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

interesting experiment

Expand Messages
  • brokenladdercalendar
    Check this out = http://betterpolls.com/do/337 this is only with 11 votes, but interesting. i did the same sets of 3-5 candidates, but with score voting and
    Message 1 of 1 , Sep 1, 2008
    • 0 Attachment
      Check this out => http://betterpolls.com/do/337

      this is only with 11 votes, but interesting. i did the same sets of
      3-5 candidates, but with score voting and irv. (only it gives you the
      results for all methods no matter which method was actually used. but
      you can look at the description for each "election" to see which data
      set is actually being referred to.)

      point is, we now should look at the cases where the irv and range
      voting winner differed, and then see how well those differences do in
      a head-to-head matchup.

      * first election: al gore, bill clinton, barack obama
      obama won with score voting _and_ irv, but there's still something
      interesting here. if you look at the extrapolations from the score
      voting election (e.g. converting it to ordinal data to see who would
      have won with instant runoff) obama basically owns it - though he ties
      with gore under the approval extrapolation. but if you look at the
      data from the irv election, gore wins the extrapolations for score
      voting, approval voting, and i.r.n.r. (Instant Runoff Normalized Ratings).

      * last election: 7 candidates, including john stewart and ron paul
      ron paul won the i.r.v. (tied with pelosi), and john stewart won the
      ratings summation.

      i think the data set is simply too small here to do anything very
      useful, but this would be an interesting idea to follow up on. it
      might be useful to do a meta-analysis of previous score voting and
      ranked voting polls.
    Your message has been successfully submitted and would be delivered to recipients shortly.