[EM] Re: proper Borda implementation

[Paul Kislanko]

This is what is frustrating to people who come here from a different
perspective - what deBorda suggested at the same time Condorcet suggested
his method is entirely unambiguous about this, but evidently the name has
been mis-applied.

Borda's suggestion was that for each BALLOT a candidate be awarded points
according to how many alternatives the voter ranked lower than the
candidate. Clearly a ballot like 

A>B>C>D 

should be accounted 3 points for A, 2 for B, 1 for C and none for D.

A Ballot like 

A>B

should be accounted 3 points for A, 2 for B, 0 for C and 0 for D.

The "Borda score" (as de Borda defined it) would be a sum over all ballots
of the points assigned to each candidate. 

There's no reason to "invent" votes for C and D that the voter who cast the
second ballot did not tell you about. It is only Condorcet advocates that
require a completed ballot because in a Condorcet-based system incomplete
ballots can't be handled without some assumptions that may or may not
represent the voter's intent.

> -----Original Message-----
> From: election-methods-electorama.com-bounces at electorama.com 
> [mailto:election-methods-electorama.com-bounces at electorama.com
> ] On Behalf Of Curt Siffert
> Sent: Saturday, September 25, 2004 4:41 PM
> To: election-methods-electorama.com at electorama.com
> Subject: Re: [EM] Re: proper Borda implementation
> 
> 
> On Sep 25, 2004, at 1:49 PM, Dgamble997 at aol.com wrote:
> 
> > In a message dated 24/09/2004 20:06:37 GMT Standard Time, 
> > election-methods-electorama.com-request at electorama.com writes:
> >
> >
> >
> > [truncated borda ballots]
> >
> >  Any thoughts as to which is better?
> >
> >  What do people who advocate Borda as a serious method (i.e. Donald
> > Saari) say you should do in situations like this?
> >
> >  If there is a consensus of opinion amongst Borda supporters as to 
> > what you do regarding truncated ballots then a proper 
> implementation 
> > of Borda will follow that consensus.
> >
> >  David Gamble----
> >
> 
> I believe that's why he was asking here.  That said, I don't 
> think there's a right answer.  I've looked into this before 
> and Borda advocates usually say that Borda requires full 
> ballots, end of story.
> 
> There's also a third option - take the points that would 
> generally be awarded to the unlisted candidates, and 
> re-allocate them to the candidates you've ranked.  And there, 
> either give an equal amount to each candidate, or allocate 
> them proportionally.  And then there's the matter of choosing 
> the proportion.  If you rank 3 candidates out of five, would 
> the proportion be 3:2:1, or 5:4:3 ?
> 
> It's the same flaw as always with rated ballots, there's 
> never a 100% right way to normalize things.  You're either 
> choosing accurate voter intent, or equal ballot power.  But 
> out of all the above choices, I'd probably pick the 3:2:1 
> option.  I'm just thinking of the guy that only cares about 
> two candidates in a 100-candidate election; one he loves and 
> one he sorta likes, the rest irrelevant .  He'd probably want 
> his power to be allocated 2:1 rather than 100:99.
> 
> Curt
> 
> ----
> Election-methods mailing list - see http://electorama.com/em 
> for list info
>