It's always a bad idea to handicap elections right after the conventions. Remember "17-point" Michael Dukakis? He managed to lose to George H.W. Bush, one of the most uninspiring presidential candidates in history. Never believe the polls that come right after the conventions.

The polls that matter are the ones that appear in October. In a tight election (and this one is, at least right now), things don't move until October. Dukakis lost his lead within three weeks, but that wasn't a close election. This one might be. Though I think it most resembles 1980, when Ronald Reagan stood next to Jimmy Carter at the debates and managed to seem competent. After that first debate, Reagan's numbers kept going up.

The only real knock on Barack Obama is that he's inexperienced. That he hasn't come through the fire and thus isn't ready. One debate with John McCain, and I think a lot of people will decide that Obama is, in fact, presidential.

I'd put McCain's chances at 3-1, which puts Obama's line at 1-2--my house doesn't take a cut. But more importantly, I think the election will be called by 10 p.m. This one won't be close, and most of the important states will have made their decisions by 8 p.m. Want to be the first to toast the new president? Here are the tea leaves to watch.

If Obama wins North Carolina and Georgia, the election is over. If he wins those states, he will also win Virginia and Missouri (not to mention Colorado, New Mexico, Nevada...). And that's game over with an open-ended landslide draw. If you see those states turn blue, Democrats will probably end the night with 300 reps and 60 senators--and Obama might win 400 electoral votes.

If McCain wins Pennsylvania, Michigan and New Jersey, he's the winner. Call it whenever you see those states fall. It won't be a landslide--in a best-case scenario he'll score 300 electoral votes. But even on the best of all possible GOP nights, the Democrats will increase their margins in both houses.

But lets say election night settles down. Obama picks off Virginia and McCain steal Michigan. What then?

The midwest probably won't tell us much. On a tight night, you'll see the candidates split the western banks of the Mississippi. Obama might take Iowa and Minnesota, while McCain grabs Missouri, Arkansas and Louisiana. And you might see some blue just to the west, as Obama just might pick off the Dakotas. But the real action comes in the mountain west. Montana is in play--the current polls bounce around within a point or two. If the election goes deep, Montana just might the kicker.

But barring some unforseen gaffe, Obama should win relatively easily. His campaign so far has played it safe, but it's obviously in control of its candidate and message. McCain's campaign, on the other hand, has chosen to spend most of its time trashing Obama without pushing the ideals of its candidate. When all the ads are about Obama, that's a plus for the Democrat.

How can I be so confident? Here are some interesting poll numbers. McCain leads Obama in North Carolina, Florida and Ohio--but not outside the margin of error. I've been around a number of poll sites and the general consensus is that Obama has between 225 and 250 electoral votes in the bag. McCain's "strong" states tally between 120 and 160 votes. There are about 100 votes in states where McCain leads by four points or fewer.

The key to previous Republican victories was in defining the other candidate. This hasn't worked for McCain. The joke making the rounds is that you can't joke about Obama. That joke happens to be funny and true. The McCain campaign is finding that Obama is a much tougher target than John Kerry or Mike Dukakis.

More importantly, the McCain campaign is already firing off its best shots. The best time to accuse Obama of playing the race card is the last week of October, not August. McCain (or his stand-ins) can't spend the next two months whining about the race card. That's a one-time shot that might score a crucial short-term burst right before the election. When you give people three months, they tend to see through the race card.

John McCain is a good guy, and a good candidate. He'd have wiped the table with Al Gore in 2000. But in a year when most folks see Republicans as tired and corrupt, it's hard out there for the GOP pimp-in-chief. And in the end, his party is what will sink him.

That or he'll say something truly stupid. In the middle of a debate. A classic YouTube moment that will echo like Howard Dean's gelding screech.

But I don't think that will happen. I think we'll have an increasingly nasty campaign until the screens get punched, and then Obama will win.

My prediction: Obama 341-197. And you'll know by 10 p.m.

Bank on it.

Jon Worley thought Dukakis had "a real good chance" on November 7, 1988.