With all the talk in recent weeks previewing Electoral College variables in the fall campaign it has been difficult to find a sensible yet thorough examination of the issue. Bill Richardson's former pollster finally offers one at Salon, even with the expected Democratic slant.

It's an excellent primer for those interested, though here are a few nuggets that are particularly interesting:

But my collection of swing states is not based on current match-up polling between Obama and McCain. I mostly ignored the polls -- come on, it's May. Instead, I looked at long-term voting trends and demographics.

Amen to that! And more here on the topic of the inherent lameness of general election polls in spring.

On our own Evergreen State: "If the Democrats are in trouble here, then it's lights out nationally." Exactly.

On New Mexico: "This one always goes down to the wire...New Mexico has deviated a grand total of 1 percentage point from the winner's share of the vote in the last three elections combined." Might be close there again, no?

On Ohio:

Could choosing Ohio's popular Democratic governor as his running mate be Obama's answer? (Ted Strickland, conveniently, hails from Ohio's Appalachian southeast.) It may be the only way to prevail here in a close race.

That's a variable that should concern Republicans, just as would an Obama VP choice of Virginia Governor Tim Kaine. If the race is close and if the Electoral Map is as topsy-turvy as McCain and Obama could make it, a VP pick that turns a sizable red state into blue could be trouble.

On Wisconsin: "Gore and Kerry each won by fewer than 11,000 votes." Enough said. It'll be a war in cheesehead country to the end.

Yes, VP candidates don't often actually do much to shift the Electoral College map. But given how this year is unfolding and current projections for swing states aplenty, it would be foolish for either candidate not to select a running mate who has known strength in at least several projected battlegrounds.