Most people have learned that that the best way to make sense of polling data is to not pay any attention to it. The result is that most people aren’t confused by polling data because what they know of it is sketchy at best. For example, most people probably think that right now Obama leads Romney in most presidential election polls by 3 or 4 percent.

Over the last 2 weeks Gallup, CNN and Newsweek polls all give Obama a 3-point lead over Romney. Fox News poll gives Obama a 5-point lead over Romney for the same period. Fox News seems to always give Obama the best numbers, making it a odd that he hates them so much. Meanwhile Rasmussen gives Romney a 1-point lead over Obama for the same time period. Real Clear Politics averages all the polls and says that gives Romney a 1.5-point lead over Obama.

This is where most readers will leave this page and go somewhere else. Who cares about these polls anyway? Are they even close to accurate? I’ll try to make this interesting.

Anomalies regularly appear when you look at several different polls, especially polls taken for different criteria. For example, Rasmussen has Obama’s job approval at 45% and gives him 44% in the Obama/Romney presidential tracking poll as of Monday, July 9th. Obama scores a 53% job disapproval rating, but Romney gets only 47% in the presidential tracking poll.

Fewer people in the Rasmussen PT poll like Romney than dislike Obama in the JA poll. This must indicate that although a majority of American’s disapprove of the job Obama’s is doing, about 8% “out of “ the 53%, (15% “of” the 53%) them aren’t all that favorable to Romney. If those split down the middle in the election, Romney and Obama will each gain about 4 points giving Romney a lead in the popular vote of 51-49. That’s a bit close and doesn’t mean he’ll win since it is the number of electoral votes that counts and not the popular vote. Obama can expect to get the electoral votes of the high population blue state on the Western and Eastern seaboards and with at least one of the middle America battleground states he’d like reach the critical 270 electoral votes.

Romney clearly needs to win over more than half of those voters disenchanted with Obama but not yet on board with Romney. He needs to keep Obama from adding anything to his electoral count that he will get just because he’s a Democrat from the Eastern and Western blue states. That means states like Florida, North Carolina, Virginia, Pennsylvania, Ohio, Wisconsin, and Colorado are vital for Romney to win.

I tend to only give credibility to the Rasmussen poll. That’s because Rasmussen polls 1500 likely voter and all the other polls are talking to registered voters or simply adults. Their sampling size is smaller, sometimes as few as 500, making their margin of error much higher. Rasmussen has the best track record of all. Rasmussen predicted that Obama would beat McCain in 2008 by 52% to 46%. The final results were 53% to 46%.

Polls right now don’t mean much because the election is still pretty far away in political time. That suggests another phenomenon that must be understood. Most pollsters are in line with the mainstream media, meaning they favor Democrats. The pollsters want two things: they want Democrats to win and they want to have a sterling reputation for accuracy. At this time they are less concerned with their accuracy than with making sure Democrats win. They know they will be judged mostly by the numbers they produce in the last 2 weeks leading up to the election. Once the election is over the polls that came before the ones close in time to the election are quickly forgotten. Therefore, if they manipulate the results of their polls now, and there are many subtle way of doing that, they take little risk to their track record. That means a lot of polling this early is not for the purpose of reporting the news but for the purpose of creating the news. They are trying to create a bandwagon effect in the minds of voters. People who don’t pay attention to politics can be influenced by polls showing Obama ahead of Romney right now. They will tend to think that maybe Obama isn’t so bad after all if he’s ahead in the polls.

Here is a shining example of how Democrat pollsters (that’s almost all pollsters) manipulate the results of their polls without actually lying, or at least not lying too much.  Today, a Washington Post poll shows the presidential race to be dead even with Romney and Obama both at 47%.  That is the lead headline of the story, and as we all know most people don’t read too much farther down in a newspaper story than the headline.  So way down near the end of a fairly long story we find out that this same poll shows 54% of those questioned disapprove of the way Obama has handled the economy. But that’s not all. The poll’s sampling was heavily slanted toward Democrats. Only 24% of respondents call themselves Republicans, with a full 60% listed as Democrat or independent. This poll should scare hell out of Obama.

I believe Rasmussen is an exception to the way most polling works. Rasmussen is the one polling group that is most closely associated with one individual, Scott Rasmussen. I believe he cares more about his reputation for accurate polling than who wins elections.  Not that he doesn’t care who wins, just that he won’t sacrifice his integrity for it. That makes him the one least likely to manipulate his results at any time, either close or far from election day. His track record shows it, as does his methodology.