Thursday, January 31, 2013

Proposal: Various Fixes

Timed out and failed, 4-6. scshunt loses 2 credibility, and scshunt loses 1 more Credibility for voting FOR. Non-voters Clucky, Murphy, Koen, and robo1995 lose 1 credibility. RiachuKFM, Skju, and Spitemaster don’t lose credibility as their FOR votes came after the first occasion at which this proposal could have been adminned. Josh

Adminned at 03 Feb 2013 06:08:54 UTC

Amend the rule “Party Consensus” to read:

A Debater of a Proposal is an Honourable Member other than its author.

A Party is Eligible to Hold Consensus over a given Pending Proposal if it is the Party of the author or if it is In Power, and if at least one Debater in that Party has voted on that Proposal. Among all the Parties Eligible to Hold Consensus over a given Pending Proposal, the Party with the largest absolute value of (F - A)/T Holds Consensus over that Proposal, where T is the total number of votes from Debaters in that Party, F is the total number of votes FOR from Debaters of that Party, and A is the total number of votes AGAINST from Debaters in that Party. Votes of DEFERENTIAL do not count in this calculation. In the event of a tie, then the Party of the Proposal’s Author Holds Consensus over the Proposal.

If a Party Holds Consensus over a Pending Proposal, then all votes of DEFERENTIAL cast on that Proposal are instead considered to be valid votes FOR or AGAINST the Proposal, whichever is more common among the valid votes of Debaters from the Party Holding Consensus. Ties are broken in favour of FOR. This effect applies regardless of the Speaker’s vote.

Currently the consensus rules don’t work at all due to a number of confusions. This brings them into line with what I believe is the intent in the written behaviour.

[Amended by the Deliberation ‘By Popular Demand’ -scshunt]

posted by scshunt at 31 Jan 2013 03:50:53 UTC Comments (23)

Comments

Spitemaster:

31-01-2013 05:41:34 UTC

for I like it.  This is much clearer, and I don’t think it changes anything significantly.

Josh: Observer he/they

31-01-2013 08:15:35 UTC

for

Purplebeard:

31-01-2013 11:37:41 UTC

for

RaichuKFM: she/her

31-01-2013 12:04:21 UTC

Wait, if there’s 7 party members, 3 For and 4 Against they’d have a score of 7/(3-4)=7/1=7, whereas 7 For and 0 Against gives a score of 1, which is smaller. Shouldn’t the T/(F-A) be flipped? Changing “larger” to “smaller” in the proposal wouldn’t work, as 5/(2-3)=5, which would be better than a larger Party equally split, not to mention that ties lead to division by zero, which is arguably larger than infinity. Only problem i can see with flipping is it’ll be either one or a fraction. Wait, but then smaller Parties at Consensus are better than larger Parties stuck at the same net value. So I’d be for this, but the fraction needs to be fixed, and having it patched after passing wouldn’t be the best of ideas in my opinion.  against

nqeron:

31-01-2013 14:43:43 UTC

against per raichu

The mathematics on this is off.  The intent is to ensure that the more significantly member of a party votes one way, then the more you should bring the others in line.  This does the opposite.  The greatest absolute values come when |F-A| is minimized, rather than maximized, as it should be.  It should probably be |F-A| / T.  This way, the greater the polarity, the higher the number.

Josh: Observer he/they

31-01-2013 14:46:30 UTC

CoV against

Purplebeard:

31-01-2013 14:53:55 UTC

against

Larrytheturtle:

31-01-2013 15:42:15 UTC

against

Spitemaster:

31-01-2013 16:08:55 UTC

against CoV per Raichu

scshunt:

31-01-2013 16:17:56 UTC

With respect, that’s a separate fix. The intent was to preserve behaviour; enacting this won’t make things any worse and a proposal to change the formula would be a different matter.

Patrick:

31-01-2013 23:26:56 UTC

against

Klisz:

31-01-2013 23:33:30 UTC

against

Skju:

01-02-2013 01:00:08 UTC

This is good. Thank you for knowing how to say things. I’ll vote for it once the Deliberation passes.

scshunt:

01-02-2013 21:16:30 UTC

Please CoV now.

RaichuKFM: she/her

01-02-2013 21:32:37 UTC

for Boop!

RaichuKFM: she/her

02-02-2013 01:22:51 UTC

Everyone else? Anyone else? It’s been fixed.

RaichuKFM: she/her

02-02-2013 01:58:48 UTC

I figured it out- the best value is T*(F-A). This one still works, but T(F-A) rewards more consensus and larger Parties, whereas the fractional ones only reward one of the two. This fraction isn’t nearly as bad as the other one, however, and I apologize for finding the error and proposing another broken solution. I also wrote in my original comment that that wasn’t perfect either, but I couldn’t think of a better one. I’m still for this, but unless it gets votes and fast I’ll have to switch. I’m going to make a patch now, though its too late for a deliberation.

RaichuKFM: she/her

02-02-2013 03:13:21 UTC

against Sorry.

Skju:

02-02-2013 03:39:31 UTC

for

RaichuKFM: she/her

02-02-2013 04:03:46 UTC

for

Spitemaster:

02-02-2013 05:37:47 UTC

for

Josh: Observer he/they

02-02-2013 08:13:48 UTC

CoV for

Josh: Observer he/they

03-02-2013 10:07:04 UTC

CoV against