But I said there were other possible scenarios. I can think of three.
The 1964/1972 scenario: Challenger disqualifies himself. Barry Goldwater and George McGovern were idealistic, intelligent senators who took positions on issues that made them unacceptable to most voters in years favorable to incumbents.
This could happen to Mitt Romney this year. And it might well have happened if some of his primary opponents had won the nomination. But he doesn't seem to be the kind of candidate who would disqualify himself. Chances for this scenario: less than 5 percent.
The 1988 scenario: Affluent voters break strongly Republican. Vice President George Bush was 17 points behind Michael Dukakis after the Democratic National Convention. But he came back to win by a 53 to 46 percent margin.
One reason is that his "read my lips, no new taxes" promise solidified his support among affluent suburbanites. His margins in suburbs enabled him to carry metro Philadelphia, metro Baltimore, metro Detroit, metro Chicago, metro Los Angeles and the surrounding states.
Since then, affluent non-Southern suburbanites have trended Democratic. And big city crime and welfare rolls -- cause for complaint in 1988 -- have declined.
Republicans' conservative stands on cultural issues and the increasing Southern influence in the party repelled suburbanites. Barack Obama carried most affluent non-Southern suburbs handily in 2008.
But Romney showed particular appeal to this constituency in the primaries. Without big margins in affluent suburbs, he would have lost Michigan, Ohio and Illinois to Rick Santorum.
Romney's proposed tax cuts and Obama's proposed tax increases pose the sharpest contrast on the tax issue since Bush beat Dukakis 24 years ago. And economics is far more important than cultural issues this year. Chances for the 1988 scenario: maybe 20 percent.
The 1980 scenario: Late break away from the incumbent. We remember the 1980 election as Ronald Reagan's landslide defeat of Jimmy Carter.
It didn't look like that during the campaign. Carter led in polls much of the time. Sometimes the race looked like a 2000/2004-style long, hard slog through target states.
But Carter's job rating was buoyed up that year by approval of his varied attempts to free the hostages in Iran. Underneath those numbers, his ratings on other foreign issues and the economy were weak.
Most voters were ready for an alternative, but were wary of Reagan, who was 69 years old and supposedly extreme conservative. He might have disqualified himself in a number of ways.
Instead, in his one debate with Carter, on the Thursday before the election, Reagan echoed a 1934 Franklin Roosevelt fireside chat, which he remembered but the press corps didn't. "Are you better off than you were four years ago?" he asked voters.
Opinion moved quickly. Weekend polling showed an unprecedented 10-point shift from Carter to Reagan. Pollster Pat Caddell had to go to the White House Monday morning and tell Jimmy Carter that he was not going to be re-elected president of the United States.
Could something like this happen this year? It is my view that Obama was helped in 2008 by a widespread belief that, in the abstract, it would be a good thing for Americans to elect a black president. I know I felt that way myself.
This year, I sense that many, perhaps most voters do not want the country to be seen rejecting the first black president. Such a feeling might be buoying Obama's support despite the lagging economic recovery and the widespread opposition to his signature policies.
If that is correct, it is possible that in the last days of the campaign a large number of voters will decide, quietly and out of public view, that they just don't want any more of what they've had for the last four years and they will try the other guy and see if he can do better.
That's what happened in 1980. Reagan carried 44 states and won the popular vote by 10 points, more than anyone else since. Chances for the 1980 scenario: maybe 20 percent.
So what remains for the chance of the 2000/2004 long, hard slog scenario? At least 55 percent. Still the best bet. But not the only one.