[EM] round robin tournaments RBJ

[Kevin Venzke]

Hi Robert,

--- En date de : Mer 29.6.11, robert bristow-johnson <rbj at audioimagination.com> a écrit :
> i'm selecting smaller segments to respond to.  i might
> select another small segment.

Alright, I was hoping you would comment on the 35 A>B 25 B 40 C
scenario.

> > Are there any other methods that have "salience
> attached to how close or
> > decisive an election is"?
> 
> sure, Ranked Pairs, Schulze, MinMax using Margins.
> 
> > It seems important to you but where does it come
> from?
> 
> i guess from Tideman or Marcus S.

Markus is surely pro-WV... He doesn't write about it much recently, but
you can read in his paper that he concludes WV should be used. When he
gives Schulze examples one should assume he is using WV.

> > Most methods don't even have a pairwise matrix, so do
> they
> > just all flunk your test, and you "just cannot get
> past" any method
> > besides margins? If not, I am curious where you find
> that other methods
> > have this kind of "salience."
> 
> i'll admit that i cannot get past the ranked-choice ballot
> (vs. the traditional vote-for-one, or the score ballot, or
> the approval ballot) and the Condorcet method of tabulating
> it.  i am less hung up on which Condorcet method, but,
> in the case of a cycle and given a choice, i can't see how
> decisiveness cannot be important.  if someone "wins" by
> just one or two votes (out of thousands), it says about the
> same thing as a tie or a loss.  the information we got
> from the voters is far less than that of a decisive defeat
> of another candidate.  it's the decisive defeat that
> should not have its result reversed in favor of the result
> of an indecisive defeat.

Not sure what else I can say to this as I was already pretty repetitive
in my last post.

In short, you are going to "respect" wins of candidates who can't win
the race, who generally will not want their wins to be respected.

Because the contests won't work in the way it seems they should (i.e.
the winner cannot benefit) they, as information, shouldn't be taken at
face value.

If pressing a button gives me an orange, look to the orange and not the
label on the button to see what it means when I keep pressing it.

And if a C>B win means A is elected.........

Kevin Venzke