...when adjusted for oversampling. A recent graph released by NumbersMuncher on twitter shows the (mainly) liberal bias of polls. If we exclude the Rasmussen, every poll oversamples Democrats by at least 5%. The Dem corps poll has an oversampling of 11%! Now, how can they get away with this? They pick and choose their data carefully: they use voter turnout data from the 2008 election, which had Democrats turning up at the polls in record numbers and a depressed Republican voter base. The result was a 7% advantage in voter turnout in favor of the Democrats. So these pollsters claim they are trying accurately represent voter turnout based on one heavily outlying data point! Now we get to the good stuff: If you adjust every poll for this oversampling and show an equal voter turnout for both sides, every poll shows Romney winning by at least 1 point, even the Rasmussen. If you do this with the recent ABC/Wash Post poll, Romney is winning by 9 points!
This situation is only if turnout is equal between the groups. If turnout is similar to the 2010 elections, (or the ongoing campaign rallies where the current president is having trouble filling venues much smaller than those he packed to capacity in 2008) then Romney wins in a landslide.
So what does this mean for my fellow conservatives? All you have to do is show up and get your friends to show up and we've won!
No comments:
Post a Comment