[EM] Margins vs. Winning Votes

[Juho Laatu]

Hello Eric,

On Jul 27, 2005, at 00:27, Eric Gorr wrote:

> Dave Ketchum wrote:
>
>> Remember that the topic is ties, rather than splitting up a district 
>> with a fixed quantity of real voters.  The district could have had 
>> 3000 real voters in 2 groups of 1500 or 3 groups of 1000 - or 
>> whatever made the desired example.
>
> I fail to see the significance of these examples. Pretend, for the 
> moment, that the odd voter did not exist and the election ended in a 
> genuine tie.
>
> I fail to see how a randomly selected winner (the most common tie 
> resolution method) could be any better or worse then a single voter 
> casting the deciding ballot...which is another common tie resolution 
> method and used in the U.S. Senate, for example.
>
> It is quite beyond me why anyone would find this odd should it occur 
> in a genuine election.

The original reason for writing these examples was to study the 
behaviour of margins and winning votes. I don't know if you already 
took position on that part when discussing the significance of these 
examples. From margins vs. winning votes point of view the question is 
if the last voter is able to pick one of the _best_ candidates of the 
"1000 voter parties" or any of the candidates.

1000:  A>B>C>D
1000:  E>F>G
1000:  H>I
1000:  J

In this example, under margins the last voter can pick one of A, E, H 
and J (i.e. one of the #1 candidates of each party, others would need 
many more votes). Under winning votes the last voter can pick one of A, 
B, C, D, E, F, G, H, I and J (i.e any of the candidates). The question 
thus is if it is acceptable that winning votes doesn't put any weight 
on the unanimous opinion on the order of candidates set by the voters 
of each party to the candidates of that party.

Regards,
Juho