Clarification of Proportional Representation Act

[windjammer]

Clarification of Proportional Representation Act

4. Where candidates are tied they shall be ranked by the total number of second preferences votes in the preceding round received and so on. Where this method proves indecisive and the tie must be broken to materially progress the count, a Condorcet count between the tied candidates shall be decisive.
(a) Where candidates are tied at any stage in the count, they shall be graded according to their status on the preceding count and so on. Where candidates are tied at each such stage, they shall be graded according to their total number of second preferences.
(b) Where still tied, by a similar analysis of their third preferences and so on.
(c) Where this method proves indecisive and the tie must be broken to materially progress the count, a Condorcet count between the tied candidates shall be decisive.
[/quote]

Sponsor: Talleyrand

[windjammer]

Talleyrand?

[Talleyrand]

Sorry, hadn't seen this up there.

This is a fix to an inconsistency in counting results for at-large elections.

Basically, this will allow ties at various stages in each round to be broken by the total number of votes in the previous round and so on, rather than trying to break them based on the added total number of second preferences or anything like that.

If that failed too, we'd be headed for a Condorcet count.

[Southern Senator North Carolina Yankee]

Shouldn't "c" be "b" with the old "b" being eliminated?


Also, what problems have arisen under the present system that this would correct?

[Talleyrand]

Shouldn't "c" be "b" with the old "b" being eliminated?


Also, what problems have arisen under the present system that this would correct?

Fortunately, we've avoided this issue in the federal at-large elections, but we had a crisis in the Northeast Assembly election last month over how exactly this was to be implemented.

As for the b/c thing, I think the speaker can just suspend the rules and fix that.

[HagridOfTheDeep]

I like the way it is now. If it's a tie, the candidate with most first preferences wins. If that's also a tie, we move onto the second preferences. Etc. That's how I've always interpreted it, and it makes sense to me...

[Lief 🗽]

Can someone explain like I'm 5 what the difference between this and the old rule is?

[Talleyrand]

I like the way it is now. If it's a tie, the candidate with most first preferences wins. If that's also a tie, we move onto the second preferences. Etc. That's how I've always interpreted it, and it makes sense to me...

The reason that doesn't make sense is because while preceding counts do indicate support for a candidate, second or third preferences do not do that, especially since votes do not match preferences at different stages of the count, if that makes sense.

For example, in an election with 20 candidates, some voters' second preferences might not be used at all, if one candidate already starts off with a strong base of support, while other voters' may eventually go down to their 15th preference, in a particular round. As the law currently stands, if there were a tie that could not be broken in the 4th round by preceding rounds, you'd have to head to second preferences (even though they're already distributed all over the place) and then third preferences and so on if that proves inconclusive.

A Condorcet count between the two candidates (or the the elimination of all candidates besides those involved) as a tie breaker would be a far more convenient, fair way of settling such elections if preceding counts did nothing but deadlock.

Can someone explain like I'm 5 what the difference between this and the old rule is?

Currently, there are three steps to breaking a tie in the PR act.

1. Preceding Rounds.
2. Descending Preferences.
3. Condorcet Count.

Basically it proposes eliminating the 2nd factor and just having 1 & 3. 

[Lief 🗽]

Thank you for the explanation. I support this.

[Blair]

I think this shows, with the current climate we really need to get some form of electoral reform passed

[Fmr President & Senator Polnut]

I'm prepared to think about this.

[Cranberry]

From what I have seen being presented in debate, this is something I am prepared to support.

[Senator Cris]

I'm still undecided about it.

[Southern Senator North Carolina Yankee]

How did the situation in the northeast come about and how likely is such to be an occurence? It is a rare fluke or is there a more systemic problem here that could repeatedly be seen?

[Talleyrand]

How did the situation in the northeast come about and how likely is such to be an occurence? It is a rare fluke or is there a more systemic problem here that could repeatedly be seen?

https://uselectionatlas.org/FORUM/index.php?topic=211434.0

Here is a link to the thread.

It probably won't happen too frequently, but my biggest issue here (in addition to clearing ambiguity) is the point I mentioned earlier with the asymmetrical movement you're going to see in lower preferences. That doesn't reflect where votes are realistically going to flow and the like and shouldn't be used for tie-breaking purposes.

[HagridOfTheDeep]

I think I still like the old way...

Tangent: PR-STV is interesting, but I do actually think it's crappy that second preferences are sometimes meaningless. It's hard to understand for new players, too. One thing I've always thought we should consider is transitioning to a points system. For example, voters could get "20 points per ballot" and distribute them among the candidates however they wished. I don't really know how this fits into the current discussion, but it seems fitting. Only because I think PR-STV gives the impression that second preferences count. Even though they often don't. Here, you could technically give meaningful second-preference support to another candidate... assuming you were okay with taking away some support from your first preference.

[Talleyrand]

I motion for a final vote. Senators have 24 hours to object.

[Southern Senator North Carolina Yankee]

I think I still like the old way...

Tangent: PR-STV is interesting, but I do actually think it's crappy that second preferences are sometimes meaningless. It's hard to understand for new players, too. One thing I've always thought we should consider is transitioning to a points system. For example, voters could get "20 points per ballot" and distribute them among the candidates however they wished. I don't really know how this fits into the current discussion, but it seems fitting. Only because I think PR-STV gives the impression that second preferences count. Even though they often don't. Here, you could technically give meaningful second-preference support to another candidate... assuming you were okay with taking away some support from your first preference.

This is kind of like what the South used at the beginning of its legislature if I recall correctly. Though not entirely, obviously. You could vote for one candidate or all candidates and one with the most cumulative vote totals one.
                   A                     B
Voter 1        X
Voter 2                                X
Voter 3        X                      X
Voter 4        X                      X
Voter 5        x

Candidate A wins 4 to 3 over Candidate B

[Talleyrand]

Senators, a final vote is now open. The vote will last 72 hours or until a majority to pass or kill the bill is reached.


AYE

[Lief 🗽]

AYE

[TNF]

Aye

[Blair]

aye

[Senator Cris]

Abstain

[Boston Bread]

Aye

[Türkisblau]

Aye