Election night is a great spectator sport, a good excuse for a party, and sometimes an exercise in group therapy. Everyone wants to know the results before the results are in, and prognosticators are thick on TV and in the blogosphere. Exit polling may reflect how things are breaking, but such polls have a poor track record. The AP is trying to increase the accuracy of polls by factoring in early voting and voters who only have cell phones, but there is no way to know how accurate they are until the real numbers come in, by which point it won’t matter.

I prefer the traditional practice of looking at key precincts or counties, which is more fun because anyone can participate. Why wait for the networks to call the race when you can do it yourself? These days state by state, county by county data are available in real time online, so you can be just as informed as the talking heads.

Indiana is a good place to start looking for numbers, since most of the polls there close at 6 P.M. The question is not who will win, but the size of McCain’s margin. The latest polling shows McCain with a slight edge, but this state has returned respectable margins for Republicans in most elections, so a slight edge is something to be worried about. McCain would like to see a 59 percent win like George Bush enjoyed in 2004.

Indiana was noteworthy for huge turnout on the Democratic side in the May 6 primary. It is unusual for the primary vote to be greater than same-party vote totals from the previous general election, but that was the case in every county but one in Indiana; on average, 31 percent more people voted in the 2008 Democratic primary than voted for Kerry in 2004.

This phenomenon, noted in other states as well, was said by some to be a reflection of the importance of the race, or the result of Obamamania. Regardless, Hillary Clinton won Indiana 636,000 to 632,000, and of the eight counties showing especially high enthusiasm (with a 65 percent or more increase over the 2004 election turnout), only one went for Obama. Crossover voting probably accounted for a great deal of the increase; Republican turnout was much lower than Democratic turnout, and the GOP race had been decided by then. McCain’s margin will largely depend on how many of Hillary’s supporters were closet Republicans, and how many of the real Democrats who cast primary ballots for Hillary will vote for Obama. So watch turnout and how it distributes, whether it reflects the dynamic of the primary or not; that will tell whether this is a transformational election as the Obama team expects, or a traditional election as the McCain team hopes.

Watch the Clinton stronghold of Dearborn County, where Gore got 6,000 votes in 2000 and Kerry got 6,596 votes in 2004. Obama got 1,782 votes in the 2008 primary, while Clinton received 5,270. Obama’s fate in that county is clearly in the hands of Clintonites, making it a good indicator of which direction the whole state may move. If Obama gets 7,000 or more votes, he’ll probably win the presidency. If he gets 5,000 or fewer, it won’t look good.

Also check Hamilton County, where Kerry got 26,388 votes in 2004 — only 25 percent of the total for that county, which is reliably Republican. Yet Obama alone received 25,351 votes in the primary, 96 percent of the Kerry 2004 total, with Clinton taking an additional 16,491. If those numbers hold across the board Obama could well win the state, which would surely mean a broad national trouncing for McCain. On the other hand, if Obama simply matches Kerry’s performance in that county, it means that he did not build on the numbers he attracted in the primary; i.e., the Hillary voters did not go his way.

The largest and most urban counties, Marion (Indianapolis) and Lake (near Chicago) also bear watching, particularly since the Obama campaign is counting on massive urban turnout to offset McCain’s strengths in the suburbs and especially rural areas. Kerry won both counties in 2004, and in 2008 Obama trounced Clinton in both. Both counties had below average “enthusiasm” numbers in 2008, but Obama still got 79 percent and 64 percent respectively of the Kerry general election totals. If Obama’s totals in these counties reach 180,000 and 130,000, it will mean he retained the total Democratic strength from the primary season, and he will be in great shape, perhaps even winning the state. If his numbers are closer to 130,000 and 75,000, it will mean he could not reach beyond his primary base and will have hard going.

Virginia, unlike Indiana, is a true battleground state, and data should be available after the polls close at 7 p.m. The latest polling shows Obama leading in the state by four to six points. It is possible for McCain to win without Virginia, but Pennsylvania would have to flip. (That seemed unlikely until Obama explained his energy policy this way: “If somebody wants to build a coal-powered plant, they can. It’s just that it will bankrupt them.” A McCain campaign source tells me the comment has spread like wildfire.)

The Virginia primary was held February 12, a week after Super Tuesday, when it was becoming clear that Obama’s challenge to Clinton was serious. Yet the enthusiasm factor was not present in Virginia to the degree it was elsewhere. Primary turnout was two-thirds that of the 2004 general election, numbers more in line with what one might expect — a little higher than usual, but nothing like Indiana.

As in Indiana, early indications will come from urban and rural votes. Three urban areas that traditionally vote Democratic by at least 2:1 are Alexandria, Richmond, and Norfolk. They also went for Obama over Clinton by at least 2:1, 3:1 in the case of Richmond. The latter will be the critical indicator. If turnout is higher than normal and Obama takes 75 percent of the vote, he will take the state. If Obama only matches Kerry, it won’t give him the win.

The rural areas usually vote Republican, and three representative counties to watch are Augusta, Orange, and Appomattox. All three chose Obama over Clinton, but by narrow margins. These will be interesting to watch, because their returns are likely to come in early, and it will be easy to assess whether Obama has added to his base with Clinton voters, previous undecideds, and disaffected Republicans.

In Augusta, Kerry got 7,019 votes, a disappointing 24 percent. Obama will seek to build on his base of 2,544 primary voters (2,213 went to Clinton). In Orange, Kerry received 5,015 votes (39 percent) in 2004. Obama got 1,817 primary votes, with Clinton netting 1,455. Appomattox has shown consistent, albeit low, levels of support for the Democrats — 2,239 votes in 1996, 2,132 in 2000, and 2,191 in 2004. During the primaries Obama attracted 616 votes, Clinton 589. Given the stability of the Democratic vote in that county, it could be a decent bellwether — which side of 2,000 votes Obama winds up will determine whose Appomattox this will be.

Naturally, when watching the returns, one should look for odd or unexpected movements that could pop up anywhere. That’s part of the fun. And if this election lives up to the hype, there are likely to be many such movements. If not, that is significant in itself.

Other things to look at and keep in mind.:

‐ David Leip’s Atlas of U.S. Presidential Elections.

‐ The IDB/TIPP poll shows unaffiliated voters breaking 43/36 for Obama but with a large number undecided. This is a sizeable group of undecided voters this late in the game, and how they decide, if they do, will determine the outcome.

‐ Missouri has been on the winning side of every election since 1960; Ohio, Kentucky, and Tennessee have been on the winning side of every election since 1964.

‐ The Washington Redskins’ loss Monday night presages an Obama win, but Bush’s 2004 reelection broke the team’s lifelong record of accurately calling presidential contests.

‐ The Weekly World News reports that the space alien switched his support from Obama to McCain. “The Alien has correctly predicted the winning president in every election for the past 28 years,” the News notes. Batboy, a hard-core Democrat, naturally backs Obama. – NRO contributor James S. Robbins is the author of Last in Their Class: Custer, Picket and the Goats of West Point.