This article provides a fairly good overview of problems with the vote this year. Many of the examples and points are linked to sources for the information. Very readable and understandable. From AlterNet: Election 2004: Worse Than 2000?
A poster named "TruthIsAll" on the DemocraticUnderground.com forums laid out the questionable results of Tuesday's election in succinct fashion: "To believe that Bush won the election, you must also believe: That the exit polls were wrong; that Zogby's 5pm election day calls for Kerry winning Ohio and Florida were wrong (he was exactly right in his 2000 final poll); that Harris' last-minute polling for Kerry was wrong (he was exactly right in his 2000 final poll); that incumbent rule #1 – undecideds break for the challenger - was wrong; That the 50% rule – an incumbent doesn't do better than his final polling - was wrong; That the approval rating rule – an incumbent with less than 50% approval will most likely lose the election – was wrong; that it was just a coincidence that the exit polls were correct where there was a paper trail and incorrect (+5% for Bush) where there was no paper trail; that the surge in new young voters had no positive effect for Kerry; that Kerry did worse than Gore against an opponent who lost the support of scores of Republican newspapers who were for Bush in 2000; that voting machines made by Republicans with no paper trail and with no software publication, which have been proven by thousands of computer scientists to be vulnerable in scores of ways, were not tampered with in this election."