[EM] Anyone got a good analysis on limitations of approval andrange voting? (long)

[Dave Ketchum]

On Nov 10, 2009, at 7:58 AM, Matthew Welland wrote:
> On Tuesday 10 November 2009 03:57:34 am you wrote:
>> Dear Matthew,
>>
>> you wrote:
>>> Suite of complicated systems that strive to reach "Condorcet"  
>>> ideals.
>>>
>>>   1. No regular bloke would ever trust 'em because you can't explain
>>>      how they work in one or two sentences.
>>
>> Well, here's a very simple "Condorcet" system which can easily be
>> explained in two sentences:
>>
>> 1. Let voters vote on candidate pairs, successively replacing the  
>> loser
>> with a different candidate, in a random order of candidates.
>> 2. As soon as all candidates have been included, the winner of the  
>> last
>> pair is declared the overall winner.
>>
>> This system is arguably the earliest example of a "Condorcet"  
>> system. It
>> was devised by Ramon Llull in the 13th century and was successfully  
>> used
>> for elections in monastaries. It is easily understood by the common  
>> man
>> since it resembles a procedure frequently used in child play.
>
> That is an excellent description. Thanks.

Gives a feel for something good enough for their purposes - if there  
is a CW the CW wins.

BUT, if there is a cycle, the best candidate can lose to another cycle  
member - so better make sure the apparent CW gets compared with every  
other candidate.
>
> I've always liked that specific system and think being Condorcet  
> compliant
> makes great sense in smaller elections or where a perfect choice is
> critical.
>
> Implementing it on a large scale seems tough and actually voting in  
> it seems
> tedious. If there are N candidates am I forced to make N-1  
> decisions? If
> there is a short circuit way to do the vote then it might be workable.

In voting you could start with thinking who you would approve of.   
Then vote for them, while ignoring the others that you like less.  Do  
any of:
      Approve them by giving them the same rank.
      Vote for the best as in FPTP.
      Rank them to show your preference for better vs lesser.
>
>
> The improvement over approval still seems marginal, especially in  
> large
> elections and I think the cost for implementing, tabulating and  
> voting is
> much higher,

Agreed the implementing costs - mostly in being able to do this.  With  
the ability you get a big help in some elections and less or little in  
others (like 0 when only two candidates).

Tabulating a ballot gets to be labor when a voter ranks many candidates.

Being able to determine winner from the N*N array was an implementing  
cost, but then easy to do via computer (assuming use of an easy  
variant of Condorcet).

Voters who were happy with FPTP will see no benefit - but no cost once  
they see they can vote as they have before.

Voters who have studied range/score will make two groups:  sad to be  
unable to express the exact size of their likes/dislikes; thankful for  
the easier decisions involved here.

Voters who want to rank higher those they like best will be thankful  
to get past approval.

Dave Ketchum
>
>> Yours, Jobst