election-methods@mailman.electorama.com

Technical discussion of election methods

View all threads

Re: [EM] Election-Methods Digest, Vol 144, Issue 11

KM
Kristofer Munsterhjelm
Sat, Aug 20, 2016 10:49 AM

On 08/16/2016 03:57 AM, Sennet Williams wrote:

Stop being ridiculous.  There is never only "three candidates" in a
contested IRV election.  Since Oakland, SF and Berkeley modernized to
IRV, there are generally 9-10 candidates.  (unless there is a popular
incumbent)  And all serious candidates campaign for the widest support
they can, from the center.)  (Unless they totally ignore the IRV system,
like Don Perata did in Oakland's first IRV election.)  He learned that
you cannot buy an IRv election.

Here's how to make a 10-candidate monotonicity problem example:

  1. Pick a 3-candidate monotonicity problem example.
  2. Add 7 other candidates and modify the ballots so that the new
    candidates are all eliminated before the 3 candidates in the original
    example.
  3. You now have a 10-candidate monotonicity problem example.

Since IRV is based on elimination, if there's a problem with IRV with k
candidates, there will also be a problem with IRV with n candidates,
n>k. So if there's a three-candidate monotonicity problem in IRV,
there's also a ten-candidate monotonicity problem.

On 08/16/2016 03:57 AM, Sennet Williams wrote: > Stop being ridiculous. There is never only "three candidates" in a > contested IRV election. Since Oakland, SF and Berkeley modernized to > IRV, there are generally 9-10 candidates. (unless there is a popular > incumbent) And all serious candidates campaign for the widest support > they can, from the center.) (Unless they totally ignore the IRV system, > like Don Perata did in Oakland's first IRV election.) He learned that > you cannot buy an IRv election. Here's how to make a 10-candidate monotonicity problem example: 1. Pick a 3-candidate monotonicity problem example. 2. Add 7 other candidates and modify the ballots so that the new candidates are all eliminated before the 3 candidates in the original example. 3. You now have a 10-candidate monotonicity problem example. Since IRV is based on elimination, if there's a problem with IRV with k candidates, there will also be a problem with IRV with n candidates, n>k. So if there's a three-candidate monotonicity problem in IRV, there's also a ten-candidate monotonicity problem.