[EM] EM criteria discussions

[MIKE OSSIPOFF]

But, though there´s been nearly no interest in my criteria on EM, that 
doesn´t mean that they haven´t been discussed on EM. They´ve been thoroughly 
discussed here. Bruce Anderson, Markus, Blake, and Richard have been kind 
enough  :-) to industriously help look for faults in the criteria, often 
very valianly and tenaciously.  But without finding problems in the 
criteria.

Well, to be fair, Blake did find that one criterion that I´d been using 
didn´t act as intended. It was what I called the Stronger Defensive Strategy 
Criterion (SrDSC). It was something that had occurred to me relatively late, 
and I´d only been using it for a short time. I dropped it.

Other than that, all the criticisms were erroneous. Bruce had claimed that 
WDSC & SDSC, which require that a majority have a way of defeating someone 
under certain conditions, require that _every_ such majority be able to do 
so simultaneously. But that isn´t what those criteria say in their 
requirement.

There´s a TV show called "Survivor". I´ve never watched it, because there 
was always at least a crummy sitcom on another channel. But I´ve heard about 
its scenario. Say there are 3 people left on the island, and they have to 
vote someone off the island. You´d surely agree that each of the 3 possible 
pairs of voters has the power to vote the other one off the island. But is 
that the same as saying that all 3 of the possible pairs of voters  have the 
power to simultaneously vote the other person off the island?

But that was Bruces´s objecion to WDSC & SDSC. He´d first mentioned it in 
individual e-mail, and I explained that, as written, and as intended, WDSC 
and SDSC don´t make the requirement that he said that they make. I thought 
that was the end of that, but then, later, Bruce recycled and re-used that 
same objection on EM, though I´d answered that elementary misunderstanding 
previously in individual e-mail.

So I explained it again. But then, later, Markus found Bruce´s objection 
postings in the EM archives, and recycled them yet again. Maybe around 
September of ´97, ´98 or ´99. Maybe 2000.

Anyway, my point is that the defensive strategy criteria have been 
thoroughly discussed and throughly evaluated for faults, by four determined 
and strongly-motivated fault-seekers, over a period of years.

If I said that I haven´t communicated with academics, that was a lie. There 
were a brief few e-mails exchanged with Steven Brams. Professor Brams has 
been a dedicated and effective advocate of Approval, and I mean no criticism 
of him. We´re lucky to have him on our side.

But, having said that, in that brief correspondence, I mentioned some 
criteria by which Approval does better than some other proposed methods. I 
thought that would be of interest. Brams e-mailed back to say that the 
individual methods´properties aren´t the important thing. The important 
thing, he said,  is the impossibility theorems, like that of Arrow, which 
tell the limits of what _any_ method can do.

You can forgive my making that error, because I thought that that was why 
were advocating new voting systems--because of their better properties. But 
who am I, an uneducated pitiful amateur, to disagree with a professor. 
Professor says that properties of individual methods aren´t important, and 
so they aren´t important, right? And don´t you forget it.

You haven´t looked at much voting system literature imuch if you haven´t 
found articles in which voting system academics say some astoundingly 
ridiculous things. Both Bruce Anderson and Niemi have written that Approval 
is worse than Plurality. Niemi said that´s because Approval gives to the 
voter too many choices. I mean look, when the voter decides how far s/he 
needs to compromise in Approval, just as s/he must in Plurality, then s/he 
has the agonizing dilemma of whether or not to vote for the candidates whom 
s/he likes better than her/his compromise :-)  It seems to me that Bruce´s 
justification for the claim was the same or similar.

Riker recommended Plurality for U.S. presidential elections, because 
plurality preserves the 2-party system.

But another academic, whose name I don´t remember, only that it was a German 
name beginning with G, apparently hadn´t read Riker about that, because Prof 
G said that Approval isn´t needed, because we have a 2-party system.

Those people all have PhDs. Some teach university courses on voting systems. 
This uneducated pitiful amateur humbly suggests that maybe a degree doesn´t 
prove that someone can be allowed to cross the street without supervision, 
or that someone is an authority. Well, Russ is proof that a degree is no 
guarantee against pretentious ignorance.

Well, authority is really important to some people, and I respect that. 
Well, not really.

Blake´s main objection to my criteria was that one won´t find them in a 
journal. To Blake that means everything. When we were discussing margins vs 
wv, Blake felt that he was playing the trump card when he said that Tideman 
advocates margins, and Tideman is an economist.

Mike Ossipoff

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/