[EM] Mass Candidates

[raphfrk at netscape.net]

From: Anthony Duff <anthony_duff at yahoo.com.au>

>--- Stephane Rouillon <stephane.rouillon at sympatico.ca> wrote:
>> any system that would leave a default value for 
unvalued/unranked/unapproved
>> candidates would help. Personnaly, I would suggest:
>> A) Let the voter precise the score, rank or state of all unexpressed
>> preferences;
>> B) I favor preference-style ballots over simple approbational 
ballots;
>> C) I favor ranking systems over scoring sytems.
>>
>> The reason behid B) is that I think preferences help to get a more 
sincere
>> result
>> because the added details are worth more than the strategical 
opportunities
>> in my humble opinion.  I wrote C) for the same reason.
>
>I don't see the logic of the last sentence.  A scoring system should 
include
>more
>details than a ranking system.  A scoring system can have an arbitrary 
level of
>precision, and can be reduced to a rank ballot.

It all comes down to how it is counted.

If I rank candidates

A>B>C

and use range

A: 100
B: 10
C: 1

I have the same ranking for both.  However, the first vote carries more
weight when comparing B and C for the ranked choice.

This means that something like renormalisation has to happen when
using range votes.  This is very labour intensive and would mean
a computer is required.

What range would be good at is when trying to decide if 2 candidates 
are better than 1.  For example, if the voting system was electing more 
than 1 candidate, the range vote would show the voter would prefer A to 
be elected instead of B and C being elected.  This might be helpful in 
some kind of declared preference voting system.