Loading ...
Sorry, an error occurred while loading the content.

Four IRV elections in 10-minutes!

Expand Messages
  • Tom Ruen
    Dear IRV supporters, I promised my coworkers I could count FOUR IRV polls in 10 minutes in our office Halloween party for best costume. We voted on four
    Message 1 of 1 , Oct 31, 2007
    • 0 Attachment
      Dear IRV supporters,

      I promised my coworkers I could count FOUR IRV polls in 10 minutes in our
      office Halloween party for best costume. We voted on four ballots for: Most
      original, Scariest, Funniest and Sexiest costumes. There were ~20
      candidates, and 17-19 ballots submitted for each election. I had one
      co-worker who helped count and we completed this task in 9 minutes 50
      seconds!

      The biggest problem with IRV with only a few ballots is ties for elimination
      order, so I performed a quicker top-two IRV, making piles of ballots by
      first choices, determining two strongest candidates, and recounting other
      ballots between these two choices. When faced by an elimination tie for
      second place, I added a mid-stage, redistributing all ballots for candidates
      BELOW the tied-for-second candidates, and this step fortunately that broke
      all the ties. The speed of counting proves how fast this process works!

      Incidentally, in 2 of 4 of these polls the "plurality first" candidate was
      narrowly defeated (by a single vote). In 3 of 4 elections it still wasn't a
      "majority choice" due to exhausted ballots, although no one lost a vote
      because of limited ranking space. (Ballots had 6 rankings and many people
      only marked a single favorite.)

      I suppose this example suggests a possible democratic principle: "If a
      majority of voters DON'T want to offer any choices below first, they have
      the political power (before the election) to overrule any runoff process and
      demand a single-vote Plurality candidate to win." Although it would be
      curious for a majority to WANT to "throw away their vote"!

      Anyway this shows hand-counting IRV is "almost as fast as plurality" and
      "slightly better" results - allowing everyone with a willing choice a chance
      to pick the winner. A "real runoff round" (second round) of voting between
      the top-two for each contest would have allowed a real majority preference,
      but at a cost of time (and patience) that was not desired for a fun
      election.

      Sincerely,
      Tom Ruen
    Your message has been successfully submitted and would be delivered to recipients shortly.