The strange, nasty US federal election has not ended; it has just become stranger and nastier, but in a way that none of us expected. The election of the next president of the United States may hinge, oddly enough, on user interface design.

While Bush may squeak out a victory in the current Florida recount, some people in Palm Beach are rumbling about the way their ballot was set up. Take a look yourself, and see how the South Florida Sun Sentinel.com illustrated the ballot confusion.

Even describing the problem is a bit of a challenge: The second hole on the right does not correspond to the second candidate on the left (Gore), but rather to the first candidate on the right (Buchanan). Glancing at the ballot, one is not sure which circle corresponds to which candidate; do you follow the line, the arrow, or does one choose the circle closest to the name?

AskTog has another analysis of the ballot. Voters "were interested in one candidate, the one they wanted to vote for. Their entire focus was on finding that candidate and punching the hole next to his or her name. In the case of Gore, that required scanning only two names down in the first column. There was never any reason at all for Gore voters to ever even see the right hand column."

Usability fundementalist Jakob Nielsen, the author of Designing Web Usability, writes on his home page: "The Florida ballot clearly had usability problems, caused by the attempt to map a two-column set of labels onto a one-column action area... A direct mapping between two single-column areas would have been much less error-prone."

The result: many Palm Beach voters claim that they voted for Buchanan when they thought they had voted for Gore. A statistical analysis seems to back this up. Till Rosenband at MIT determined that the "probability that Buchanan would get so many votes in Palm Beach by pure chance is less than 1 in 3,000,000,000,000,000." And if you look at the graph, you can see that Palm Beach is outside the expected bell curve.

Even Pat Buchanan is saying that the votes in Palm Beach aren't his. About 20% of Buchanan's votes in Florida came from Palm Beach county. In yesterday's discussion about this on Metafilter, some suggested that if people were confused by the ballot, it was their own fault, that they should have read the ballot more carefully. But the only test of whether something is confusing or not is whether or not people find it confusing. There were over 19,000 double-punched ballots.

I can read the ballot. But I have to say that the Palm Beach ballot seems typical of a lot of American ballot design I have seen: The form is designed for the machine to read, not for people to use intuitively.

My partner is a US citizen who voted by absentee ballot in Oklahoma. Different ballot, similar problems, including a layout that placed the marking circles in different places depending on the page, switching from columns to rows. It didn't take us long at all to figure it out, but the same might not be the case for a senior with poor vision.

Dan Bricklin, the creator of Visicalc and Trellix, agrees. His site offers a detailed analysis of the ballot's flaws, and asks, "I wonder what the usability testing for this was like, or even if there was any such testing."

An interesting question remains: how did other usability issues involving other ballots effect voting outcomes not just in other parts of Florida, but other states as well? What will be the result of legal challenges to ballots in those states that Al Gore has won? - Discuss at Metafilter