As you might have noticed, we use plurality for voting (we are only allowed to choice 1 candidate for office). Plurality is an artifact of voicevoting:

A clerk would record verbal votes. For speeding things, the clerk would only allow 1 candidate for office.

Plurality suffers from Duverger's Law:

If 2 candidates are similar, they split the vote and both loose. This leads to 2 unassailable political parties. A vote for anypony else than the lesser of 2 evils is a vote for the greater of 2 evils.

¿Does it have to be this way? ¡No!:

If one removes the OverVoteRule so that one can approve multiple candidates, one removes vote splitting. Let me give an example:

Amy gets 60% of the vote. Bob gets 40% of the vote. Clara, who is very similar to Amy joins the race. The new totals are:

Amy:
30%

Bob:
40%

Clara:
30%

Let us try approval (the name of the voting system where one can approve multiple candidates):

Amy:
60%

Bob:
40%

Clara:
59%

¡Amy wins!

By removing the OverVoteRule, ¡we eliminate votesplitting and allow independents and 3rd parties to win!

5617772
For a more complete explanation of how our voting system is broken and how it can be fixed, I highly recommend this playlist:

5617779

CGPGrey understands what is wrong with Plurality, but he fell for the lies of FairVote:

FairVote pushes IRV (Instant RunOffVoting). IRV is a reform which does not solve the problem. In 2000, it would give us Gore, but can never give us Nader:

In IRV, secondary, tertiary, et cetera votes from eliminated candidates get transfered to noneliminated candidates. This leads to 1 of the 2 dominating parties to win (all votes roll downhill toward the 2 dominating parties). It also sufferes from nonmonoticity (giving more votes to a winning candidate can cause that candidate to loose) because of the kingmakereffect the votetransfers from eliminated candidates can cause. Because of the multiple rounds of counting, IRV is slow to count and increases the risk of recounthell from results of rounds being close (if one has 9 rounds of eliminations, one has 9 chances for results in a round to happen to be very close).

I suspect but cannot prove that either the Republicans or Democrats are behind FairVote.

5617806

but he fell for the lies of FairVote:

Because anyone who disagrees with you is obviously being malicious. No honest disagreement allowed, right?

IRV is a reform which does not solve the problem. In 2000, it would give us Gore, but can never give us Nader:

1. That's a rather black-and-white way to look at the issue. It would certainly be an improvement.
2. No sensible voting system would have ever given us Nader; he was a fringe candidate.

I suspect but cannot prove that either the Republicans or Democrats are behind FairVote.

Republicans and Democrats are both perfectly fine with the current voting system. Why on Earth would they back any reform. Your paranoia doesn't even make sense.

It's all about ranked choice voting

5617857
5617865

Both plurality and IRV are susceptible to clones.

> "Earth would they back any reform. Your paranoia doesn't even make sense."

They do not:

Ponies get fed up with republicats and democans. They want reform. They want 3rd parties and independents. FairVote drops in with its extensive resources and money and offers IRV. They claim that 3rd parties and independents can win under IRV. Ponies adopt IRV and only republicrats and democans win after the adoption.

¡Nothing like a reform which changes nothing, from an organization which cannot account for its resources!

5617924 You don't think voting patterns would change over time?

5617924
A major point against approval voting is that it fails to account for voter's preferences of one candidate over another. It can lead to situations where the ‘meh’ candidate that isn't anyone's first choice ends up winning. Approval voting and ranked-choice voting have different strengths and weaknesses, and therefore one is not objectively better than the other.

Edit: And yes, ranked-choice voting is better than our current system. Your all-or-nothing attitude is not helpful.

5617926

> "You don't think voting patterns would change over time?"

In order for a 3rd partier such as Johnson or Stein to win under IRV, that candidate would need a plurality of the votes and not transfer those votes to 1 of the 2 major parties. By the way, this is exactly what it would take to win under plurality. Granted, one could try ramping up the vote total for the 3rd party without making the greater of 2 evils win under IRV, but in order to win the election, when the 3rd party finally gets a plurality, one must get one's supporters to stop hedging their bets by marking either the republicrats or democans as a 2nd choice, or the votes will transfer to 1 of the 2 big parties. This is why 3rd parties do not win under IRV. It is mine opinion that IRV is nothing more than a way for the 2 major parties to prevent real reform by giving voters a false reform, which, although it prevents spoiling like Nader in 2000, does not let 3rd parties and independents win. Just remember that Australia has used IRV longer than any other country and is still a duopoly.

5618100

> "A major point against approval voting is that it fails to account for voter's preferences of one candidate over another."

I am glad you brought that up because it is possible to approve some candidates more strongly and to actively reject candidates and with different degrees of rejection. That was suppose to be the next installment, but I see now that I have to postpone it and cover something else. I started with Approval because it requires no modifications to anything, just removal of the OverVoteRule. The next step would require ballotredesigns, but would still use paperballots and can be hoofcounted, in the precincts.

Both you and Super seem to believe that the only ranked voting system around is IRV. This is not true. Hundred of ways of counting raked ballots exist for both single-winner and multiple-winners. Here is a partial list of ranked systems for single-winner:

Preferential/ranked/ordinal systems

* Borda count
* Bucklin voting
* Condorcet methods (Copeland's, Dodgson's, Kemeny-Young, Minimax, Nanson's, Ranked pairs, Schulze)
* Coombs' method
* Instant-runoff (alternative vote)
* Contingent vote
* Oklahoma primary electoral system

I shall cover ranked voting systems for single-winner in the next installment and get back to Approval with indicators for strength of approval later. Before I go, I would just like to comment on Bucklin:

Bucklin is also known as American Instant RunOffVoting (as opposed to Australian Instant RunOffVoting which is IRV). Bucklin invented it in the Progressive Era as a way to let 3rd parties and independents win. It worked, so the republicrats and democans joined forces to kill it because it was a threat to their duopoly. This is how it works:

# 0 . The voter ranks the candidates.
# 1 . One counts the votes for the lowest numerical rank not yet counted.
# 2 . If candidate(s) have a majority, then stop (the candidate with the greatest majority wins), if no candidate has a majority, go to line # 1 .

This is how it would have worked in Florida in 2000:

No candidate has a majority on 1st choices. 2nd choices push both Gore and Bush over 50%, but Gore is farther past 50%, so he wins.

It is interesting that FairVote pushes the Australian system which does not allow 3rd parties and independents to win instead of the American system, which lets 3rd parties and independents win.

5618530

Both you and Super seem to believe that the only ranked voting system around is IRV.

Uh, no. You brought up one particular method of voting and I presented another, and you claimed the one you presented was better than the one I presented. That is not me ignoring other voting systems, that's just me staying on topic.

This is how it would have worked in Florida in 2000:
No candidate has a majority on 1st choices. 2nd choices push both Gore and Bush over 5r0%, but Gore is farther past 50%, so he wins.

I like how you make a big deal about third party votes, and then present an example that doesn't go to a third party, and in fact gives the same results as IRV would. But no, IRV is evil, because… uh… let me think a minute…

5618562

> "I like how you make a big deal about third party votes, and then present an example that doesn't go to a third party, and in fact gives the same results as IRV would. But no, IRV is evil, because… uh… let me think a minute…"

I gave an example most know:

In 2000, Nader took votes away from Gore. To a lesser extent, Buchanan, took votes away from Bush.

To give to you what you want, I must create an hypothetical example because voters know that any vote not for the lesser of 2 evils is a vote for the greater of 2 evils. Since this scenario is counterfactual, take it with a grain of salt:

For the general election of the USA in 2016, the USA would use Approval. Doctrix Jill Stein decides that she does not have to pander to antivaccers and would support vaccines. Since voters can approve multiple candidates without negative candidates, Stein and Johnson would receive so many votes that 1 of them would win and the other would come in 2nd while 1 of Trump and Clinton would come in 3rd and the other would come in 4th.

That is your hypothetical example.

5620115

That is your hypothetical example.

I would argue that IRV would give a similar result. You have yet to give any convincing reason why your proposal is better.

5620120

With IRV, votetransfers would let either Clinton or Trump win.

5620138

With IRV, votetransfers would let either Clinton or Trump win.

Unless, of course, they aren't the top two choices of the vast majority of voters… which they aren't. Fear of the spoiler effect is the main driver of the big candidates in this election, and IRV removes that fear.

5620151

I do not see why you do not like just eliminating the OverVoteRule. We can use current ballots and equipment and it is hoofcountable. IRV is worse and many municipalities would have to redo their ballots get more equipment for a system inferior to Approval.

5620163

I do not see why you do not like just eliminating the OverVoteRule.

I never said I didn't like it. What I did was dispute this:

IRV is worse

You have yet to demonstrate this claim. When pressed on it, you dip into conspiracy theories about who you think might be pushing it, make unrealistic predictions about how people would vote under it, and the best of all: Refuse to acknowledge what I've said, as in your avoidance of my last comment.
Skepticism is not just a stance, it is not simply being against the mainstream. It is a process, and the previous paragraph is a taste of what it looks like.

5620404

> " > 'IRV is worse'"

> "You have yet to demonstrate this claim.

IRV is worse than Approval:

* It works with existing ballots.
* It works with existing equipment.
* It requires less voter-instruction/training (voters have to understand that they are not allowed to corank or skip ranks or they spoil their ballots (personally the NoCoRankRule and the NoSkipRankRules seem as stupid to me as the OverVoteRule)).
* It is easier to count (all of those eliminations, transfers, and recounting are hard on pollworkers).
* It is slow to to count (all of those eliminations, transfers, and recounting slow down the count).
* IRV is far more susceptible to recounthell. (One generally does a rapid hoofcount of several ballots per minute. One can get results within an hour of closing the polls. If Alice gets 45% of the vote to Bob's 55%, one can be pretty sure that Bob won. ¿What if they differ by less than 1%? These rapid counts have about a 1% error. One now must do a slow hoofrecount, where one inspects each ballot for more than 1 minute. One will not not get the results in less than a week. Each round of eliminations is another chance for the eliminated and retained candidates to be with 1%.)
* IRV is not precinctsummable (one must either ship the ballots to a counting center —— ¡a perfect opportunity to switch the ballotboxes! —— or have the pricincts in contact with a central tallier, where the precincts report their totals for each round and learn which candidate to eliminate for the next round).
* IRV is nonmonotonic (because the transfers from eliminated candidates can act as KingMaker, a wing candidate gaining more support, could loose because of changes in the order of eliminations. This generally does not happen in established duopolies adopting IRV (all of the Greens would have the Democrats as their 2nd choice and all of the Libertarians would have the Republicans as their fallback, so republicrats and democans would still win) but countries going straight from nothing to IRV have results which look just like Sortition).

In cases of an irrelevant alternative, IRV is better than plurality. I shall demonstrate:

Alice:
55%

Bob:
45.

Cathy is a clone of Alice:

Alice:
40%

Bob:
45%

Cathy:
15%

On the whole though, IRV does not seem better than Plurality and Approval kicks its arse.

5620727
You only did half of the work, though. To show that approval is better than IRV, you have to list all of the unique faults of approval that IRV doesn't have, and show why they aren't as bad. It isn't enough to disparage one side; the claim in question requires a comparison.

5620817

You already found the fault:

Approval cannot distinguish strength of support or rejection. I wrote that I shall show how one can indicate degree of support or rejection in Approval. Unfortunately, the misunderstandings of ranked systems for single-winner means that rather than write about it this weekend in the next installment, it will be in a fortnight. Since you chop at the bit —  ¡everypony here is a pony! —— I shall write the answer now:

¡ScoreVoting!

0thly, I mustmake this clear:

All scales are really from 0 to 1 or from -1 to +1. We use different ranges for avoiding fractions and radices (in decimal (Base10), decimals).

one rates the candidates on a range from -9 to +9 (all allowed values have single-digit length). One sums the votes. This is how an optical-scan ballot (pony/machine-readable) would look:

Alice: (+) (-) (1) (2) (3) (4) (5) (6) (7) (9)

Notice the lack of 0; this forces voters to come down 1 way or another. Voters can skip politicians, but they skipped politicians automatically get -9. These are the VoterInstructions. We only allow single-digit vales, which reduces errors.

How To Rate:

* Rate the best candidate in the race who is best as +9.
* Rate the worst candidate in the race as -9
* Rate all other candidates relative to the best and worst candidates

How to fill in the ballot:

* Pencil in whether the candidate is positive (+) or negative (-).
* Pencil in one numeral.
* The result is a score, e.g. such as -7 (negative seven) for bad candidate or +5 (positive five) for a good candidate.
* One may abstain from rating a candidate.

Rules for preventing spoilage:

* One must mark either (+) or (-) but not both nor neither.
* One must pencil in only one numeral.
* At least 1 candidate in each race must have a score of +9 (positive nine).
* At least 1 candidate in each race must have a score of -9 (negative 9).

The pollworkers simply sum the votes. The candidate with the highest votetotal wins. Here is an example with 2 voters and 1 candidate:

Alice gives Bob -7 votes. Cathy gives Bob +2 votes:

-7
+
+2
=
-5

Alice has -5 votes.

ScoreVoting is approval voting, but with an expanded range of allowed values:

Approval:
0, 1

ScoreVoting:
-9, -8, -7, -6, -5, -4, -3, -2, -1, +1, +2, +3, +4, +5, +6, +7, +8, +9

5621105

You already found the fault:

The fault? Singular? Really?

Notice the lack of 0; this forces voters to come down 1 way or another.

So… that really stupid thing psychologists like to do, then. That doesn't sound like a positive, to me.

I see a massive, potentially fatal, fault in score voting. There are people who have trouble understanding how to use the simple voting machines currently in place. How do you expect to get people to reliably follow a much more complicated system?

5621551

Sorry for the late response, but my country is doomed.

> " > 'Notice the lack of 0; this forces voters to come down 1 way or another.

> "So… that really stupid thing psychologists like to do, then. That doesn't sound like a positive, to me."

¿What purpose would letting ponies give 0 votes do? It also causes problems in interpreting -0 and +0. Maybe, you would like to map thing thus:

-09, -08, -07, -06, -05, -04, -03, -02, -01, -00, 00, +00, +01, +02, +03, +04, +05, +06, +07, +08, +09
-10, -09, -08, -07, -06, -05, -04, -03, -02, -01, 00, +01, +02, +03, +04, +05, +06, +07, +08, +09, +10

The 1st line would have only 1 numeral, but I write the line with leading 0s so that its members would line up with their mapping values. Not having 0s forces voters to comedown 1 way or another and simplifies figuring out voter intent (¿is -0 the same or different than 0 or +0?).

> "I see a massive, potentially fatal, fault in score voting. There are people who have trouble understanding how to use the simple voting machines currently in place. How do you expect to get people to reliably follow a much more complicated system?

The system uses paperballots. I the voter does not abstain, then the voter fills in 2 bubble:

* Either (-) or (+).
* A Numeral.

The paperballots are machine/human-readable.

5622630

¿What purpose would letting ponies give 0 votes do?

Not forcing them to choose a positive or negative rating for a candidate that they don't have strong feelings about?

Not having 0s forces voters to comedown 1 way or another

Which is exactly the problem. It forces people to choose something they might not want to.

The system uses paperballots. I the voter does not abstain, then the voter fills in 2 bubble:
* Either (-) or (+).
* A Numeral.

You specified a rule about having to rank one candidate +9 and one candidate -9. Vast numbers of people aren't going to get that.

5622649

> "You specified a rule about having to rank one candidate +9 and one candidate -9. Vast numbers of people aren't going to get that."

That rule exists because it makes sense:

If a moron would give a top score of only +4 and a bottom score of -3, that voter would be screwed. ¿Would you not give the best candidate in the race + 9? and ¿the worst candidate in the race -9? ¿Do you realize that if you vote moronically with a constricted range and the opposition votes rationally with the full range, you almost guarantee your defeat? One should give the best candidate in the race +9 and the worse candidate in the race -9 and then rate the other candidates in the race relative to these 2 extremes in the race.

5622724

That rule exists because it makes sense:

I didn't take a position on whether or not it makes sense. I pointed out that a large number of people wouldn't do it, because people, as a rule, are idiots.

5622837

> "I pointed out that a large number of people wouldn't do it, because people, as a rule, are idiots."

That is why we have BallotValidators in the precicts:

The Ballots are Pony/Machine-Readable. A BallotValidator is an optical scanner voters can use for finding BallotErrors before casting the Ballot.

5625568
Sounds expensive and time-consuming.

5625884

A ballot validator costs only about 100.00 U$D, processes a ballot in less than 1 second, and informs the voter about skipped races and mistakes which could spoil the ballot. A voter may want to skip a race, but surely the voters, after going through all of the trouble to vote, want to know about accidentally spoiled ballots, so that they can request a new ballot and fill it in correctly.