[EM] nonmonotonicity of the combo method

[MIKE OSSIPOFF]

Markus gave an example outline, and then said:

>Could you please demonstrate that such a scenario isn't possible?

No, I can't do that for you. Your example outline sounds perfectly
possible, doesn't it?

But note that the combo wasn't one of my suggestions in my
most recent message. I said that if Tideman(wv) & BeatpathWinner were
the only methods under consideration, it would make the most sense
to simply vote between those two. And I said that, if those aren't
the only methods considered, then I'd suggest Voter's Choice.

I suspect that Tideman(wv) & BeatpathWinner are the most popular
rank-counts on EM. But if we were going to vote on what method to
use for subsequent votes on EM, I'd advocate Voter's Choice as thew
way to choose. Because, when we consider the list members who aren't
the most active posters, and when we don't limit our consideration
to rank methods, it isn't at all clear that it's only a choice between
Tideman(wv) & BeatpathWinner.

By the way, as I said in a reply to Demorep, I think there's a
question about how truncated rankings should be interpreted. The
voter should be able to indicate whether he means his ranking to
only compare its ranked candidates, or whether he also means to say
that his ranked candidates are better than his unranked candidates.

The question is which of those interpretations should be the default
interpretation. Surely it would be better if the voter didn't have
to actively make that choice. But what should it be taken to mean
if he says nothing about which interpretation he wants? In other words,
which is the natural one that deserves the benefit of the doubt?

Mike Ossipoff

_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.

Share information about yourself, create your own public profile at 
http://profiles.msn.com.