states that are in a tie right now.
