22 Mar 2005: Change to voting


From: Mark Goodge <m.goodge@ukvoting.org.uk>
Date: Tue, 22 Mar 2005 20:48:44 +0000
Subject: RESULTS: Change the Voting Guidelines FAILS
Newsgroups: uk.net.news.management,uk.net.news.announce,uk.net.news.config

-----BEGIN PGP SIGNED MESSAGE-----

RESULTS: Change the Voting Guidelines FAILS

Summary: Change the Voting Guidelines fails as "No" was preferred to "Yes" by 
31 votes to 12 and preferred to "Re-open Discussion" by 28 votes to 11.

Voting closed at 23:59:59 GMT 10 March 2005

Proponent: David Mahon <dmahonRFD@amigo.co.uk>

Votetaker: Mark Goodge <m.goodge@ukvoting.org.uk>
Secondary Votetaker: Richard Ashton <richard.ashton@ukvoting.org.uk>

Distribution: 
uk.net.news.announce
uk.net.news.management
uk.net.news.config

======================================================================

The full results follow below in the following order:

1) Information on the ballot
2) Results
3) Individual Vote Details
4) Votetakers Comments
5) Voting and Appeal Information
6) Rationale and Proposal (from the CFV)

======================================================================

INFORMATION ON THE BALLOT

Voters were asked to show their preference on the question:

   Replace the "Voting Procedures Within The UK Hierarchy" with
   the wording as given in the proposal

with three options on the ballot:

     A. Yes               
     B. No                
     C. Re-Open Discussion 

======================================================================

RESULTS

There were 49 vote emails received:

45 Valid and counted votes
 3 Valid votes subsequently superseded by the voter
 1 Invalid vote subsequently corrected by the voter

This was a multi-option vote. Multi-option votes in the uk.* hierarchy 
are decided by a two-stage procedure. All options are first compared with 
votes for the status quo, and any which are not preferred to SQ by the 
required margin of 12 votes are eliminated. The remaining options, if any, 
are then evaluated by the Condorcet method.

STAGE 1

Option B (No/SQ) beat Option A (Yes) by 31 votes to 12
Option B (No/SQ) beat Option C (Re-Open Discussion) by 28 votes to 11

Option B is therefore the winner. A full Condorcet count was not required,
but the tables are included here for reference:


Table 1 - Absolute Figures (Raw Preferences Grid)

     A     B     C 
A  [n/a] [ 12] [ 12] 
B  [ 31] [n/a] [ 28] 
C  [ 26] [ 11] [n/a]  

In this grid, each number represents the number of votes which were
cast that preferred the option named in the number's row to the option
appropriate to the number's column. For instance, 12 voters preferred
Option A to Option B while 28 voters preferred Option B to Option C.

Table 2 - Relative Figures (Opposite Comparison Grid)

     A     B     C 
A  [n/a] [-19] [-14] 
B  [ 19] [n/a] [ 17] 
C  [ 14] [-17] [n/a]   

In this grid, each number represents the number of votes which were
cast that preferred the option named in the number's row to the option
appropriate to the number's column less the number of votes which were
cast that preferred the option named in the number's column to the
option appropriate to the number's row. For example, 19 more people
preferred Option B to Option A than preferred A to B, and 17 more
people preferred Option B to Option C than preferred C to B.

The winning option in the absolute table is that where all the numbers
in its row are higher than the number in the equivalent position in
its column. The winning option in the relative table is that which has
all positive numbers in its row and all negative numbers in its
column. (On this ballot, this is Option B).


======================================================================
INDIVIDUAL VOTE DETAILS

Notes:

1. Voters were asked to rank the options from 1 (highest) to 3 (lowest). 
Equal rankings are allowed, and any unranked options are considered to be 
ranked below all ranked options.

2. Voters used a variety of different marks (or lack of them) to indicate
unranked options. Irrespective of the actual character used, all unranked
options have been represented in the table by '-' (dash).

3. The votes column shows the ranking given by the voter to each option.
For example:

               A B C
   voter 1     1 2 3
   voter 2     - 2 1

Voter 1 ranked option A first, then B,  and then C.
Voter 2 ranked Option C first, then B, and left the other option unranked.


Name                 Email                                    A B C
====                 =====                                    =====
Adam Price           adam+usenet?pappnase@co@uk               3 1 2
Alan LeHun           ukvote?lehun@clara@co@uk                 3 2 1
Alex D. Baxter       alex-news?oenone@demon@co@uk             3 1 2
Alex Holden          votes?alex-holden@co@uk                  3 1 2
Barry Salter         ukvotes?southie@me@uk                    3 1 2
Blippie              chris?horse-back@co@uk                   1 - -
Brett Paul Dunbar    brett?dimetrodon@demon@co@uk             1 3 2
Charles Lindsey      chl?clerew@man@ac@uk                     3 1 2
Clive Feather        clive?davros@org                         - - 1
Clive R Robertson    voteproc051?clive@org@uk                 1 2 3
Dave Fawthrop        hyphen?hyphenologist@co@uk               3 1 2
Dave Hillam          brillig?dsl@pipex@com                    - 1 -
Dave Johnson         requiem?freeuk@com                       3 1 2
Dave Mason           xctv?broonzy@org@uk                      3 1 2
Dave Mayall          dave?research-group@co@uk                3 1 2
David Mahon          dmahon_vote_xctv?amigo@co@uk             1 3 2
David Uri            daviduri?bigfoot@com                     1 2 3
Dick Gaughan         dickg?dickalba@demon@co@uk               3 1 2
Geoff Berrow         bl?ckdog@co@uk                           - 1 -
Graham Drabble       graham@drabble?lineone@net               3 1 2
Jonathan Allen       jonathan?barumtrading@co@uk              1 - -
Jonathan Wheeler     jfw1?rl@ac@uk                            1 - -
kat                  kat@news?ntlworld@com                    - 1 -
Lt. Cmdr. Jim        jim?us-lot@org                           3 1 2
Maurice              mbickers?netcomuk@co@uk                  1 - -
Mike Fleming         mike?tauzero@co@uk                       1 3 2
Mike Stanton         mike@stanton?dsl@pipex@com               1 3 2
MJ Ray               mjr?dsl@pipex@com                        3 2 1
Molly Mockford       xctv?mollymockford@me@uk                 3 2 1
Owen Rees            owenrees?waitrose@com                    3 1 2
paul                 {voter}?watman@clara@co@uk               - 1 -
Paul Cummins         paul?cummins@ie@eu@org                   - - -
Paul Harper          paul?harper@net                          - 1 -
Pedt                 pedt?dontspame@com                       3 2 1
Pete The Gardener    pete_the_gardener?hotmail@com            1 - -
Peter Duck           pduck?zetnet@co@uk                       1 3 2
Peter Robinson       pmrobinson?gmx@net                       3 2 1
Peter Smyth          psmyth?gmx@net                           3 1 2
Richard Clayton      richard?highwayman@com                   3 1 2
Robbie Irvine        robbie?babblings@org@uk                  - 1 2
Roger Hayter         roger?hayter@demon@co@uk                 3 1 2
Stuart O'Donnell     stuart@o'donnell?venicones@demon@co@uk   - 1 -
Thomas Lee           tfl?psp@co@uk                            3 1 2
Tim Sharrock         tim-nospam?sharrock@org@uk               3 1 2
Wm                   tcnw61?tarrcity@demon@co@uk              3 1 2

======================================================================
VOTETAKER'S COMMENTS

A fairly simple vote, with no major problems. A surprisingly large number
of people appeared to misread the voting instructions and vote as if this
were a first-post-the-post election, placing a character such as "X" or
the word "Yes" against a single option. As usual, these have all been 
converted to a Condorcet ranking of "1" for their favoured option and 
"-" for the others.

======================================================================
VOTING AND APPEAL GUIDELINES

This vote was conducted by a neutral third party member of UKVoting.
UKVoting is a group of independent votetakers who count votes on behalf
of the uk.* hierarchy and other 3rd parties.

The rules under which votes for the uk.* hierarchy are taken are posted
regularly to uk.net.news.announce or can be found at the following URL:
<ftp://sunsite.doc.ic.ac.uk/usenet/news-info/news.answers/uk/voting>

The UKVoting web pages can be found at <http://www.ukvoting.org.uk/>

There is a five day discussion period after these results are posted to
uk.net.news.announce.

Allegations of irregularity should be sent to control at usenet dot org dot uk.

======================================================================

RATIONALE and PROPOSAL from the last RFD

RATIONALE

The voting procedures have been written and added to over a period of time. Whilst containing most of the information required, they are perhaps overly long and complicated. They do not read well and could do with some general tidying up.

The only significant change I believe I have made is to give SQ a +10 bonus before the calculations begin. I think this makes the whole computation process easier than saying that each option has to beat SQ by 12 or more before then progressing to the Condorcet count. +10 seemed more logical than +11 IMHO, but it will, of course, be covered in the discussion (some people may advocate a higher or lower figure).

It is expected that this proposal may require more than one RFD and I encourage anyone who can think of any changes to submit them during the discussion phase. I also think the other documents could also do with tidying up (and pruning) but I don't really want to take on more than one at a time (it could also make the newsgroup rather busy).

PROPOSAL


Replace the "Voting Procedures Within The UK Hierarchy" with:

VOTING PROCEDURES WITHIN THE UK HIERARCHY

The following words where used in this document have the precise meanings shown here:

SHALL
any RFD which fails to follow this guideline is invalid

SHOULD
in all but exceptional cases an RFD ought to follow this guideline

MAY
whilst this guideline is accepted practice, it remains optional

These rules are provided for use when an RFD requires a vote. Hopefully most RFDs can be completed without a vote.

Votes SHOULD be conducted by a neutral third party.

The current group(s) of volunteer votetakers used for votes within the uk.* hierarchy, together with their email address(es) are:

UKVoting: ukvoting@ukvoting.org.uk.

The Vote

1. If after the discussion following a Request For Discussion (RFD) it becomes necessary to hold a vote, the proponent SHALL ask one of the votetaking organisations to formulate a Call For Votes (CFV). The CFV SHALL be posted uk.net.news.announce together with all those newsgroups and mailing lists that the latest RFD was posted to.

The CFV SHALL include:

      A summary of the discussion
      An indication of any differences between the final 
       proposal and the last RFD

      The rationale
      
      One of the following two sets of information:

      The name of the group
      The newsgroups line
      The charter
      (alternative versions of some or all of these MAY be offered)
      OR
      The proposed guideline change
      (alternative versions MAY be offered)

      The voting instructions
      The exact date that the voting period will end
      The ballot form, or alternatively, instructions on how 
       to obtain a ballot form

Only minor changes are permitted between the last RFD and the proposal being voted and these SHALL have arisen directly from the discussion.

The ballot form SHALL:

    Include the voting address
    State the closing date for receipt of the ballot
    State clearly the question(s) being put
    Be completely even-handed as regards voting FOR, AGAINST or ABSTAIN

If a status quo for the proposal exists, the voter SHALL be given the opportunity to vote for it, either by voting explicitly for the status quo or by voting against the proposal.

2. The voting period MUST last at least 18 days and MAY last up to 28 days.

3. A repeat of the CFV SHOULD be posted half way through the vote.

4. Only votes emailed directly to the official voting address SHALL count.

5. Votes SHALL be submitted on an approved ballot in accordance with the voting instructions.

6. Votes SHALL be explicit answers to the questions put.

7. Votes SHALL NOT be transferred to other, similar proposals.

8. Voters SHALL include their name and a valid email address, both of which are published in the result.

The votetaker SHALL determine the validity of each vote with respect to the voting instructions and MAY seek further clarification from the voter. If the votetaker determines that a vote is invalid, the votetaker SHOULD inform the voter.

9. Several votes for related groups MAY be included in the same CFV, provided the voter has the opportunity to vote independently for each one. A particular vote MAY be specified as being dependent on the result of an earlier one but decision trees of excessive complexity SHOULD be avoided.

10. For a vote between several mutually exclusive options, the voters SHALL be asked to indicate their relative preference amongst the given options, which SHALL include the status quo (if one exists). It is permitted to give the same preference level to more than one option.

11. An option to re-open discussion (ROD) SHALL be included the first or second vote in any ballot:

      Involving several mutually exclusive options
      If the committee so request
      If 4 or more people petition for it by e-mailing 
        control during the RFD stage
      
The ROD option SHALL NOT be offered on a third or subsequent vote on any one proposal.

12. Where the ballot paper has been sent via an auto-responder, or direct mail from the votetaking organisation, only ballot papers that have been requested by these methods SHALL be counted.

13. The votetaking organisation SHALL send a formal acknowledgement to the voter's email address within 5 days of receiving a vote, which SHALL include the ballot details, received. If a vote is missing required information, the acknowledgement SHALL indicate that the voter MAY resubmit his vote. It SHOULD be pointed out that the vote has NOT been counted in its current incomplete status.

14. The votetaking organisation or the committee MAY call a halt to the Vote if any serious irregularity becomes apparent. If it can be rectified, the votetaking organisation may recommence or restart the Vote.


The Result

1. At the completion of the voting period, the votetaking organisation SHALL post the result to uk.net.news.announce and to all the other groups or mailing lists that the original CFV was posted to. It SHALL include the names and e-mail addresses (suitably munged) of all the voters, together with which way each one voted, so that the results can be verified.

2. The Status Quo (if it exists) SHALL receive 12 bonus votes (votes preferring SQ to all other options) prior to calculating the result.

3. For a vote between multiple options, the Condorcet method SHALL be used.

4. In the event of a tie or Condorcet loop, the last RFD SHALL be posted again and the discussion re-opened.

5. If multiple votes are submitted using a single email address, only the last one SHALL count.

6. Where it is believed that several votes have been submitted by a single voter in an attempt to bypass these restrictions, those votes SHALL all be rejected.

7. There SHALL be a 5 day waiting period, beginning when the voting results appear in uk.net.news.announce, to allow for appeals. Appeals SHALL be sent to control@usenet.org.uk within this period. Following this period and after resolution of any appeals, control will implement the proposal.

8. All objections and appeals to the result SHALL be decided by the Committee. Their decisions SHALL be posted to uk.net.news.announce.


END PROPOSAL
===================================================================

-----BEGIN PGP SIGNATURE-----
Version: GnuPG-v1.2.4-(GNU/Linux)
Charset: noconv

iQCVAwUBQkCEsGOfGXkh8vHZAQFHhwP/Y/opskRUFarNoii59ppwp3eyTP97CnWb
4MmwQTyDqmLX2z0bekCEM48wE8fHbgYW32LllRaIkPkpbYC4D2fRA8oQ+dTzXbSK
hRffC4k43wWTTFZF95mad1Pi6gki2X8k6qD5UsV2bw156hCQNUWHWIsRNsyT21uI
8gOX8W2DfDw=
=F/P6
-----END PGP SIGNATURE-----


You can also see the raw article.
Back ot the UKVoting homepage