Believe in Your Data - Database Weekly (Aug 25, 2008)

  • Steve Jones - SSC Editor

    SSC Guru

    Points: 715053

  • mcostax

    Newbie

    Points: 9

    good article...

    I'm brazilian, and here's a link about Brazilian Elections (in english):

    http://en.wikipedia.org/wiki/Elections_in_Brazil

  • David Reed-223505

    SSCrazy

    Points: 2456

    Before we moved north, I got the opportunity to use a couple different electronic machines in Houston. I didn't like the way that the UI worked personally, but I imagine that they're simpler to explain to non-geeks than what I would design. Don't know who the manufacturer was, but the outcome of the elections was never in doubt in the districts I lived in, at least not for the federal offices.

    For a humorous take on voting machines: http://xkcd.com/463/

    Be sure to hover the image for the alt tag humor.

  • Anonymouse

    SSC Eights!

    Points: 915

    Voting with paper and pencil is still not bug-proof.

    Recently in the city I live in, we had an election where the votes were tied (no electronic voting machines). A recount was done, and candidate A won by 3 votes (out of thousands cast). Candidate B requested another recount. In this recount, Candidate B won by one vote.

    So at this point, election officials decided to count "spoiled ballots". These could be where the voter did not vote for a mayor, but voted for a councillor in their ward. IMO, these ballots are not spoiled; instead; the voter chose only to use a portion of their vote. This didn't solve the problem either because in the end, there needed to be another election to elect a councillor in this ward.

    So electronic voting may not be perfect - but neither is hand counting ballots.

  • David Reed-223505

    SSCrazy

    Points: 2456

    I've seen proposals for electronic voting widgets that print out the ballot for later stuffing into boxes, maybe some exist already, but like Penelope says, nothing's human-proof.

    😛

    Having been involved more than the average schmoo with party politics, I just think that most people are OK with the wacky hijinks that happen with our long-time friend Mr. Paper, but most luddites are still not comfortable with the newer, more fun wacky hijinks that can happen with technology. Survey says that most people don't think their vote matters in the long run, hence the high rate of folks who don't even vote, and even those who do vote are jaded about the possible outcomes.

    I've recently discovered that everyone in Washington state is apparently a permanent absentee voter and we all vote by mail. Hijinks indeed. Heh.

  • Steve Jones - SSC Editor

    SSC Guru

    Points: 715053

    I voted by mail as well last year. Seems to be more and more people that want to use that, don't have to make a trip to the polls. How can we be srue that's actually done correctly and counted? Maybe card readers are jus tthe best solution, and we recount if we're within some margin of error, like someone wins by < 2%.

  • webrunner

    One Orange Chip

    Points: 29863

    No voting system is perfect, but it seems that electronic manipulation of votes is harder to track than manipulation of paper ballots. That's just my gut feeling, though. I may be wrong.

    It made me ask myself, How long did it take for electronic hacking take off after computers came into common use?

    It just seems that we have just seen the tip of the iceberg regarding electronic voting manipulation and no doubt more is coming down the pike. In some cases, only circumstantial evidence such as a mismatch between the voting-age population count and the number of electronic votes tallied can give a clue that something has gone wrong.

    Well, that and seeing that an Access DB is being used. I couldn't believe that. At the very least, I would want to see:

    (1) SQL or Oracle used without access given to poll workers,

    (2) some process of encrypting the data,

    (3) writing the results to two places for bookkeeping, and only one person with access to both places (for accountability),

    (4) a list of everyone who is authorized to access the data and voting equipment.

    This list is off the top of my head and I am sure subject to much criticism, but I bet it would be a lot more secure than an Access DB. I haven't specified any usable detail for security, but I would design something around not giving access to anyone who is a poll worker, and instead having an independent, randomly assigned database specialist to assist with any problems using the system. I'd also opt for somewhat cheaper machines and have more of them, so that in case of a problem the whole voting line would not be delayed. Because I would want each reported problem that required tinkering with the machine or getting database help to be logged for security purposes and disclosure.

    Probably pie in the sky and full of holes and prohibitively expensive, but again, I think it's a better mindset than relying on Access for such a mission-critical system.

    Then there is the issue of which voting system is used to begin with. Even assuming complete data integrity, some have argued that certain voting systems are inherently unfair. For example, Gaming the Vote by William Poundstone. I haven't read the book, and I see from the Amazon reviews that there are some criticisms of it. But it's just another thing to think about regarding voting, even if it is not a data issue as such. Namely, how do you best capture the will of a people as to which candidate most of them want?

    Just my two cents,

    webrunner

    -------------------
    A SQL query walks into a bar and sees two tables. He walks up to them and says Can I join you?
    Ref.: http://tkyte.blogspot.com/2009/02/sql-joke.html

  • David Reed-223505

    SSCrazy

    Points: 2456

    I remember reading a post (I thought it was Schneier) on using cryptography, public key-style, to provide for anonymous authentication and repudiation of election results. When I originally read it, I kept trying to figure out how I would explain to my parents and siblings how to use the election decoder ring... Heh.

    Don't see what I remember on Schneier's website, so it must've been someone else. :hehe: Another false attribution to the cryptography god. He's complained about that recently, too.

    Turned up some interesting links while looking, though:

    Vendor misinformation in the e-voting world

    http://www.freedom-to-tinker.com/?p=1304

    Diebold Accidentally Leaks Results Of 2008 Election Early (I remember watching this originally on my TiVo)

    http://www.theonion.com/content/video/diebold_accidentally_leaks

    More voting machine news from Schneier

    http://www.schneier.com/blog/archives/2007/12/more_voting_mac_1.html

  • Longv-278915

    SSC-Addicted

    Points: 412

    It is incredibly difficult to fathom how someone could have been allowed to use Access for such an important application. IMHO, that fact alone tells you all you need to know regarding why the voting machines are not operating properly.


    Kindest Regards,

    VL

  • StarNamer

    SSCrazy Eights

    Points: 8633

    webrunner (8/26/2008)


    Then there is the issue of which voting system is used to begin with. Even assuming complete data integrity, some have argued that certain voting systems are inherently unfair. For example, Gaming the Vote by William Poundstone. I haven't read the book, and I see from the Amazon reviews that there are some criticisms of it. But it's just another thing to think about regarding voting, even if it is not a data issue as such. Namely, how do you best capture the will of a people as to which candidate most of them want?

    There was an article in the UK magazine New Scientist a few months back which discussed voting systems and pointed out that most a flawed. To quote from the start of the article:

    ONE person one vote is the mantra of democracy. And as Americans prepare to elect a new president this year, they'll be weighing up who to cast their precious vote for. Yet giving each citizen just one vote may not serve democracy's best interests. It can all too easily throw up a winner who in a straight fight with the runner-up would not be the majority's choice - surely a negation of democracy.

    So perhaps the goal of an acceptable voting system should be to elect the candidate which most voters find least objectionable rather than the one they want most - of course, in a 2 horse race this is the same thing, but with 3 or more candidates it may not be.

    Strangely, after looking at over 30 voting systems (which weren't all described), the author concluded that the 'fairest' voting system is probably the style used by YouTube and similar systems to 'vote' for videos or music. Basically, 'voters' assign a score from 1 to 5 to any 'candidates' they wish to vote for (and simply ignore the rest) and the candidate with the highest average score is declared the winner.

    Unfortunately, I can't see any political party rushing to advocate this system! 🙂

    In addition, the 'YouTube' system adds a whole new level of complexity and pretty much mandates electronic voting!

    Derek

Viewing 10 posts - 1 through 10 (of 10 total)

You must be logged in to reply to this topic. Login to reply