Slashdot Mirror

is the Answer! Mathematical proofs of lower regret index scores and essays on why RV will save us all Here.

Its really tough to game the Hugo voting system. Here's a description:

http://www.thehugoawards.org/?page_id=4

Many US Libertarians feel that we should use this system (or perhaps, a system with a similar design) to vote in US elections:

http://rangevoting.org/rangeVapp.html

I think the general societal norm is quite different from the political norm in developed countries. Otherwise, I agree.

Nerds are deviants under (say) 1950's norms because they realize those norms are untenable in the face of advancing science and technology. I think nerds tend toward radical (libertarian/anarcho-socialist/green) ideologies because they see an irreconcilable conflict between current mainstream political ideologies and technological/social progress.

There are a lot of reasons for the political decay in most countries, ranging from cultural melting-pot instability, to the game-theoretic 2-party mess to the institutionalization of congress (essay of the same name, Polsby, 1968), to the lack of a frontier for radicals to inhabit, constructing their own societies free from external influence.

Until those problems are mitigated, political norms will remain skewed from general societal norms.

The evidence would be in the notion of what you are trying to do. That is divide something into sections that don't change the political layout of the districts after they have been changed with the intent of disfiguring the political layout.

What I mean is that once the district has been appropriated because of political advantage, any deviation from that advantage or to another would have the same appearance and effects of gerrymandering. The only difference would be the intent of the redistricting.

Imagine I have 4 groups of people, I have divided them in a way that leave one group pro democrat and the others pro republican but there is actually a 50/50 split between the people of all the groups. First, How do you re portion the groups without it looking like your getting rid of the pro republican advantage in favor of the democrats? Next, people tend to group themselves together with like minded people so how do you use math to divide the groups without favoring one side or the other when their locations might be split? Third, About any change, even if it was intended to be neutral would have some political backing behind it, it would in effect be Gerrymandering because of the motivation to equalize the field (IE give the pro democrat some power back).

It would be more about what people perceive then what really happens. A lot of times the funny shaped and long entails into other districts are because of population growth more then anything else. It is true that some are specifically done with the intention of aiding the power of one party over another but that isn't always the case.

And there are some people who are trying to use math to draw district lines. They are using what they call the Shortest splitline algorithm They even have produced some images for each state. I would like to see them make images from previous census readings to show how different the districts looked before the new census numbers. One of the problems is that when a new district needs to be made, how do you make it without changing the representations of the old districts? I would be pretty upset if every ten years, I got a new congressman who both didn't campaign in my area and I had no chance of voting for. Somehow I think using a blind formula to district things would have that effect without regard to the goal.

I think the US could benefit from a better voting system but I don't think Instant Runoff Voting as used in Australia is the way to go. IRV might sound good initially but it really doesn't do much for small parties and has some scary consequences.

There are lots of different systems but Range Voting looks pretty good to me.

That's only one layer of the discussion. There are several ways it could be implemented in a federal system to elect a president. Unfortunately, its a tedious boring subject for most people.

I think the US could benefit from a better voting system but I don't think Instant Runoff Voting as used in Australia is the way to go. IRV might sound good initially but it really doesn't do much for small parties and has some scary consequences.

There are lots of different systems but Range Voting looks pretty good to me.

That's only one layer of the discussion. There are several ways it could be implemented in a federal system to elect a president. Unfortunately, its a tedious boring subject for most people.

The Center for Range Voting has a solution to this problem here: http://rangevoting.org/GerryExec.html

They've got some other great ideas that just might work if they had enough people on board.

We need range voting. Voting in plurality voting or IRV single-winner elections merely lends credibility to a corrupt method of candidate selection, encouraging dominance by two parties which have a game-theoretic interest in remaining close to moderates on most issues. You can vote for third parties, but it's literally not worth your time to fill in the bubble or press the button, and it's certainly not worth your time to research candidates who have no chance of winning.

Nobody cares whether Libertarians or Greens get .5% or 1% of the vote (unless it's in a swing state), but if national election turnout falls to 20% or 10%, that might catalyze real change.

The rest of the problem is voter stupidity, but that's unsolvable.

Our lack of an "open multi-party system" in the U.S. is due to the horrid Plurality voting system that we (and most other countries) use for national elections. IRV single-winner -- which is foolishly used in a few (foreign) national elections and a few lower-level elections here and abroad -- is not much better. It encourages speculative, dishonest, strategic voting. Its only dubious distinction (compared to plurality voting) is that it allows voters to provide more (potentially dishonest) information about their preferences.

Range Voting is a very nice alternative. Condorcet is another possibility; while not quite as good as Range Voting, it is vastly superior to Plurality and IRV, and it can use the same ballots as IRV. Its main drawbacks are its complexity, and its few flaws that stand out when compared to the arguably ideal Range Voting system.

The way you eliminate election fraud is with a simple method from the "R" in RSA: 3ballot.

http://rangevoting.org/Rivest3B.html

I'm not that optimistic about the silver lining. Our plurality voting system ensures a 2-party-dominated system*, except for very occasional shake-ups. Aberrations like the Ventura victory (at a state level) or the Perot catastrophe (at the federal level) are due to idealistic grassroots efforts that ignore, or in rare cases can overcome, the mathematical reality of the voting system. Look at most of the "independent" reps and senators in the Federal government. They didn't start as independents. They started in a major party, gained reputation, and voters got attached to them; only then could they survive as independents.

Support Range Voting.

* see, for instance, Downs' An Economic Theory of Democracy, chapter 8.

http://rangevoting.org/IRVLetter.html

http://rangevoting.org/

http://rangevoting.org/IRVLetter.html

http://rangevoting.org/

You make an important point, though it seems that the example you give assumes that each voter must rank all of the candidates. IRV supports the option that you don't vote at all for the candidates you don't want. Also there are two kinds of monoticity:

"Monotonicity" may sound intimidating, but it is not a big deal. The term actually has several definitions. Pairwise voting methods are monotonic with respect to swapped pairs. This means that, on a ballot marked "Anderson,Reagan,Carter", if you swap Reagan and Carter so the ballot reads "Anderson,Carter,Reagan", the voting method ensures that Carter will not lose if he were already the winner, and Reagan will not win if he were a loser. IRV does not satisfy this, because this may cause Reagan to be eliminated, and the next choices of Reagan's voters could cause someone other than Carter to win. IRV is monotonic with respect to added rankings. If you add a ranking to the end of the list on your ballot, or you add a ballot with a single ranking, it will always help that candidate win, and never hurt any higher-ranked candidates. Pairwise methods do not satisfy this, as demonstrated in a previous answer. The fact that each of these voting methods satisfies one type of monotonicity and not the other is just another reflection of the tradeoff between compromising and gambling on a higher payoff that is inherent to all voting methods.

You make an important point, though it seems that the example you give assumes that each voter must rank all of the candidates. IRV supports the option that you don't vote at all for the candidates you don't want. Also there are two kinds of monoticity:

"Monotonicity" may sound intimidating, but it is not a big deal. The term actually has several definitions. Pairwise voting methods are monotonic with respect to swapped pairs. This means that, on a ballot marked "Anderson,Reagan,Carter", if you swap Reagan and Carter so the ballot reads "Anderson,Carter,Reagan", the voting method ensures that Carter will not lose if he were already the winner, and Reagan will not win if he were a loser. IRV does not satisfy this, because this may cause Reagan to be eliminated, and the next choices of Reagan's voters could cause someone other than Carter to win. IRV is monotonic with respect to added rankings. If you add a ranking to the end of the list on your ballot, or you add a ballot with a single ranking, it will always help that candidate win, and never hurt any higher-ranked candidates. Pairwise methods do not satisfy this, as demonstrated in a previous answer. The fact that each of these voting methods satisfies one type of monotonicity and not the other is just another reflection of the tradeoff between compromising and gambling on a higher payoff that is inherent to all voting methods.