Mike an Ebbie,

Two states split their electorial ballots based on the popular vote. 25 or so of the rest give all the votes to whoever won the popular vote in the state. The rest don't specify but generally vote forwhoever won the State.

After the Fla. Supreme Court decison, NPR had an interesting analysis of what might happen if the recount was delayed for too long and the Florida legislature certifed electors favorable to Bush. Then the recount gives the State to Gore and the Court certifies a slate of Democratic electors. It all goes to congress to sort out. If they can't decide on which slate to accept, the ultimate authority is the State executive charged with certification of the vote: In Florida's case Sec of State Kathleen Harris. Just to confuse things even more.

More fun facts: Florida law requires that a ballot be counted if the intent of the voter can be determined, and provides no guidelines or standards. Texas law is very specific and includes dimpled or pregnant ballots.

There's an arguement that the Electoral College can meet, vote and elect a President without any electoral representation from Florida.

I haven't found any hard statistics on hand counts. I know that there were a number of studies based on data entry and bar codes. Manual data entry yielded one error per thousand keystrokes, bar codes yielded one error per million. Not sure how that translates into manual counts, plus it dealt with single entry point, not redundant counts.

Neither manual nor automated counts resolves the core issue: that in Florida (and a lot of other States) there are no uniform standards for determining the "intent of the voter". The Canvassing Board in my County (that uses an optical system) is now debating standards for future elections. How much of the circle has to be colored in before the "intent" is clear. What if a voter uses an "X"? Or a check mark. Neither will be picked up by the scanner.

Regards

John