[EM] RE : Re: simulator code for 2 Approval methods

[raphfrk at netscape.net]

From: stepjak at yahoo.fr
 > Thanks. Looking at that result, compared with other methods' results

> on your site, and noting that the original study obtained nearly the

> same results for Approval and Condorcet... I have to wonder if there

> isn't something wrong here. Is it possible that there are many ties

> occurring between green and red, that are being broken invariably in

> favor of red? If so we'll have to be more careful about that...



You could be right.  2 candidates could easily be above the threshold and

thus get exactly the same totals.  The function breaks ties the same

way every time.  The way to check it would be to take a random experiment

and reverse the order of the candidates.



In general, the simulation could be setup so that it runs 2 elections with

the candidates order forward and 2 elections with the candidates ordered

backwards instead of 4 ordered forwards.  If candidate order matters, 

then there would be areas which are solid mixed colour.



I think the 'poll' case should be alot better as it will allow the 

virtual voters to place their threshold better.

    

   Raphfrk
 --------------------
 Interesting site
 "what if anyone could modify the laws"
 
 www.wikocracy.com 
   
________________________________________________________________________
Check Out the new free AIM(R) Mail -- 2 GB of storage and industry-leading spam and email virus protection.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20061216/c387c46a/attachment-0001.htm>