[EM] New tryIRV free IRV survey website online

[robert bristow-johnson]

On Jul 7, 2011, at 7:26 PM, Dave Ketchum wrote:

> Ouch!

i missed it.

> .     As Kristofer just wrote, Condorcet is a much better method  
> than IRV for what you are promising - Interesting that Condorcet  
> offers (more than) the same voter ranking capabilities as IRV, but  
> does much better counting.

i think the major argument for Condorcet is that it is the most  
consistent with the binary election of any pair.  isn't that sorta  
what Pareto efficiency is about?

we all agree how an election between only two candidates should be  
evaluated given equal weight between voters (that is the true meaning  
of "One person, one vote" and i'm still appalled that this slogan was  
used by the IRV-repeal people).  it should be no different if a third  
candidate is added unless that third candidate beats both A and B.   
there is no justification for why this third candidate should reverse  
the preference of the electorate regarding A and B.  if it's Condorcet  
compliant and if there is a Condorcet winner, then the outcome is no  
different than it would be if the CW runs against any of the other  
candidates.  the electorate, when asked and given equal weight to  
voters, say that they prefer this candidate over every other candidate.

> .     CIVS offers, available now, what you seem to be trying.   
> Recommend you study this description of CIVS and consider what it  
> offers:   http://www.cs.cornell.edu/andru/civs.html
>
> Dave Ketchum
>
> On Jul 7, 2011, at 10:25 AM, Sand W wrote:
>
>> I hope everyone is interested in a new online survey site intended  
>> to prove how much better IRV-enabled surveys are than traditional  
>> "one choice" or approval surveys.

can you provide a ranked-choice survey that is Condorcet compliant  
rather than IRV?

if your survey page has the ranked ballot that IRV uses, you can  
evaluate the survey by different methods.  why not give the users a  
choice?  some might pick Borda (cough, cough).

hey, this would actually be useful information for academic study.   
make the tools available (like in the website that performs the  
surveys) and the choice of several election methods, including  
traditional vote-for-one/plurality, Approval, ranked-choice (whatever  
Condorcet, IRV, Borda, Bucklin), and Score voting.  find out which  
ones are more preferred by users of the survey tools.

just an idea.

--

r b-j                  rbj at audioimagination.com

"Imagination is more important than knowledge."