Everyone talks about it. Do a poll posing a theoretical Democrat against a theoretical Republican candidate for president and there is a 15% or more advantage for the Democrat.

But plug in the names and you see as low as a 2% difference in the polls. Why?

I think there are two factors. One is all about racism. Yep. That ugly, nasty, embarassing part of America-- which, apparently, is particularly manifested by over 60 whites.

The other factor is the pollster policy. They are using the same numbers they've been using for a long time to create their polls-- about 40% Democrat, 38% Republican and the rest independent. But things have changed. First, the young people-- the "Millenials" who will change this election and elections for the next fifty years to come, use cell phones. They get left out of the polls and they are a good two to one for Obama. Then, there is the reality that the old proportions are no longer accurate. Now, the numbers should be more like 46-48% Democrat and 28-32% Republican. If they compose a poll group with the old proportions, it will give the results they get. If they modified the polls demographic components to reflect all the cell-phone using millenials and adjusted the dem/rep proportions, the results would be very different.

The handy thing about using inappropriate, unscientific, yet habitual approaches is it makes things look better for McCain. That may help his fundraising, but it also helps Obama, since Barack's team can send out calls for volunteers and funding that point out how close things are.

The offensive thing is the way MSM newscasters and pundits use the stats to malign Obama, with remarks like, "What's wrong with him?"

But now, you know better. This election, unless there is some kind of October Surprise, in the form of a new war, probably with Iran, initiated by a Tonkin Gulf phony event, and maybe even then, will go heavily for Obama. The pollsters have it wrong. They're blowing it and they are stuck with their old ways.