Derek's correction: when does Vilfredo dominate Condorcet

[Pietro Speroni di Fenizio]

Derek sent me a mail regarding what I wrote about Condorcet vs Vilfredo. I could have kept it private, but I think it is good for everybody if we slowly move our communications about Vilfredo from private mail to this list. 

In my talk in Milan I claimed that if A dominates B according to Vilfredo than it also dominates it according to Condorcet. This is imprecise. Meaning that it is not always true (which should mean that it is false). But it is also not false.

If people only vote yes, no to each proposal. That is either vote in favour of various proposals or against, then we can still apply Condorcet. We treat all the proposals that have been voted the same as being... voted the same. Nothing in how we apply Condorcet voting systems requires that people strictly order the proposals. A partial ordering where proposals are either above or below is also acceptable. And if this is how people voted, then yes Vilfredo domination implies Condorcet domination. But.

But this is not how we usually vote when we are going to apply Condorcet. Usually people order all the proposals and the we calculate the graph of pairwise comparisons. But of course if people vote in this way, the a Vilfredo domination does NOT imply a Condorcet domination. 

Let us look at an example. 

Let us suppose we have 5 people that prefer A>C>B, Not only but all of them find C the least acceptable proposal, so if pressed to just say yes or no they will vote for A and C (and say no to B). Which would look like {A, C}>B

Then suppose we have other 3 people that prefer B>C>A. And again C is the least acceptable so they vote B and C. Which would look like {B, C}>A

The Vilfredo domination would say that C dominates A and B.

The Condorcet domination, on the other side would say that A dominates C (5 people prefer A to C and 3 prefer C to A). 

So according to Condorcet using the full ordering:

A>C (5 to 3)

C>B (5 to 3)

A>B (5 to 3)

According to Condorcet using the partial ordering:

C>A (3 to 0)

C>B (5 to 0)

A>B (5 to 3)

and finally according to Vilfredo C dominates B and A:

C>A

C>B

and A, B are non comparable.

So Vilfredo implies Condorcet on the partial ordering, but it does NOT imply Vilfredo with the full ordering.

While this sounds like bad news it is not. Because if you are not satisfied by the Vilfredo Pareto Front, and look for the Condorcet winner for Partially Ordering you will get one of the elements in the Pareto Front (or a subset of those). But if you apply Condorcet on the full ordering to the full set of elements you might get a winner which is not even in the Pareto Front.

But what if you first apply the Pareto Front, and then apply the Condorcet on the full list among the proposals in the Pareto Front? This might be better.

I would say it requires more studies. And I would say we should implement the possibility for people to range vote the proposals as soon as possible. So we can then look in various experiments what comes out.

I want to thank and congratulate Derek for correcting me. I should now correct the slides.

[Giovani Spagnolo]

this should be posted publicly in a Vilfredo "dev" blog :-) !

--
--
You received this message because you are subscribed to the Google
Groups "Vilfredo Develops in Athens" group.
To post to this group, send email to
vilfredo-deve...@googlegroups.com
To unsubscribe from this group, send email to
vilfredo-develops-i...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/vilfredo-develops-in-athens?hl=en
 
Vilfredo Develops in Athens is the mailing list for the developers of Vilfredo goes to Athens. The website can be found at
http://Vilfredo.org
the code can be found at
https://github.com/pietrosperoni/Vilfredo
On twitter we are @Vg2A
---
You received this message because you are subscribed to the Google Groups "Vilfredo Develops in Athens" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vilfredo-develops-i...@googlegroups.com.
To post to this group, send email to vilfredo-deve...@googlegroups.com.
Visit this group at http://groups.google.com/group/vilfredo-develops-in-athens.
To view this discussion on the web visit https://groups.google.com/d/msgid/vilfredo-develops-in-athens/a47da5b2-634c-4ddf-ba0f-073eccfbc2fc%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Pietro Speroni di Fenizio]

Hi Giovani, 

thanks for posting.

Should we have a dev blog at all? 

Usually developers speak through mailing lists and the blog are for public announcements.  Or am I missing something? 

Also this post is not much more technical than some of the latest we had on this mailing list. So I would be interested to know what you would put in the dev blog (I never run an OS project, I am learning :-) ).

Cheers,

Pietro

 

To view this discussion on the web visit https://groups.google.com/d/msgid/vilfredo-develops-in-athens/C4A4E0E6-C467-4261-84C1-8E5C026F66FF%40webyes.com.br.

[Giovani Spagnolo]

mailing lists are surely the way to go to discuss internally around issues, features, ideas, brainstorming, ecc...

a dev blog instead is the way to communicate to the outside world (also better SEO indexing) with release announcements, to bring new people in more easily because they read something interesting about the theories behind the software (for example your explanation below that, if happens only in a Mailing list, would be irrelevant after some time because hard to find and sunk in people's mailbox - also unavailable easily to others outside the ML), show code samples for API integration, or screenshots of working parts of the system explaining how it works related to why this should work this way, etc...

there are many dev blogs around, their use do not preclude using the ML to discuss "internal" issues.

see also some famous dev blogs and how what they publish is different from what happens in mailing lists:

http://googledevelopers.blogspot.it/

https://developers.facebook.com/blog/

https://dev.twitter.com/blog

http://make.wordpress.org/core/

giovani

To view this discussion on the web visit https://groups.google.com/d/msgid/vilfredo-develops-in-athens/CAMX8Vg9j-EBqWcdgefSCKJ9e3i1rS96zPYFNeyyV0EWfAB%2BYXQ%40mail.gmail.com.