The most common sign-off question to “experts” today is “which states should we be watching on Election Night?” The answer is usually “Pennsylvania” or “Ohio” or “Indiana.”

Wrong.

If Obama is going to win, he is going to need to do between 3 and 4 percentage points better than Bush did nationwide in 2004. Because there is a strong national tide in any presidential election that hits all states, we can figure out how the night is likely to go without focusing solely on the so-called battleground states.

For the Nervous Nellies among us (myself included), we likely will be able to see what kind of night it is going to be by examining election returns in the earliest states to close their polls, regardless of “toss-up” status. For my money, the two states to look at are Georgia and Kentucky. Virginia might be added, if you want to include a battleground state.

Why Georgia and Kentucky? Both close their polls at 7pm. Both have a large number of counties (159 in GA and 120 in KY). Both have strong capacities for centralized election data reporting. Georgia is kinda close, so will behave a bit like a true “battleground state.” Kentucky is safe McCain, so will show a “pure” swing, less encumbered by local mobilization factors.

The number (and heterogeneity) of counties is important because returns will dribble in across the night. The cross-election partisan swing within a state will be even more uniform than the swing across states. Therefore, data from even a few counties will be informative in helping to predict how the whole state will go.

I particularly like Georgia in this regard, since it has a single voting technology statewide and because the SoS promises to e-mail me (and anyone else who asks) as the results come in. Kentucky is mostly DRE, but not entirely, and doesn’t have a single machine type statewide.

What about Virginia? It has some of the same characteristics as Kentucky and Georgia (closing time, lots of counties/independent cities), plus it’s right on the battle lines. My experience from watching election returns in past years has led me to believe that Virginia will be slow in reporting election returns, however. I could be wrong, and I will certainly be hitting my “refresh” button all night on my web browser. For now, however, I’m placing my bets on Georgia and Kentucky.