Addition to earlier post

[Mike Ossipoff]

donald at mich.com writes:
> 
> Greetings,
> 
> I wish to make an addition to my earlier post today dealing with Electorial
> College reform.
> 
> I wish to increase the number of candidates in the runoff of top candidates.
> 
> The change is as follows:
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
> The candidates in the runoff will be the top two plus any other candidates
> that can win the national election if they gain all the electoral votes of
> the reformed states.

Very good; now you're partly copying the original proposal. Except that
this "runoff" would include the top 2 candidates (by whatever
measure you'd use) even if they couldn't win a the national election.
A little pointless to give all those electoral votes to someone
who can't win, in terms of the goal of the original proposal--
getting rid of the lesser-of-2-evils problem.

> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
> 
> I feel this addition is an imporvement when we consider that as the number
> of reformed states increase the number of possible winning national
> candidates also increases. For example: When the reformed states' electoral

An improvement over your previous proposal? Sure. But still doesn't
make the grade in terms of the goals stated when the original proposal
was posted.

> votes hit 135 it is possible to have three candidates also with 135 votes
> each in the UNreformed states. Some other possibilities are below:
> 
> Reform State Votes          Possible number of candidates
>                        One   Two   Three   Four  Five  Six  Seven  Eight  etc
>        135             135   135    135
>        180              90    90     90     90
>        203              67    68     67     68    67
>        216              54    54     54     54    54    54
>        225              45    45     45     45    45    45    45
>        232              38    39     38     39    38    39    38    38
>        etc             etc
>        etc             etc (now you are being technical)
> 
> With three or more candidates in the runoff I'm thinking that you pairwise
> people will want to use one of your single-winner pairwise methods - I see
> no reason not to. (now you really are being nice to these technical guys)

There never was a reason not to use one of the best single-winner methods
among the candidates who, with the reform states' electoral votes, would
have an electoral college majority. That's what we were saying all
along, Don. 

Steve's original proposal was to find the reform-states' winner, and
then ask if he can win nationally, and, if he can't, then pick the
reform states' 2nd winner, and ask if _he_ can win nationally, and
thereby give the electoral votes to the highest finishing candidate
(in the reform states' count) who can win nationally.

Of course this is very much like starting out by simply eliminating
all the candidates who can't win, even if they received the reform
states' electoral votes, and using the a good single-winner method
to choose from among the remaining ones (the ones wom the reform 
states can cause to win), to determine to which of those the reform
states will give their electoral votes.

This latter procedure is simpler to define & carry out. 

1. Find which candidates would have an electoral college majority
   if the reform states gave all of their electoral votes to him/her.

2. Use a good single-winner method to choose to which of those
   candidates the reform states will give all of their electoral
   votes.


Have we arrived at a compromise wording?

> 
> As I see it - when we have three or more candidates in the runoff the thing
> to do to reassign all the votes of the dropped candidates and then work the

If you're using a method that uses a single transferable vote. 
You're saying the count should be done by your favorite single-winner
method, Instant Runoff. Actually, the method used is a separate issue.
Besides, Instant Runoff was thoroughly trounced. Won't you let it
go?

> three or more remaining candidates with some single-winner method.
> 
> This addition of this increase in candidates will allow you to introduce
> your single winner methods at an earlier point in time.

Not earlier than we were before. In the initial proposal it was
used immediately. In my version (defined in this message, above),
it is used immediately after it's determined which candidates are
winnable--it's used to choose between them).

What could be simpler & more obvious than to use a good single-
winner method among the candidates determined to be winnable
in the national electoral college count. And then give the
reform states' electoral votes to the candidate who wins by
that method.


> 
> Donald
> 


Mike Ossipoff

> 
> .-
> 


--