A Final Report on the Election

[Jon Seattle]

You have made the point I have been trying to make exactly. I think that we should point out the ethics of voting without changing the law. Thus, ethical action should be voluntary -- not forced by legal constraints. It should be enforced through electoral action -- voting against unethical parties (any party refusing to pledge not to abuse the system).

Yes, I have come much closer to this position. There are still a couple of things that ought to be cleaned up. First, lets use zero-based instead of 1-based Borda counts. Second, lets fix the system so that elimination works more or less as intended, that is elimination should not get stronger for each faction the more you eliminate.

Back to my expressiveness example. Instead of using the lowest count for each eliminated faction, you can reverse the direction. For ranked factions use counts:

3, 2, 1, 0

And for eliminated factions use counts:

_, 1, 0, -1 (if you want more severe use _, -2, -1, 0; or _, -3, -2, -1)

So the scores are:

3, 2, 1, 0 (none eliminated)
3, 2, 1, -1 (one eliminated, subtract one point last faction)
3, 2, 0, -1 (two eliminated, subtract one point from each of the two factions)
3, 1, 0, -1 (three eliminated, subtract one point from all three)

Now the counts for the eliminated factions would be applied evenly across all eliminated factions. This means that when you eliminate faction A, it has the same effect for faction A as if you eliminate faction A and B. In other words it allows the expressiveness of elimination, without the trick that makes your vote stronger by eliminating all you can. Its a small fix, but it makes elimination work in a way that is much closer to our intuition about it.

[Cindy Ecksol]

Second, lets fix the system so that elimination works more or less as intended, that is elimination should not get stronger for each faction the more you eliminate.

Back to my expressiveness example. Instead of using the lowest count for each eliminated faction, you can reverse the direction. For ranked factions use counts:

3, 2, 1, 0

And for eliminated factions use counts:

_, 1, 0, -1 (if you want more severe use _, -2, -1, 0; or _, -3, -2, -1)

So the scores are:

3, 2, 1, 0 (none eliminated)
3, 2, 1, -1 (one eliminated, subtract one point last faction)
3, 2, 0, -1 (two eliminated, subtract one point from each of the two factions)
3, 1, 0, -1 (three eliminated, subtract one point from all three)

Now the counts for the eliminated factions would be applied evenly across all eliminated factions. This means that when you eliminate faction A, it has the same effect for faction A as if you eliminate faction A and B. In other words it allows the expressiveness of elimination, without the trick that makes your vote stronger by eliminating all you can. Its a small fix, but it makes elimination work in a way that is much closer to our intuition about it.

Jon, you may not be comfortable with the idea of giving zero points to any faction that is eliminated by a voter, but _I_ am very uncomfortable with the idea of allowing one voter to subtract points from a faction that were given to it by another voter. In theory you could end up with factions that score "less than zero" votes which is pure nonsense in real world terms even though it makes sense mathematically. Those who are comfortable with the mathematical implications may not boggle about this, but some of us will have a difficult time with the concept. In addition, your proposal still coerces a voter who feels very strongly that a faction deserves no support at all into adding points to that factions total against he/her will. I will always advocate keeping things simple enough so that everyone can understand how they work over a system that takes significant brain-power (or a computer) to evaluate, and against coercing support for any faction or individual whom the voter does not choose to support.

Rather than monkeying with the counting system (which many of us think is not quite as broken as you perceive) , why not resolve the REAL problem and do something that will give every voter except the most partisan an incentive to vote for more than one faction? I've already suggested one such incentive: allow voters to rank the individual candidates within each faction they rank. There are several positive aspects to this suggestion. First, it addresses (although it does not completely resolve) the issue of voters who wish to rank a faction other than the one they "belong" to first. Second, it gives those who have opinions on which individuals within each faction they support can do the best job in the RA a chance to express those opinions. And third, it gives the community a chance to balance out the tendencies of factions to become dependent upon a single individual who may then dominate faction operations to the detriment of the CDS community at large. The only negative factor is that it allows "non-faction-members" to have a say in who represents each faction. But since we don't have a clear way of determining exactly who belongs to which faction (and many may choose not to be active in any particular faction) I don't see this as a significant problem.

If my proposal doesn't appeal to you, I'm sure that if we put the collective brains of the community together we can surely come up with other incentives that will resolve the problem you perceive without coercion of any sort.

Cindy

[Jon Seattle]

allow voters to rank the individual candidates within each faction they rank. There are several positive aspects to this suggestion. First, it addresses (although it does not completely resolve) the issue of voters who wish to rank a faction other than the one they "belong" to first. Second, it gives those who have opinions on which individuals within each faction they support can do the best job in the RA a chance to express those opinions. And third, it gives the community a chance to balance out the tendencies of factions to become dependent upon a single individual who may then dominate faction operations to the detriment of the CDS community at large. The only negative factor is that it allows "non-faction-members" to have a say in who represents each faction. But since we don't have a clear way of determining exactly who belongs to which faction (and many may choose not to be active in any particular faction) I don't see this as a significant problem.

In effect though, this allows voters at large to reach in and determine the policies and priorities of another faction. So for example, NuCARE could decide to rank SP last and then re-arrange it's internal people and policies. At that point why bother with factions?

As Beathan has point out elsewhere, there is an advantage to retaining a system that puts parties and platfoms first, and discourages popularity contests.

[Jon Seattle]

Those who are comfortable with the mathematical implications may not boggle about this, but some of us will have a difficult time with the concept.

Well, I ask you Cindy, if I eliminate a faction under the current system, each time I do, how many points are taken away from that faction? Your only answer is some complicated formula that depends on how many I cross off.

My proposal, on the other hand, is very easy to understand. Every time you cross off a faction you take away a fixed number of points. It always works the same way, wether you eliminate one, two, or three.

You have some problem with the (very improbable) idea that a faction may end up with a negative total, but:

1. People are most concerned with figuring out how their own vote changes the scores of the factions. In fact we should display this to them when they vote so they know for sure. Most people are not playing this as a game where they need to figure out how other people voted before they can decide on their own choice.

2. It is a lot easier to understand the totals if you have a clear statement: Faction A got so many Borda counts through 1st place, 2nd place, .. minus so many points though elimination. It makes the whole system more comprehensible, because you can immediately tell how many people eliminated the faction.

3. Its the seat allocation method is what matters in the outcome of the election. And guess what? The allocation method we use could not care less about positive and negative.

[Beathan]

Jon, you may not be comfortable with the idea of giving zero points to any faction that is eliminated by a voter, but _I_ am very uncomfortable with the idea of allowing one voter to subtract points from a faction that were given to it by another voter. In theory you could end up with factions that score "less than zero" votes which is pure nonsense in real world terms even though it makes sense mathematically. Those who are comfortable with the mathematical implications may not boggle about this, but some of us will have a difficult time with the concept.

Cindy -- Back at my college, this is exactly how we did elections for distribution of student body funds. I found that the system worked well -- and was widely understood (even outside the math department) in short order. I agree that there is the possibility that a faction could end up with a negative number of votes. However, far from being a drawback, I consider this a strength of a system that allows us not merely to measure support but also to measure opposition. I further believe that such a system will have the side benefit of reducing partisan extremism and preventing demagoguery by making it risky to play up controversial positions knowing that it will increase support in some quarters. Because such actions will reduce support elsewhere -- and both support and opposition would matter -- the result would be and increased moderation and civility in CDS politics.

In addition, your proposal still coerces a voter who feels very strongly that a faction deserves no support at all into adding points to that factions total against he/her will. I will always advocate keeping things simple enough so that everyone can understand how they work over a system that takes significant brain-power (or a computer) to evaluate, and against coercing support for any faction or individual whom the voter does not choose to support.

No, it doesn't. As I explained above, the system Jon describes is more sensitive to opposition than the current system. What it is less sensitive to is partisan manipulation -- but gaming a system for the benefit of the party a voter most supports is not the same thing as expressing opposition to parties the voter opposes.

Rather than monkeying with the counting system (which many of us think is not quite as broken as you perceive) , why not resolve the REAL problem and do something that will give every voter except the most partisan an incentive to vote for more than one faction? I've already suggested one such incentive: allow voters to rank the individual candidates within each faction they rank. There are several positive aspects to this suggestion. First, it addresses (although it does not completely resolve) the issue of voters who wish to rank a faction other than the one they "belong" to first. Second, it gives those who have opinions on which individuals within each faction they support can do the best job in the RA a chance to express those opinions. And third, it gives the community a chance to balance out the tendencies of factions to become dependent upon a single individual who may then dominate faction operations to the detriment of the CDS community at large. The only negative factor is that it allows "non-faction-members" to have a say in who represents each faction. But since we don't have a clear way of determining exactly who belongs to which faction (and many may choose not to be active in any particular faction) I don't see this as a significant problem.

I have several problems with this. First, as I have written elsewhere, such a move would tend to personalize politics in the CDS -- which is something we have tried to avoid. If we thought the partisan squabbles of the past, which frequently involved personal hostility and personal attacks, were ugly -- why should we ratchet up the ugliness by personalizing the campaign? Second, this proposal would allow members of one party (a larger party) to control the representation of smaller parties. For instance, if there had been three SP candidates in the last election (as there had been in previous elections), the SP representatives would have been chosen by the CSDF, because the CSDF was so much larger than the SP. A system that allows one party to dictate another party's leadership and representation is just wrong.

Beathan

[Cindy Ecksol]

Those who are comfortable with the mathematical implications may not boggle about this, but some of us will have a difficult time with the concept.

Well, I ask you Cindy, if I eliminate a faction under the current system, each time I do, how many points are taken away from that faction? Your only answer is some complicated formula that depends on how many I cross off.

My proposal, on the other hand, is very easy to understand. Every time you cross off a faction you take away a fixed number of points. It always works the same way, wether you eliminate one, two, or three.

Actually, Jon, under the existing system, it's very easy to see what happens: if I vote for a faction I give them 3, 2, or 1 points and if I don't vote for them I give them zero points. No points at all are "taken away." All your rhetoric about the varying impact etc. etc. is true as far as it goes, but it's simply not relevant to most people. They want to vote FOR a faction they want, not AGAINST anyone.

You have some problem with the (very improbable) idea that a faction may end up with a negative total, but:

1. People are most concerned with figuring out how their own vote changes the scores of the factions. In fact we should display this to them when they vote so they know for sure. Most people are not playing this as a game where they need to figure out how other people voted before they can decide on their own choice.

2. It is a lot easier to understand the totals if you have a clear statement: Faction A got so many Borda counts through 1st place, 2nd place, .. minus so many points though elimination. It makes the whole system more comprehensible, because you can immediately tell how many people eliminated the faction.

3. Its the seat allocation method is what matters in the outcome of the election. And guess what? The allocation method we use could not care less about positive and negative.

My point, Jon, is that a system which requires a computer in the voting booth to display "possible impacts" is a system that is too complex for the real world. You are correct: the allocation system doesn't care about "positive" and "negative" and mathematically-speaking what you say makes sense. But humans are not mathematical constructs: they work on intuition and emotion and that influences their voting logic as well as their interpretation of results. And by and large they don't understand math that you may find childishly simple. The fact is that people want to see the totals, and if you think the current confusion over "who won" the recent election and how the seats were allocated has been something to see, you haven't seen ANYTHING until you see the confusion when vote totals include negative numbers! I'm sure it will be exciting for you to explain to all of us, but personally I'd still be more comfortable with a system that allows me to reward factions I like by giving them the "vote points" I have available to distribute. I have no desire to punish factions I don't care for by taking away points that someone else has awarded to them. The current system is nice that way: with four factions in the running I have six points to distribute and it's very easy to decide how to do that without supporting a faction that I don't approve of. I think I'm not alone in this view of voting as a process of giving votes (or points) rather than taking away. I just don't like the negative feeling of the system you propose nor its complexity.

A little tip for those who are interested in implementing complex systems: you might want to consider participation in the Metaverse Republic.

Cindy

[Cindy Ecksol]

Cindy -- Back at my college, this is exactly how we did elections for distribution of student body funds. I found that the system worked well -- and was widely understood (even outside the math department) in short order. I agree that there is the possibility that a faction could end up with a negative number of votes. However, far from being a drawback, I consider this a strength of a system that allows us not merely to measure support but also to measure opposition. I further believe that such a system will have the side benefit of reducing partisan extremism and preventing demagoguery by making it risky to play up controversial positions knowing that it will increase support in some quarters. Because such actions will reduce support elsewhere -- and both support and opposition would matter -- the result would be and increased moderation and civility in CDS politics.

Well, I think we're just going to have to disagree on that perception, Beathan. I think a system that actively allows people to award "negative points" is one that encourages thoughts about "voting against" rather than "voting for." To my mind that makes it a system that encourages contention and divisiveness, not one that encourages moderation and cooperation.

Rather than monkeying with the counting system (which many of us think is not quite as broken as you perceive) , why not resolve the REAL problem and do something that will give every voter except the most partisan an incentive to vote for more than one faction? I've already suggested one such incentive: allow voters to rank the individual candidates within each faction they rank. There are several positive aspects to this suggestion. First, it addresses (although it does not completely resolve) the issue of voters who wish to rank a faction other than the one they "belong" to first. Second, it gives those who have opinions on which individuals within each faction they support can do the best job in the RA a chance to express those opinions. And third, it gives the community a chance to balance out the tendencies of factions to become dependent upon a single individual who may then dominate faction operations to the detriment of the CDS community at large. The only negative factor is that it allows "non-faction-members" to have a say in who represents each faction. But since we don't have a clear way of determining exactly who belongs to which faction (and many may choose not to be active in any particular faction) I don't see this as a significant problem.

I have several problems with this. First, as I have written elsewhere, such a move would tend to personalize politics in the CDS -- which is something we have tried to avoid. If we thought the partisan squabbles of the past, which frequently involved personal hostility and personal attacks, were ugly -- why should we ratchet up the ugliness by personalizing the campaign? Second, this proposal would allow members of one party (a larger party) to control the representation of smaller parties. For instance, if there had been three SP candidates in the last election (as there had been in previous elections), the SP representatives would have been chosen by the CSDF, because the CSDF was so much larger than the SP. A system that allows one party to dictate another party's leadership and representation is just wrong.

Well, again, I think we're just going to have to disagree. I'm not the only one who sees the tracks in CDS politics of the negative impact of individuals who have become entrenched in their faction leadership positions to the eventual detriment of the community at large. The resulting explosions are in large part an artifact of a pure faction-oriented system that permits and even encourages such entrenchment. I strongly believe that no "pure" system, either faction-based or individual-based is workable over the long term. Factions are necessary for building cooperation, but to assume that individuals are completely fungible denies the reality that a strong faction leader is not necessarily the most productive member of the RA. There SHOULD be times when one wants to vote for a faction, but believes that particular individuals who are members of the faction should not sit on the RA. There should also be times when one believes that an individual would be a great RA member, but one cannot support that faction. There's nothing at all in our voting scheme that addresses that issue, but that's what one-on-one discussions about principles are for: one can always encourage an individual to re-align his or her faction affiliation in the time between elections

Cindy

[Jon Seattle]

My point, Jon, is that a system which requires a computer in the voting booth to display "possible impacts" is a system that is too complex for the real world.

We have a preference system, and not a system where people decide to spend so many points on this and so many points on that. There is nothing complicated about taking away 1 point (or n points) from a faction. And even more important, under my proposal, it is not as easy to confuse people about the consequences of not ranking. I think it is clear that you would prefer to scrap the preference voting for one that involves voting for your favorite faction only, but in fact we have had times (the DPU in this case) where a faction has become the leader mainly because it was a second place favorite. I, and many others, consider that a benefit, not a liability.

[Cindy Ecksol]

My point, Jon, is that a system which requires a computer in the voting booth to display "possible impacts" is a system that is too complex for the real world.

We have a preference system, and not a system where people decide to spend so many points on this and so many points on that. There is nothing complicated about taking away 1 point (or n points) from a faction. And even more important, under my proposal, it is not as easy to confuse people about the consequences of not ranking. I think it is clear that you would prefer to scrap the preference voting for one that involves voting for your favorite faction only, but in fact we have had times (the DPU in this case) where a faction has become the leader mainly because it was a second place favorite. I, and many others, consider that a benefit, not a liability.

We've always been in agreement about the benefits of a preference system, Jon. "Winner take all" systems tend to produce results that foster a "mandate mentality" where the winning faction feels that it has the right to force its ideas on everyone else. Where we don't agree is in the philosophy of the preference system. You seem to prefer a philosophy that produces a system that assumes that voters want to express "negative preferences" and take away points or "punish" factions that they dislike. I'm more interested in a system that assumes that voters want to reward factions that they prefer. That's how I vote, and I expect that most others do the same.

As for complexity, well, I guess I'll just look forward to you introducing a bill to change the system as you propose, and if it passes I'm going to enjoy sitting back and watching you explain the results of casting all those "negative point" votes to the rest of the CDS after the election. You'll get the system you're after, but it'll be via coercion rather than persuasion and it WILL cause conflicts and confusion.

And as for me, hearing all of this discussion and talking offline with some about possible changes to the system has convinced me that the system really isn't broken enough that it needs a fix right now. What I AM concerned about is the "no peeking" bill (or the alternative constitutional amendment that you proposed). I think it's critical that we enact something of this sort before the next election, and I hope that the RA agrees and quickly develops a consensus around a solid bill. I also think that Beathan's proposal for RA procedure changes is very important. In particular if we want to encourage participation by all, it would be a good idea to improve the ability of citizens and RA members to get their proposals discussed rather than relying on the LRA to completely control priorities. And of course adding a "good and welfare" item to the standard agenda will ensure that other items are discussed as well even if they are not yet ready to be put in the form of a bill. That's going to take some discipline on the part of RA members to keep the discussion focused, but learning that discipline is just part of the process of running a democracy. The good news is that even the relatively disorganized discussion at last week's RA session sparked a discussion about adding a new sim on the forum, and hopefully that will result in a bill before too long.

So after all the sturm und drang, seems like we're moving along in a positive direction and I look forward to that continuing.

Cindy

[Sonja Strom]

Thank you to everyone who has put so much thought into the details of our vote-counting system, especially Jon and Cindy!
Your efforts here and elsewhere will help the CDS a great deal to improve what has been created up to now.

We still do have a few months until the next election, and I believe by then some bills will be passed to incorporate at least some of the changes you suggest, if not most of them.

Naturally the details have yet to be worked out. Yet, now we know what most of them are, with heartfelt thanks again.