Re: Numbers of ranked preference votes

From: Steve Chessin <steve_dot_chessin_at_sun_dot_com>
Date: Fri Apr 30 2004 - 03:18:02 CDT

>From arthur@kellers.org Thu Apr 29 13:33:03 2004
>Date: Thu, 29 Apr 2004 13:32:53 -0700
>To: voting-project@lists.sonic.net
>From: Arthur Keller <arthur@kellers.org>
>Subject: Re: [voting-project] Numbers of ranked preference votes
>Cc: voting-project@lists.sonic.net
>
>At 4:24 PM -0400 4/29/04, David Mertz wrote:
>>>>Looking at the numbers, it is extremely likely that with a
>>>>moderate number of candidates and slots, no two voters at a
>>>>precinct will vote the same way. In fact, it is possible for no
>>>>two voters at a state level to vote the same way (e.g. 7 ranks
>>>>among 16 candidates allows 63M different votes).
>>>two canvassing steps: (1) tabulating the vote lists to create
>>>counts of people who vote the same way, and (2) doing the "run off"
>>>algorithm based on the tabulations.
>>>(1) can be done incrementally, but (2) can only be done at the end.
>>
>>True. But it's quite possible the incremental step doesn't do much:
>>You will often wind up with a precinct reporting exactly one vote
>>for each preference-order. It's a tabulation, but kinda a trivial
>>one.
>>
>>Actually, in practice, you're right that a few combinations (like
>>only zero or one candidate ranked) will occur more often. But the
>>number of like-vote tallies moved up from a precinct won't be a
>>whole lot less than the number of raw votes. Which isn't to say not
>>to do it, just that it doesn't save all that much space.
>
>Wouldn't in help in combining multiple tallies from multiple
>jurisdictions? Suppose you did statewide IRV for a primary, for
>example.

You might still have to report the intermediate results by precinct, so
it wouldn't help.

This is what California Law currently requires in the Statement of Vote
(from
<http://www.leginfo.ca.gov/cgi-bin/displaycode?section=elec&group=15001-16000&file=15370-15376>):

15374. (a) The statement of the result shall show all of the
following:
   (1) The total number of ballots cast.
   (2) The number of votes cast at each precinct for each candidate
and for and against each measure.
   (3) The total number of votes cast for each candidate and for and
against each measure.
   (b) The statement of the result shall also show the number of
votes cast in each city, Assembly district, congressional district,
senatorial district, State Board of Equalization district, and
supervisorial district located in whole or in part in the county, for
each candidate for the offices of presidential elector and all
statewide offices, depending on the offices to be filled, and on each
statewide ballot proposition.

This is what we (Californians for Electoral Reform) would like to
see added:

   (c) For each office elected by IRV, STV, or any other election
method involving multiple rounds of counting, the statement of the
result shall show all of the following:
      (1) The number of votes cast at each precinct that are counted
for each continuing candidate in each official round of counting.
      (2) The number of exhausted ballots in each precinct in each
official round of counting.
      (3) The total number of votes cast that are counted for each
continuing candidate in each official round of counting.
      (4) The total number of exhausted ballots in each official round
of counting.

I believe that in Ireland, where they use IRV to elect the President
and count the ballots by hand, each county counts just the first
choices and then phones the results in to a central location. When all
the first choice results have been phoned in, the central location
determines who to eliminate and phones back all the counties to tell
them. The counties then do the transfers, compute new results, and
phone them back into the central location. They repeat this process
until they have a winner. So it isn't necessary to have all the ballots
in one location.

BTW, if you want to see one method of representing ranked choice
ballots, follow these links:
http://votingsolutions.com/irvtest/irvdata.txt
http://votingsolutions.com/irvtest/data.htm

They show the counting of the ballots here:
http://votingsolutions.com/irvtest/IRV%20Round.htm

You'll find similar information pertaining to STV here:
http://votingsolutions.com/camb2001/index.htm

Also, they're planning to GPL their code; see
http://www.fairvote.org/choiceplus/ for details.

--Steve
==================================================================
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
==================================================================
Received on Fri Apr 30 23:17:26 2004

This archive was generated by hypermail 2.1.8 : Fri Apr 30 2004 - 23:17:29 CDT