Presidential Election Overview

Can the Polls Be Off Enough for a Romney Win?
Any prognosticator that knows politics would tell you that in any given election the polls can indeed be off by several percentage points. However, the track record of poll averages and trend lines will also tell you that such an event is rare. In fact the last time state polls were off by a significant margin across the board was way back in 1992, in which polls were biased against republicans by 2.9. Such a scenario in reverse for 2012 would not only give Romney the election, but do so quite convincingly. Even the Princeton Election Consortium that is showing a land slide victory for Obama indicates that such a scenario would make Romney the likely winner. This sharp discrepancy is one of many reasons why I really question their predictions.

However, Romney does not need something this dramatic to occur to hand him the presidency. Contrary to what Nate Silver has been saying, this just is not the case. All we need is the polling in one state to be off by about a three point margin (Ohio) and a a couple small things to go Romney's way. This type of scenario happens a lot, including 2008 in which several swing state poll averages had a republican bias between 3-4 points, but in the end the total bias among all the polled states was only 0.8 points. Obviously, in this case we would still need Virginia (which is dead even) to go for Romney and then one state from all the others to go his way. (Colorado, New Hampshire, Iowa, etc.) So republicans are not in need of some type of incredible polling bias as some have tried to state, but the polls to be off in basically one state.

Some would argue that Romney also needs every state that the polls show him narrowly ahead to also break his way and not have a substantial counter-bias in any state that would be pro-republican. With all due respect I do not see much validity for any republican bias in this race. Why?
1.) Reports are showing that even democrats internal polling are showing the same basic results as the media polls. Generally speaking  internal polls are based upon somewhat dream scenarios that they see as likely happening, as it applies to voter turnout ratios. In this case, democrats have the rare "advantage" of having their likely scenarios represented in the published media polls. Back in 2008 democrats claimed that their internal polls were clearly showing Obama's margin being larger then the polls were indicating and this ended up being the case. According to articles I am reading, democrats are not finding much discrepancy between their polls and other polls for the most part.
2.) Almost every poll is showing a repeat of the record seven point advantage that democrats had in the 2008 voter turnout. In fact some Ohio polls are showing this margin even growing to a nine point advantage. I understand that this democrat bias is not intentional and is happening because of their polling formula they use to determine such things, but the notion that we are going to see anything like 2008 in voter turnout this timer around is just plain ludicrous! (Yes, I will stand behind this statement to the end.)
3.) Romney has something going his way that is giving prognosticators nightmares and that is popular vote polls have been showing this a dead even race, which does not collaborate with the state polls whatsoever. I have stated numerous times that my research has shown the state polls to be generally more accurate then national polls, but if the state polls have any reasonable republican bias then this is going to likely mean that the popular vote polls will be off by a very significant margin that has not been seen for a long time.

Will My Formula Consider Polling Bias?
After looking at the internal numbers behind numerous polls I am convinced that we could see dramatically different results then the polls are showing right now. But, like most prognosticators my predictions and margins are based mostly upon polling data. Otherwise, predictions would be based upon conjecture and feelings at the time. One does not need to go very far in my blog to see that my feelings have been all over the board. So if the polls are way off then my final predictions will probably be way off. For the record though I would still have to say my personal conjecture and feelings are still telling me Obama will win by a narrow margin.

The only thing my formula does do is take into consideration polls that have a track record of bias for one party and compensate polls that are outliers to a degree that is based upon that polls past accuracy. This and my trend line compensation are why my margins are usually different from the basic average of all the polls on RCP.

Are Both Candidates Worried and Confident?
When is a presidential candidate not worried, but portraying confidence would be a better question. Yet, from the reports I am reading while both campaigns feel like they are the favorite according to their numbers, both campaigns are losing sleep over one major issue...What if the other sides numbers are correct. After looking at the logic behind both sides numbers I can say that I see the validity behind both sets of opinions. However, republican pollsters admit that if what they see as likely were to happen then it would be a dead even race, with a very slight victory for Romney. However, if the democrats are right then Obama would win by a more comfortable margin.

Obama's Likely Path to Victory
Win the states that are already showing him with reasonable margins of victory; Nevada, Iowa, and Wisconsin all of which he is strongly favored in. Then he would just need one of the following states: Ohio, Virginia, or Florida. He could lose all three of these big states and still win the election if he pulled out both Colorado and New Hampshire as well.

Romney's Likely Path to Victory
Romney needs wins in Virginia (dead even) Florida (currently up by about one point) and Ohio (down by about three points) then he would be considered in the drivers seat on the basis of him only needing one additional state from a long list...Colorado, New Hampshire (both of which are extremely close) or longer shots like: Iowa, Nevada, or Wisconsin.