[EM] Poll Ballots, from poll-topics poll

MIKE OSSIPOFF nkklrp at hotmail.com
Thu Apr 5 16:53:17 PDT 2001



Rob--Thanks for counting the rankings for Cloneproof SSD & Ranked Pairs!

So Voting Systems wins by every designated method. And when each
ballot that has designated a method gives an automatic Approval vote
to that method's winner, and to everything ranked higher on that ballot,
and when any ballot not designating a method gives (by default) an
Approval vote to the alternatives marked on its Approval ballot, then
Voting Systems wins that final Approval count.

Alright, then even if I still haven't found out how to run a program
on my computer by 4 weeks from tomorrow, you can do the count for
pairwise count methods that are designated in the Voting Systems poll.

I'll post the details of the Voter's Choice final Approval count
later today.

Here's what I was suggesting last night: Let's start the nominations
and campaign discussion for voting systems today, and then, exactly
2 weeks after midnight tonight, April 5, California time, let's say
at that time nominations are concluded and the balloting period begins.
Let's say the balloting period is exactly 2 weeks.

So the nominations deadline is the midnight that ends April 19th,
California time. At that time the balloting period begins. The
balloting deadline is the midnight that ends May 3, California time.
That's a 2-week period for nominations & campaign discussion, and
a 2-week balloting period.

Now that I'm assured that pairwise count methods will be counted,
the Voting Systems poll will be exactly like the Poll Topics poll,
and any method can be designated. The default designation is "Manual".

I'd like the Voting Systems poll to have 3 minor differences, if no
one objects:

1. I should specify that if the method that you've designated returns
   a tie, then, instead of using Random Ballot to solve that tie, your
   ballot simply gives an Automatic Approval vote to each of the winners
   that your designated method has returned, and to everyone whom
   you've ranked higher than at least one of those winners.

2. Random Ballot is used if the final Approval count returns more than
   one winner. I'd like to do Random Ballot in a more random way than
   the way that I specified (but didn't have to use) this time:

   Each letter of the alphabet is represented by a number, so that A is
   1, B is 2, and Z is 26.

   Add up the initial letters of all the last names on the ballots.
   Subtract from that sum the largest multiple of 26 that is less than
   that sum. Multiply the resulting number by 10, and subract from that
   product the largest multiple of 26 that is less than that product.
   Call the resulting number "K".

   For each ballot's last name's 2nd letter, add K to that
   2nd letter, and subtract from that sum the largest multiple of
   26 that is less than that sum.

   The ballot for which the resulting number is largest will be the
   tiebreaking ballot.

   If that ballot ranks 2 or more of the winners equally, then repeat
   the 2 paragraphs before this one, but for the 3rd letter of each
   ballot's last name instead of the 2nd letter. Use the same K.
   And in general, if the use of the nth letter of the last names doesn't
   return just one winner, because it equally ranks 2 or more of the
   winners, then use the (n+1)th letter of the last names.

   If it's necessary to repeatedly go to using the (n+1)th letter of
   the last names, and if some of the names don't have an (n+1)th
   letter at some point, then for those names, the 1st letter becomes
   the next letter.

3. I was saying that we're on an honor system to not make use of
   information about previously-posted ballots to do offensive
   order-reversal. Let's say that we're also on an honor system to
   vote sincerely in CR if we vote a CR ballot. That means if someone
   doesn't vote a CR ballot, we don't infer a strategic one from their
   Approval ballot, and we just don't count them as voting a CR
   ballot.

   In a public political election, people would soon begin voting
   strategically in CR, and there's nothing wrong with that. But
   here, the reason for CR is to find out how its winner differs from
   those of the other methods, and it's of more interest to get a
   sincere CR winner than a strategic one that merely duplicates the
   Approval result.  And if some are strategizing in CR (as I did
   in the poll-topics poll), that isn't fair to those who choose to
   not strategize in that method.

As I said, I'm going to post the details of the Voter's Choice final
Approval scores later today. You can easily check the Approval and
CR counts, based on the posted ballots, to verify that, as I said,
Voting Systems wins by Approval & CR. Rob reported that Voting Systems
also wins by Cloneproof SSD & Ranked Pairs(margins).

Mike Ossipoff




_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com



More information about the Election-Methods mailing list