Mathew Gross has challenged, and I have answered. Here are my predictions, based on:
a) the economy
b) the numbers and preferences of self-identified registered voters AND leaning independents
c) how I think the campaigns have gone so far, and
d) past performance in elections of candidates
e) all other random factors that I could think of that affect my perception of the race.
Bush: 52% of the popular vote
Kerry: 48% of the popular vote
Bush victory: about a 65% chance (MLE estimate)
Electoral college: I think that PA, OH, WI, NV and ME will go for Kerry, and Kerry will have 268 electoral votes, leaving Bush with 270 votes (giving him NH, IA, NV and AK).
I think that adds up...trust a statistics concentrator to have made a regression model for the popular vote and then not be able to add up the electoral college votes correctly. Let me just say right now though that I REALLY REALLY hope I'm wrong. No one will be happier than me to find out that I am way off.