52 Names need no Elimination

[Mike Ossipoff]

Robert's Rules recommends Instant Runoff (referred to as "preference
voting") for postal balloting, probably because it's the only rank-
balloting count method they've heard of. They propose, for meetings,
Repeated Balloting, which, as I was saying in a previous post, is
a very good method, for something that isn't Condorcet's method.

As I was saying, Repeated Balloting is good for a meeting vote by
show of hands, when there isn't time for pairwise voting, but there's
time for Repeated Balloting. They, it seems to me, don't specify
any particular number of ballotings. Their rule is to just keep
taking Plurality votes till something gets a majority.

I'd modify that by changing the Plurality votes to Approval votes,
and specifying that the ballotings will continue till either 
something gets a vote total ar least equal to the number of voters,
or there have been N-1 ballotings (with N alternatives), whichever
happens 1st, and that at that time the winner is the alternative
with most votes.

Of course it should be pointed out to voters that it's surely to their
advantage to just vote for 1 in the 1st balloting, and then, perhaps,
if necessary, add one in each next balloting. As I said, this would
make this method very much like a show-of-hands Bucklin.

One problem could be in a large group where some are voting & some
aren't, it would be necessary to know how many are voting, whereas
it wouldn't if voting wre byk Plurality, & the simpler 1-vote definition
of "majority" were used (a majority of all the votes, rather than
at least half as many votes as there are voters). 

Also, if time is really short, that too could be a reason to use
Plurality instead of Approval. But that too has its problems, because
how are you going to make sure no one votes twice in a large meeting,
unless you write people's names down. I prefer Approval when there's
time for it, for the repeated ballotings.

And in meetings where there's more time, or fewer people &/or
alternatives, the show of hands vosting could be Condorcet///Approval
or BeatsAll//Approval, or the Smith version of those. 

BeatsAll//Approval comes into its own when it isn't feasible to
gain agreement about a circular tie-breaker.


Mike

p.s. I've decided that there's no reason to not delete Don's
letters as soon as they appear in the e-mail index, because
I don't believe that he's taken seriously, either on EM or
ER. That means I don't have to answer him, as I did recently
on ER. If he posts on ER again, I'll probably answer 1 more time,
but with a note that I'll just delete future postings from him,
and that therefore when they aren't answered that mustn't
be construed to mean that he's said something irrefutable :-)
Of course this, right now, is such a notice on EM.

I was going to do that before, but I was concerned that he
might be taken seriously on ER, but conversations since have
convinced me that there's very little danger of that.

Mike




--