As ever, it's too close to call.

Electoral-vote.com now projects Kerry as likely to win more electoral votes than Bush - but, significantly, Florida is one of three states in which the candidates are polling as exactly tied (the others are Maine and Nevada), which means that the projection does not give its votes to either candidate. In any case, if Bush takes Ohio and Kerry Pennsylvania,m then neither candidate can win without Florida. Put another way, whoever takes two of these three states will be the next President. Using a somewhat more problematic methodology in which no state is interpreted as exactly tied, and in which Bush wins Florida and Pennsylvania but not Ohio, Electionprojection.com is predicting a Bush win.

It appears that Kerry must continue to work very hard to win Ohio and Florida - not exactly news - and that he must also hold on to a very teensy lead in Pennsylvania. Protecting the rights of voters of color in places like Cleveland, Cincinnati, Philadelphia, Pittsburgh, and Miami is going to be absolutely critical. In addition, it would also be good for him to pull ahead in Maine, a state he once expected to win easily but where Bush has pulled even. In addition, there's evidence that far more overseas ballots will be cast this year than ever before, and this largely unknown group could actually decide the election.

All of the above is, of course, barring election fraud.

I'm mulling over taking a trip to Ohio to help with canvassing and other Kerry activities sometime before I go to Oregon on October 26. Anyone who might be interested in joining, please get in touch!

Mark Morford has a cute column in the SF Chronicle headlined, Who The Hell Is "Undecided"? And why do so many election polls leave you angry and stupefied and drunk?