[EM] NPV vs Condorcet

[Raph Frank]

On Wed, Oct 22, 2008 at 6:39 AM, Dave Ketchum <davek at clarityconnect.com> wrote:
> Wile an Approval ballot could be recorded as if a Condorcet ballot, its
> information could not be reconstructed from state total election counts
> (this topic was part of noting that FPTP counts are different).

Ahh, I see.

I wonder if a State wide condorcet ballot would be justified in the
case of approval.

>>
>> If States with 40-50 EC votes (and a reasonable balance of Rep/Dem
>> States) joined, they would swing every election, unless it was a
>> landslide.  I doubt a non-condorcet winner would be able to landslide,
>> so it should not be a major disadvantage for anyone.
>
> ???

One of the issues that have caused a NPV veto is the fear that the
State will vote against its majority (because the national majority is
for the other party).

The benefit of the 270+ rule in the compact is that the result will
always go to the majority winner, so neither party is disadvantaged.

My thoughts are that unless it is a landslide, giving 50 EC votes to
the NPV winner should result in him winning and only a NPV winner
should be capable of a landslide anyway.

Looking at previous elections where NPV winner didn't match the final winner.

2000:  271-266
1888: 233-168
1876: 185-184
1824: 84-99-41-37

In 2000 and 1876, a 50 EC compact would have swung the vote to the NPV winner.

In 1888, the compact would have needed 65 votes (out of a smaller EC).

In 1824, the democractic-republicans ran 4 candidates for some reason.
 The House decided the race between the top 3.

In the 2 of the 4 cases where the compact would be needed, a 50 EC
compact would change the vote to match the plurality winner.

In one of the cases, an additional 15 votes would be needed and in the
final case, party strategy was to circumvent the EC entirely.

This means that even a small compact would give most of the benefits.

In fact, a compact between California and Texas would probably be
sufficient alone.


>> On the amendment, calling a convention could be used to prompt Congress.
>
> Dangerous - you might succeed.
>
> Threatening to call a convention could be productive.

Right, that seems to be how previous convention calls have worked.

> Note that such scaling could be applied to the contents of N*N arrays.

Yeah.

Also, under the statistics for the Penrose method, small States are
actually discriminated against even with their +2 boost.