I’ve been meaning to write this for awhile, and this post from Nate gives me a good point of reference to work from.

Bascially, one thing to keep in mind as you digest coverage of the race is the different ways in which pollsters (and pundits) view than race than the way strategists look at it. Succinctly, pollsters and pundits tend to view the rack in the quasi-technicla way of being 50 distinct races that poll differently and could hodgepodge together randomly to produce a winner. In the real world though, we know that’s not the case. There are red states, blue states, swing states, and some swingier than others.

The other fact of the matter is that, new electoral college obsessions notwithstanding, the Presidential race is still, in effect, a national race. Nate has been one to fervently point that out (and nothing in this post is directed at him), by demonstrating the severe unlikelihood of a popular vote-electoral college split. In truth, demographics play the largest role in voting patterns, and those tendencies remain more or less the same nationally. Hispanics, evangelicals, blacks, people making under $30,000, etc. tend to have the same general breakdown in votes regardless of state of residence.

Which brings me to Georgia. Georgia is nothing more than the passing obsession of pundits. “Can Barack Obama win Georgia?” is like the perfect area of cable news talk. It sounds downright stupid, but a case can be made that it’s possible, and the outrageous factor makes for “entertaining” talk. But at the end of the day, it’s totally meaningless. If we’re even talking about Obama being close in Georgia, then there’s no point in talking about anything else because Obama has won the election pretty handily regardless.

Here’s how this works; there’s basically a sliding scale that correlates increased popular vote margins to electorl victories. So, for example, if Obama wins the popular vote by 1% this year, he probably wins Iowa and Colorado. 2% brings in New Mexico, 3% Ohio, and so on. That goes on down the list of Bush states from 2004 in what essentially ranks the swing value of every state. If Obama gets all the way down to pushing for Georgia, then there is no contest. It’s more or less unthinkable that Obama could be extremely close in Georgia while not winning Colorado, Iowa, New Mexico, Ohio, Nevada, Virginia, Missouri, Florida, Indiana, Alaska, Montana, and North Carolina. So at that point, Georgia becomes nothing but a margin padder for Obama.

Again, this is not to say there’s anything wrong with pollsters per se, but they don’t view the race the same way a strategist does, which makes their analysis a little shaky.