Slate has a good article discussing additional reasons why voters may have doublepunched their ballot, invalidating it. It's an informative read.

From the outside it looks like the DNC is pushing seperate avenues to challenge the Palm Beach vote: recounting and legal challenges. First they do the recount, then the legal challenges, and they've seemed to be kept seperate.

However, now that the first Palm Beach recount is complete, both parties want another recount to verify. The Republicans have petitioned for an electronic recount, while the Democrats are asking for a recount by hand. As I understand it, a by-hand recount wouldn't be a comprehensive tally, but a representative recount of randomly selected precincts, to verify statistically that there's no pervasive error in the electronic tally.

This most likely serves two purposes. One, as referenced in this eyewitness account of the Palm Beach recount is to see how pervasive the 'partial punch' problem is. Apparently those who voted by absentee ballot used less efficient tools for punching the cards, such that the inner part wasn't completely punched out. These cards are read by the electronic counter as not punched out. Each time they pass through the machine, they're more likely to have that bit pulled off, and will be accurately tallied on subsequent passes. Apparently, this is the primary cause for the increase in Gore votes in the Palm Beach recount. A manual recount could see how pervasive this problem is, and count these votes appropriately.

I speculate that the second reason the Democrats want a manual recount is to build a profile of the discarded 'doublepunched' ballots. by performing a manual recount, they can determine the nature of the discarded ballots, specifically, what portion of these doublepunches were punched Gore and either Buchanan or McReynolds, the two candidates immediately above and below on the punch-card. This would provide essential evidence for the upcoming legal challenges, allowing them to make the case that people misvoted and corrected, due to an unclear ballot, and that this error favored Bush, because there wasn't an upper hole to confuse on his portion of the ballot.

An unrelated bit that bears mentioning is the ordering of the ballot. The case has been made that the order that the candidates were presented was in keeping with Florida law (that the parties are listed inthe order of voting results from the previous election) because the ballot was designed to be read column by column, top to bottom. What's interesting to note is that the numbers on the ballot next to the arrows have a 3 for Bush, a 4 for Buchanan, and a 5 for Gore, clearly indicating an improper listing order.

That's all for now. Until next time, here's a comprehensive analysis of visual interpretations of the Palm Beach ballot (site temporarily overloaded), incorporating my own organizational schemas with those of other HCI scholars.

My cousin, Craig Fox, a professor at Duke's Fuqua School of Business, has independently written up a brief statistical analysis of the Palm Beach election returns.

Definitely worth a read.

In all the twisted theorized scenarios flying around, here's one I haven't seen yet:

Gore, through recount, revote, or other means, gets the Presidential vote. The Senate race in Washington goes to the Democrat. Now the Gore camp has a problem: If Lieberman is to be vice president, he has to abdicate his spot in the senate, and the Republican governor in his state selects a Republican to replace him, shifting the balance of the Senate back to 51 Republican, 49 Democrat.

Unless Lieberman voluntarily leaves the ticket, and the Electoral Voters are instructed to vote for a new Vice President. There's already been speculation on who Gore might choose for the role, but how about Bill Clinton? The 22nd Amendment states that a person can't be elected to the role of President more than twice, or more than once after serving two years or more as President (in the case of rising to the presedency from a lower office), but it doesn't outlaw a person from becoming VP, or even President, should he rise to that role by a means other than a federal election.

In this scenario Al and Bill would still run the White House, and the Democrats would hold a majority in the Senate, with Bill presiding and holding the tiebreaking vote on the floor.

I'm not advocating this position, I'm just bringing it up to bring to light yet another way this election could twist things away from what anyone had ever conceived.

One interesting endnote is that this wouldn't be without risk. It would only take a handful of Electors to decide that their constituents didn't vote for Gore/Clinton, and they still vote for Gore/Lieberman or Gore/??? instead. In the electoral college, the Presidential and Vice Presidential races are seperate, so while gore would win there, Clinton/Lieberman/??? might not pick up the majority. If Cheney didn't pick up 270 or more votes, which he wouldn't unless several Democratic Electors flipped and voted him for VP, then the Lame Duck Congress would vote, amongst the top two contenders, who would be VP. As it would be a Republican led congress doing the voting, it's probable that the final outcome would be Gore as President, and Cheney as the Vice President.

I have no idea what it is, but I have the strong feeling that the Bush Campaign has (or thinks it has) one more major card to play... First Florida was up in the air, then there's the recount, then the selected second recounts, then the overseas ballots, then the Palm Beach court challenge. This seems to be the current order.

But I think there's something else that's going to be inserted into this chain of events by the Bush camp. I have no idea what it is, excpet that it'll be something in Florida, and not the extra recounts the press says are being considered in Iowa and a few other states.

We'll see...

I've put together visual and textual explanations of three 'incorrect' ways that the Palm Beach ballot may have been interpreted by voters, resulting in miscast votes. I encourage you to look at it, comment on it, and pass the link on to anyone else who might be interested.

To those few who have flamed me for even talking about the possibility of a misvote, this sort of thing is the very crux of interface design. That's why I'm so keen on it.

If you have any doubt as to the effects the irregular ballot in Palm Beach has on the returns in that county, check this out. It's a data visualization applet that shows without question that the results in Palm Beach are not only over 4 standard deviations out of the norm (literally odds of 15000 to 1 of it being an accurate tally), but that the number of votes mistakenly attributed to Buchanan is between 2300 and 2600.

This isn't going to be over with the recount, or any time soon... Think OJ, think Lewinsky, then think the first major test of the constitution this century, which may ultimately result in the elimination of the Electoral College.

It's living history, and it's only just beginning...

An irregular and potentially confusing ballot layout in Palm Beach County, Florida, may have caused enough user error in the election to sway the outcome of the national election. Last I've heard three Palm Beach voters have filed suit in Florida court, and this may just be the beginning.

I cant wait to see what they do with this site.

It's not going to be over anytime soon... Even after a winner is declared in Florida, this might be the push needed to change the nature of voting from electoral to popular. Also, if Bush carries Florida, but Gore gets the popular vote, it's possible they might decide to move away from the electoral college early. It would only take a small handful of electorates to change their vote to the popular winner for Gore to win anyhow. Though very unusual (8 times this century out of thousands of electoral votes), it's not illegal for an electorate college member to vote against party lines.

It's not over any time soon... Arg.