[EM] The Coming California Single Seat Election

[Eric Gorr]

At 5:56 AM -0400 8/16/03, Donald Davison wrote:
>The Coming California Single Seat Election
>
>I think both Approval and Condorect would fail this election.

There is no logical reason to believe this that I can see.

>Approval would have a leading candidate, but without a majority of the voters.
>
>Condorect will need about ten thousand pairings, most of which will have
>too few lower choices in order to yield valid results.

With even a 135 candidates, there will be clear front runners among 
the voters and the vast majority of candidates can simply be ignored. 
The voting population knows this implicitly.

The only way that Approval would not pick someone with a majority of 
the votes would be if all 135 candidates would be divided into at 
least three polarized groups each of which capable of obtaining votes 
that would not go to at least one other group. It would be 
interesting if you could make a case for this.

Now, the reality of the CA election is that the vast majority of 
those candidates will not be taken seriously by (again) the vast 
majority of the voting population. Voters are simply smarter then 
that. I would guess, that, at best, there are 10 candidates that have 
any chance of gaining a significant number of votes.

Now, I am not following the CA election very closely, but it would 
seem likely (at the moment) that on the majority of Approval ballots, 
Arnold would be there and in the case of Condorcet, he would also 
place highly in the ranking of the majority.

The vast majority of those candidates would find themselves either 
un-approved or left unranked.

The wonderful thing about the CA election is that people are actually 
being given some genuine choices. The sad part is that because of the 
election system, they are not being allowed to express their 
preferences in any meaningful way.

>There's very good chance that the so called `Condorect Winner' will be the
>Pluralty Winner.
>I suggest using Irving, but that's me.

There's a chance the Condorcet Winner would share the winner among 
many various methods, Approval and Irving included. Unfortunately, 
with Irving, we can have no confidence that the person who won is the 
person who was indeed able to garner the broadest base of support - 
unless the results were also computed under some other election 
method.

Also, wanted to return to your prior example:

30:A>C>B
10:A>B>C
05:C>A>B
30:C>B>A
05:B>A>C
25:B>C>A

You claimed this to be a realistic example and I agree. What I also 
find most interesting is that C is the candidate who was able to 
obtain the broadest amount of support and would unambiguously win 
under RP Condorcet and many other election methods as well.

Unfortunately, what this also means is that the example fails utterly 
to provide any meaningful comparisons between various election 
methods.

I would be interested to see another realistic example from you that 
would select a different winner under Irving then it would RP 
Condorcet as judged by http://www.ericgorr.net/condorcet/. Once you 
developed such an example, an explanation of why your preferred 
method selected the correct winner while RP Condorcet selected an 
obviously wrong one would be rather enlightening.






-- 
== Eric Gorr ========= http://www.ericgorr.net ========= ICQ:9293199 ===
"Therefore the considerations of the intelligent always include both
benefit and harm." - Sun Tzu
== Insults, like violence, are the last refuge of the incompetent... ===