[Election-Methods] RELEASE: Instant Runoff Voting (Chris Benham)

[Kathy Dopp]

On Sat, Jun 21, 2008 at 5:03 PM,

Sorry bout my impatience Chris, but I try not to waste time on stupid
ideas and I've already wasted over 6 weeks of this year considering
IRV which is an incredibly stupid voting method at first glance after
15 minutes of study IMO.  We cannot train our brains to solve the
tough problems of life and think clearly while wasting brain cycles
studying dumb ideas like the new voting method you just introduced us
to in your email - which I will precisely describe for you here so
that you can hopefully understand how little it has to do with either
range or approval voting methods:

1. Have voters rate the two candidates in the race from 1 to 100

> 40: A100, B98
> 25: A98,?? B1
> 35: B100, A1


2.  Vote counting method: Drop all but the top choices of voters who
have just rated the same candidates from 1 to 100 and give each
remaining candidate one approval vote.

> 65: A
> 35: B

3. The winner is the candidate with the most first choice "approval" votes.

4. Introduce a third candidate and have voters rate him as well, using
the prior ratings for the first two candidates, so this example of a
rating voting system is the same except for the additional candidate
for voter ratings.

>
> 40: A100, B98, C1
> 25: C100, A98,?B1
> 35: B100, C98, A1

5. Change the rules (or is the rule for your new voting method always
"approve a number of top candidates equal to the total number of
candidates minus one" for each voter?) and this time drop all except
the top two choices of voters and give the remaining candidates one
approval vote each.

> candidates, to give the Approval ballots:
> 40: AB
> 25: CA
> 35: BC

Voila. This hair-brained voting method DOES exhibit the spoiler
effect!  Good going Chris. While I will bet that you can invent any
number of hair-brained voting methods which violate the spoiler effect
like this

"voter-ratings-from-1-to-100-converted-to-top(N-1) candidate approvals
worth 1 each" where N is equal to the number of total candidates,

your example shows ZIP about the approval method. However, in your
example the two range voting examples (with and without the third
party candidate) show *no* spoiler effect. B wins both times.

PLEASE try to use the range or approval voting methods, rather than
inventing a new method that no one would ever think was a good idea,
as you did above, when you try to come up with an example which is
supposed to show how the range or approval methods are susceptible to
the spoiler effect.

I hope I have adequately described your method here so that you
understand that it is not the same as approval voting.

Please do not send us any more ill-considered emails Chris.  Some of
us have more important things to get done than to discuss hair-brained
voting schemes like IRV and the method you just proposed as being
equivalent to approval voting but which is not even close to it.

Cheers,

Kathy