Prediction Time

Election is Tuesday, for those of you who live off the grid. At this point, the biggest unknown is the youth turnout. Early voting patterns have firmly established the fact that the Dems in general and African-Americans in particular are motivated to deliver a large turnout. Probably augurs a large Republican turnout as well. Here's my prediction for the outcome:

Electoral Vote: O:338, M:200 (Obama wins Ohio, Pennsylvania, Virginia, Florida, New Mexico, Colorado and Nevada; loses North Carolina, Indiana and Missouri; no second-tier swing states like Georgia, Arizona, North Dakota or Montana swing blue).

Popular Vote: O:51.7%, M:47.3%, others 1%

Senate: Dems: 57 seats not including Leiberman and the socialist dude.

Upside Scenario: Obama wins popular vote 53% to 46%, wins electoral vote 375 to 163, with Indiana, Missouri, North Carolina going blue and no other states swinging. Insane (and kind of irrational) Upside Scenario: Obama wins popular vote 54% to 44%, wins electoral vote 406 to 132, picks up Georgia, North Dakota, Montana and Arizona, and has near-miss in one of the southern states like Louisiana, Mississippi, or South Carolina.

Downside Scenario: McCain loses popular vote but wins electoral vote 273 to 265, with Obama trading Pennsylvania for Virginia and McCain running the table in every swing state. More Likely Downside Scenario: Obama wins popular vote 50.5% to 48.5%, wins electoral vote 306 to 232 by holding Pennsylvania, Ohio and Virginia.

State to Watch: Indiana.  Polls close early, so we'll see some data before other swing states. High early voting (450K through Halloween -- 10% of registered voters and in some counties 3X 2004 early voting turnout). We should therefore get a sense of general election turnout and demographics that could be a preview for other states where we've seen similar early voting patterns (North Carolina, Georgia, Nevada, Florida, etc.).  If Obama is close or leading in Indiana with most precincts reporting, he's going to have a big night. If McCain blows him out by 5%+, this will be a nail-biter.

Trend to Watch: Pundits over-valuing early voting results. To the extent these early votes are reported before election day votes, you may see sizable Obama leads that evaporate as the night wears on -- in Louisiana, Georgia, and other states. North Carolina may look like a blowout (Obama +8%) when polls close, and may end up with <1% popular vote separation at the end of the day.