# better way of voting

Page 2 of 3 [ 40 posts ]  Go to page Previous  1, 2, 3  Next

Roman
Veteran

Joined: 17 Mar 2005
Gender: Male
Posts: 2,301

05 Mar 2012, 4:20 pm

VIDEODROME wrote:
Yeah it did seem similar. I wasn't sure about the fraction of a vote versus declaring a score. This is fair because I sometimes wonder if people should all be able to score 1-10 or if the available points they can distribute to the candidates on their ballot should be limited somehow? I mean the way I worded it could a person just vote 10 for two of their favorites? So I guess either a split vote as you put it would work or if using a score system it might need to have the points limited. .

What I proposed is basically scores from -10 to 10. Thus we can give negative scores to candidates we want to vote out.

But actually we don't need to impose any restriction at all -- after all, our votes get normalized. So if we give 1000 points to candidate A, 10 points to candidate B and 1 point to candidate C, and -500 to candidate D then they will end up getting 1000/1511, 10/1511 and 1/1511 and -500/1511 respectively. In other words, normalization automatically limits everything to range from -1 to 1. But at the same time, if we give very high scores, we are able to obtain fractions with more and more decimal points, thus indicating our view with ''a lot of precision'' if thats what we want. But that is all finer pionts. In general yes it is basically simiar to the system you described -- especially since for most people the 1/10 precision is good enoough.

VIDEODROME wrote:
Also one big thing I would be interested in is can we eliminate the Primary Election? If we can have a system that accommodates multiple candidates let's vote for everyone in the General Election.

That is actually a good question on its own right. Even if we didn't do fraction votes; still, why not let Hillary and Obama both go to country-wide election, and then democrates will show their preference to Obama then. Doing primary is pointless and in fact it only distorts results -- perhaps some of the mcCain voters would have voted for Hillary if that option was available.

VIDEODROME
Veteran

Joined: 20 Nov 2008
Age: 43
Gender: Male
Posts: 2,766

05 Mar 2012, 4:51 pm

This also makes me think of the 3rd party folks such as Nader who was unfairly labeled the Spoiler of 2000. That by itself shows how screwed up the voting system is and how 2 entrenched minorities are in a position to kick others out of the debates.

enrico_dandolo
Veteran

Joined: 20 Nov 2011
Age: 29
Gender: Female
Posts: 866

05 Mar 2012, 4:53 pm

VIDEODROME wrote:
The system I like is sometimes called Range Voting. It could be a ballot where you give everyone a score instead of picking a single favorite.

The typical analogy for it is how you choose the best person in an Olympic contest. The judges don't just pick a single favorite they give every participant a score. The person with the highest average score wins.

I think Citizens at the ballot box should be able to act like the Olympics judges and use some kind of score system to declare the merit they see in each candidate. The candidate with the highest score should win.

That wouldn't work really either. There is no reason not to give only 10's and 0's. This means it is like allowing simply to vote for several candidates, which can be workable of course; but then, why not just ask to put choices in order?

Idem with the negative. In the end, it is still binary, or at best ternary (people I want, people i don't care about, people I don't want). It is all very nice, but it makes little difference, and it is not much less trouble than other, better options.

As for normalization, it is needlessly complicated, as it involves mathematical operations for every ballot without really improving representativeness.

VIDEODROME
Veteran

Joined: 20 Nov 2008
Age: 43
Gender: Male
Posts: 2,766

05 Mar 2012, 7:06 pm

I think the idea is still worth exploring, though I admit the finer detail of how to either score or rank each candidate has to be worked out.

I would still like to see experimental trials using a method like that to test it. I do find it interesting that it's accepted for use by Olympic judges and on the small scale with multiple candidates it seems to work fine.

Though I have read there are perplexing rules to follow when there is a Tie Breaker. Maybe people can be given a limited number of points based on the number of candidates. Maybe everyone gets 19 points, so they have enough to give one person a 10 and another person a 9, or they can rank their favorite a 10, another a 6, and another a 3.

Yeah maybe it seems convoluted but I think it's a direction worth exploring. Either that or the OP's fraction split vote.

ruveyn
Veteran

Joined: 21 Sep 2008
Age: 83
Gender: Male
Posts: 31,726
Location: New Jersey

05 Mar 2012, 7:15 pm

AstroGeek wrote:
ruveyn wrote:
Look of Arrow's Theorem. All methods of voting when there are more than two alternatives are flawed.

ruveyn

True, but that doesn't mean that you can't improve on them though.

Only up to a point. No method of voting for more than two choices can be completely fair.

ruveyn

AstroGeek
Veteran

Joined: 28 Jan 2011
Age: 26
Gender: Male
Posts: 1,582

05 Mar 2012, 8:39 pm

ruveyn wrote:
AstroGeek wrote:
ruveyn wrote:
Look of Arrow's Theorem. All methods of voting when there are more than two alternatives are flawed.

ruveyn

True, but that doesn't mean that you can't improve on them though.

Only up to a point. No method of voting for more than two choices can be completely fair.

ruveyn

Yes, but the first past the post system used on the USA (and Canada--I can't act all superior on this one) has no advantage over several other voting systems, such as mixed member proportional representation, but several disadvantages.

enrico_dandolo
Veteran

Joined: 20 Nov 2011
Age: 29
Gender: Female
Posts: 866

05 Mar 2012, 8:41 pm

I has one advantage: it is plain simple and easy. Even a child can understand it -- and, of course, notice how unrepresentative it can be.

Oldout
Veteran

Joined: 9 Feb 2012
Age: 69
Gender: Male
Posts: 1,539

06 Mar 2012, 11:59 am

Vote counting is a problem and its solution is tricky. How candidates get on the ballot is, I believe, a much more important question. I have previously suggested to Mayor Bloomberg to setup an Independent Lottery Party. Persons with a least a H.S. diploma, registered independent would pay say \$5,000 to put their name the lottery bin. (They must also commit to a minimal amount of campaign work for the winner.) These people would then get the necessary signatures to get the winner on the ballot. Ask yourself this question -- could these candidates be any worse than what we have now. At the very least they provide a clear vote against the two parties oligarchy.

ruveyn
Veteran

Joined: 21 Sep 2008
Age: 83
Gender: Male
Posts: 31,726
Location: New Jersey

06 Mar 2012, 1:52 pm

AstroGeek wrote:
Yes, but the first past the post system used on the USA (and Canada--I can't act all superior on this one) has no advantage over several other voting systems, such as mixed member proportional representation, but several disadvantages.

Then chose the unfairness that appeals to you.

ruveyn

Fnord
Veteran

Joined: 6 May 2008
Gender: Male
Posts: 41,599
Location: Stendec

06 Mar 2012, 2:03 pm

AstroGeek wrote:
ruveyn wrote:
Look of Arrow's Theorem. All methods of voting when there are more than two alternatives are flawed. ruveyn
True, but that doesn't mean that you can't improve on them though.

...

Here's some reasoning while we wait:

Majority Rule: In a two-choice election, the majority of voters determines which choice wins the election (51% to 49% ... 51% is a majority of voters). While not completely "fair", this method is more "fair" than allowing a minority to determine the outcome.

Minority Rule: In an election with more than two choices, a minority of voters can determine which choice wins the election (34% to 33% to 33% ... 34% is a minority of voters). Oligarchies and dictatorships are extreme examples of Minority Rule.

Roman
Veteran

Joined: 17 Mar 2005
Gender: Male
Posts: 2,301

06 Mar 2012, 4:12 pm

Fnord wrote:

Consider the following situation

1. There are as many democrates as republicans

2. Democrates favor candidates A and B

3. Republicans favor candidates C and D

4. There are SLIGHTLY more democrates that favor A than the ones favoring B

5. There are SLIGHTLY more republicans that favor C than the ones favoring D

6. If democrates are to be told that either C or D wins, they would OVERWHELMINGLY hope that D wins rather than C

7. If the republicans are to be told that A or B wins, they will OVERWHELMINGLY hope that B wins rather than A

Now consider two situations: the standard voting -- people can only vote for one candidate, and my version of voting when we can vote for more than one.

I. STANDARD VOTING (you can only vote for one person). Candidate A gets ''slightly more'' than half of democratic vote. Candidate B gets ''slightly less'' than half of democratic vote and none of republican vote. Thus, candidate A gets more vote than candidate B. Candidate C gets slightly more than half of republican vote and none of democratic vote; candidate D gets slightly less than half of republican vote and none of democratic vote. Thus, candidate C gets more vote than candidate D. Thus, the president will be either A or C.

2. MY VERSION OF VOTING (you can split votes). Candidate A gets ''slightly more'' than half of democratic vote and NONE of the republican split votes (republicants prefer B over A). Candidate B gets ''slightly less'' than half of democratic vote and A LOT of the republican split votes. Since A LOT of republicans were willing to give ''split votes'' to candidate B, these split votes outweigh the ''few'' democratic votes that candidate B lacked. As a result, the DEMOCRATIC candidate B was able to do better than another democrate, A,, solely because of REPUBLICAN votes. Similarly, the REPUBLICAN candidate D will be able to do better than another republican, C, solely because of DEMOCRATIC votes. As a result, the president will be either B or D; this should be contrasted with the outcome of ''standard'' case when the president is either A or C.

BOTTOM LINE: in order for democrates to ''help out'' one republican over the other, we NEED split votes -- if there is no split voting, the democrates would be more concerned with their own (democratic) candidates and they would have NO impact in the event republicans are to win. Now, if there are more than one republican, the democrates will ''help out'' the one that is closer to the middle rather than the one on extreme right. Likewise, if there is more than one democrate, the republicans will help out the one ''closer to the middle'' rather than extreme left. Thus, allowing democrates/republicans to ''help out'' the candidates of opposing party will put candidates ''closer to the middle'' at an advantage while puting both extreme right and extreme left to disadvantage. Now since ''split votes'' is what allows this ''helping out'' to occur, the ''split vote'' system will favor the ''middle ground'' candidates while more than standard system does.

Fnord wrote:
Here's some reasoning while we wait:

Majority Rule: In a two-choice election, the majority of voters determines which choice wins the election (51% to 49% ... 51% is a majority of voters). While not completely "fair", this method is more "fair" than allowing a minority to determine the outcome.

Minority Rule: In an election with more than two choices, a minority of voters can determine which choice wins the election (34% to 33% to 33% ... 34% is a minority of voters). Oligarchies and dictatorships are extreme examples of Minority Rule.

The problem with minority rule is that it allows the possiblity of an extremist being elected. My way of voting takes steps in preventing it. Suppose for example we have 3 candidantes, A, B and C. And suppose the candidate C is an extremist: 34% of population like him, while the rest of 66% hate him. In case of standard voting system, the candidate C will still win simply because 34% is greater than 33%. On the other hand, if we have my version of voting, the extremist will lose. After all, the people voting to C will also give partial votes to A and B. On the other hand, people voting for A and B will never give partial votes to C simply because they hate him so much. Thus, the candidate C will get less than 33% because a portion of 34% will be wasted on A and B without any ''compensation'' in a from of split votes from people on the side of A and B; on the other hand, A and B will get more than 34% becauise they got ALL of the votes from their supporters (after all, the supporters of A and B would NEVER give any portion of their vote to C ) and in addition some fraction of the votes from people on the side of C, as well. Thus, A and B would win, which would make most ppl ''safe'' from C.

AstroGeek
Veteran

Joined: 28 Jan 2011
Age: 26
Gender: Male
Posts: 1,582

06 Mar 2012, 4:28 pm

ruveyn wrote:
AstroGeek wrote:
Yes, but the first past the post system used on the USA (and Canada--I can't act all superior on this one) has no advantage over several other voting systems, such as mixed member proportional representation, but several disadvantages.

Then chose the unfairness that appeals to you.

ruveyn

Oh, believe me I'm trying.

enrico_dandolo
Veteran

Joined: 20 Nov 2011
Age: 29
Gender: Female
Posts: 866

06 Mar 2012, 5:55 pm

Roman wrote:
2. MY VERSION OF VOTING (you can split votes). Candidate A gets ''slightly more'' than half of democratic vote and NONE of the republican split votes (republicants prefer B over A). Candidate B gets ''slightly less'' than half of democratic vote and A LOT of the republican split votes. Since A LOT of republicans were willing to give ''split votes'' to candidate B, these split votes outweigh the ''few'' democratic votes that candidate B lacked. As a result, the DEMOCRATIC candidate B was able to do better than another democrate, A,, solely because of REPUBLICAN votes. Similarly, the REPUBLICAN candidate D will be able to do better than another republican, C, solely because of DEMOCRATIC votes. As a result, the president will be either B or D; this should be contrasted with the outcome of ''standard'' case when the president is either A or C.

No it wouldn't.

If the Republicans give part of their vote to the Democratic candidate, they also help the Democratic candidate to win over the Republicans to start with. They would prefer one Democrat over another, but they still don't want either.

It is like asking: "Would you prefer an ice cream, a banana muffin a kick in the face or a stab wound in the chest?" (Let us imagine people like pain where I live.) I may want the ice cream more than the banana muffin, and prefer the kick to the stab; but if helping the kick over the stab hinders my chance of getting either ice cream or muffin, why would I even consider it? Even more: if I don't really like muffin and love ice cream, and both have equal chance of winning, why would I even vote for the muffin if it hinders my chance of getting ice cream?

A way to make your objective more workable would be to involve Republican voters in the Democratic primaries, and vice versa -- but even then, helping the other party choose a moderate would help them win too, so it is unlikely it would work. "Split votes" are useless and disadvantageous. Unless you force people to split their vote, there is no reason for them to do it.

With half intelligent voters, your system is not really different from the standard one.

AstroGeek
Veteran

Joined: 28 Jan 2011
Age: 26
Gender: Male
Posts: 1,582

06 Mar 2012, 6:45 pm

Fnord wrote:
AstroGeek wrote:
ruveyn wrote:
Look of Arrow's Theorem. All methods of voting when there are more than two alternatives are flawed. ruveyn
True, but that doesn't mean that you can't improve on them though.

...

Here's some reasoning while we wait:

Majority Rule: In a two-choice election, the majority of voters determines which choice wins the election (51% to 49% ... 51% is a majority of voters). While not completely "fair", this method is more "fair" than allowing a minority to determine the outcome.

Minority Rule: In an election with more than two choices, a minority of voters can determine which choice wins the election (34% to 33% to 33% ... 34% is a minority of voters). Oligarchies and dictatorships are extreme examples of Minority Rule.

I do not disagree with either of those points. Which is why a winner-take-all system, like those of the USA, Canada, and Britain are inherently bad. I will use an example from Canadian politics. Say I live in riding where there is a tight three-way race between a Liberal (centre), a Conservative (right), and an NDP (left). The Liberals and the NDP each get 30% of the vote. The Conservative gets 40% and wins. This is minority rule and what we have now. However, given the choice between a Liberal or a Conservative, an NDP would almost certainly choose the Liberal. So in the end you get a party that the majority of people would not want. This could happen in the USA as well if there were more than 2 serious parties.

One way to eliminate some of the problems with this would be instant run-off voting. A voter would go into the voting booth and rank their choices from most preferred to least preferred. During the first round, the candidate who was the first choice of the fewest people would be eliminated. Those who had voted for him or her would then have their second choice counted in the next round. This would be repeated until someone wins by a majority. The problem with this is that it tends to reinforce centrist views to the exclusion of the left and the right.

And as an even better example, look at the 2011 election. The Conservatives won a majority of the seats with 40% of the vote. As the party is strictly controlled on how they vote on legislation, and because our constitution does not have a clear separation of legislative and executive powers, this gives the prime minister near absolute power. A prime minister that 60% of voters didn't want. This is another example of your minority rule.

The single best solution would be some form of proportional representation. There are some models that preserve local representation, but allow the distribution of seats in the House of Commons to reflect the distribution of votes. Although the prime minister might be elected with a minority of the vote, his or her power would be constrained because he or she would have to get approval from other parties in order to pass legislation. Alternatively, coalitions could be used so that the parties forming government would, in total, represent a majority of the vote.

I find all of this to be considerably better than limiting oneself to a two-party system. This allows more views to be heard, and for people to vote for a party that better represents what they believe in. Also, if the two main parties start to become corrupt, then there is room for another party to emerge and take over.

AstroGeek
Veteran

Joined: 28 Jan 2011
Age: 26
Gender: Male
Posts: 1,582

06 Mar 2012, 6:50 pm

enrico_dandolo wrote:
Roman wrote:
2. MY VERSION OF VOTING (you can split votes). Candidate A gets ''slightly more'' than half of democratic vote and NONE of the republican split votes (republicants prefer B over A). Candidate B gets ''slightly less'' than half of democratic vote and A LOT of the republican split votes. Since A LOT of republicans were willing to give ''split votes'' to candidate B, these split votes outweigh the ''few'' democratic votes that candidate B lacked. As a result, the DEMOCRATIC candidate B was able to do better than another democrate, A,, solely because of REPUBLICAN votes. Similarly, the REPUBLICAN candidate D will be able to do better than another republican, C, solely because of DEMOCRATIC votes. As a result, the president will be either B or D; this should be contrasted with the outcome of ''standard'' case when the president is either A or C.

No it wouldn't.

If the Republicans give part of their vote to the Democratic candidate, they also help the Democratic candidate to win over the Republicans to start with. They would prefer one Democrat over another, but they still don't want either.

It is like asking: "Would you prefer an ice cream, a banana muffin a kick in the face or a stab wound in the chest?" (Let us imagine people like pain where I live.) I may want the ice cream more than the banana muffin, and prefer the kick to the stab; but if helping the kick over the stab hinders my chance of getting either ice cream or muffin, why would I even consider it? Even more: if I don't really like muffin and love ice cream, and both have equal chance of winning, why would I even vote for the muffin if it hinders my chance of getting ice cream?

A way to make your objective more workable would be to involve Republican voters in the Democratic primaries, and vice versa -- but even then, helping the other party choose a moderate would help them win too, so it is unlikely it would work. "Split votes" are useless and disadvantageous. Unless you force people to split their vote, there is no reason for them to do it.

With half intelligent voters, your system is not really different from the standard one.

They are useless in a two party system. But in a multiparty system they might serve some purpose: I might really like Party A and give them 3/4 of my vote. But I know that only parties B and C stand a serious chance of winning. As I really hate C, I give the remaining 1/4 of my vote to B to help that party out against C. That said, I still think that instant runoff voting would work better in this situation.