Modifying ballot papers (Re: [EM] Utilities?

Craig Carey research at ijs.co.nz
Mon Sep 6 01:41:07 PDT 2004


At 2004-09-05 22:57 +0200 Sunday, Jobst Heitzig wrote:
...
>............. It is a
>trivial task to design a ballot in which the voter can choose to either:
>  (i) mark one candidate as first choice or several as the top set, or 

That could extend the rules to the Block vote (k marks for k winners).

>  (ii) assign ranks or ratings to all or some candidates, or
>  (iii) answer a pairwise comparison question for every pair.
>
> Allowing voters to vote in a certain way is not requiring others to do
> the same!
>
...

I don't know what the word "others" is about.

It could be difficult to adjust the rules. 

The "ratings" part of (ii) might be excessively difficult (those are ratings
for preferences perhaps).

My guess is that the author does not want rules defending *fairness* at
this time; i.e. 'the time is not right'.

I rejoined very recently and I could have missed rules representing
fairness, that Mr Hetzig wrote. Not defending fairness is normal here,
and a sentence of Perot or Gore can distract.

Option (iii) would introduce cycles. So the method could auto-delete
cycles. But there would be parallel paths for the weight of the ballot,
to move through:

   A->P->Q->B, A->F->G->B

To remove parallel paths could result in the STV ballot design.

Fairness is imposed by rules.
MR HEITZIG perhaps missed the idea of simple rules defending goodness
worded up to be about perturbations (and any larger change).

---

Here is two weak rules that have to be rewritten to handle the new papers.

--------------------------------------------------------------------------
[1] "0 <= power"

   A loss of candidate A is sustained when any change in the subsequent
    preference list starting at and/or after preferences naming A, is done.

--------------------------------------------------------------------------
[2] "power <= 1"

   The win-lose state of both A and B is sustained for SOME x, with
   0<=x<=1, when this change to added papers is done:

   (pAB) --> x*(pA) + (1-x)*(pB)

--------------------------------------------------------------------------

That is simple, and the rules are differential -- so they have a good
chance of being ignored here at the Election Methods List.

---

Mr HEITZIG might ask some EML users of REDUCE to solve the problems.

Here is a Polish webpage that tries to save some bits of REDUCE before it
all spirals into the overcharging in Germany's zib.de:

   http://www.wcss.wroc.pl/wcss/biul/bi5.html (In Polish, with links)

Here is REDUCE's home page (available through telnet too):
   http://www.zib.de/Symbolik/reduce/






More information about the Election-Methods mailing list