One thing not mentioned in this video is that plurality voting also "incentifizes lying about your preferences" not just multiple rounds. This happens (kind of) in the US, where people might want to vote for a third party (ex. green party, libertarian party, etc.) might end up voting for the democratic/republican candidate because they think that the third party they support is unlikely to win.
And it gets worse - not only does supporting a third party not help them win, but it also draws votes away from the major party you otherwise would have supported. Too often, this leads to your worst option getting elected. I like to describe Plurality as mathematically driving us to a two-party system, and this is why.
@@robbe1534 Oh that's really cool and interesting, I didn't know that existed. However, wouldn't that (over time) incentivize voters to assign all 100 points to their candidate of choice since, if they don't, someone who does assign all points to their candidate of choice will inevitably overshadow their first choice
This depends on how you normalize the scores. Personally I prefer a simple free scoring system. Its more intuitive and still allows better expression than any of the ranked methods shown on the video.
@@jacobandrews2663 DV eliminates the worst candidate by redistributing the points, so if you give all 100 points to a candidate who is eliminated among the first, the vote becomes irrelevant. To avoid this it's better to distribute the points. Also, you consider that using range [0,10] (then converted to 100 point votes), it's difficult to accumulate points.
@@pepemotion DV can use votes with range [0,10] (then converted to a 100 point vote), appearing from the voters' point of view, as a free scoring system.
Whichever way will be fine as long as it is not China's way China's way = the non-elected govt makes all the decisions forever for the country with the claim that these are the people's unanimous decisions.
All too often we laud "popular vote" as a majority system when in reality it's most often a plurality in most countries and this video very well illustrates how that system can be flawed and other systems in comparison. Kudos!
Yes, USA for example, has most of the population in East and West coasts. So if they had a popular vote system, everything would be unfair to the farmers in the vast land in the middle. Current system they have is worse, but popular vote doesn't solve everything.
@@phs125 However, there are far less farmers in the interior of the country than there are people who live on the coasts. Does it really make sense to divide people up by where they live? No. The votes of the people on the coast should count just as much as anywhere else, and the election should be about the policies the candidates espouse, because not everyone who lives on the coast is going to vote in one way and not everyone who lives in the heartland will vote the same way.
@@Warrka4 There's a similar one. One side picks odd, the other even, then like RPS both show their numbers at the same time, the total sum of the numbers will decide if even wins or odd.
JonatasAdoM oh yea odd or eve. That's basically like cricket though so it doesn't have the same system as rock paper scissor cause it in voles more of statistics than this.
@@moustafakhattab8142 No, Russia is an oligarchy led at the moment by Putin. The basilar freedoms are limited, (not eliminated) and it's impossible to candidate aganist Putin (so he can win everytime)
*_“Many forms of Government have been tried, and will be tried in this world of sin and woe. No one pretends that democracy is perfect or all-wise..._* *_Democracy is the worst form of Government except for all the others.”_* - Winston Churchill, 1947
@@samuelforesta It happened from time to time, usually in areas like china where the king/emperor grown to powerful for external treats and forgot where his power came from. At one point the people have enough and replace the emeror with one of they own. This new emperor family will behave for a few generations until they forget the lesson as well and the circle continued. In europe uprisings were rare because most of the time the ruling familys had to defend themself from competition and needed the support of the people.
@@stonferen584 No, I mean a parliamentary system elected under MMP, which allows for broader representation than STV. Look at CGP Grey's videos on STV and MMP to see what I mean.
I live in Switzerland and most people love how the president is chosen. We vote for the party we want and then the party’s with the most votes can get on to seven on the seven seats available. No one has more than 2). Then all 7 elected have the same power and collaborate. They are each president for one year but have no additional power.
I can imagine US parties splitting in the same way monopolies "don't exist" while all being owned by mega corps. That is, Dems or Reps just form subsidiary parties that are legally distinct but the same rotation of people with the same financial interests. Our political parties are very good at subverting systems meant to benefit everyone, though.
Seems pretty fair. They also apply direct democracy methods. Switzerland has a system closer to being called "democracy" than anyone else. Their methods indeed have similarities with original democracy in ancient Greece.
The Swiss system is a close to a real democracy as it gets but that works better with a small, affluent population. But it also means that sometimes it takes too long a time to reach decisions and of course its success depends on voters being well informed and interested in politics.
@@TheFreakyable the problem with the people deciding is that it is easy that many do missinformed decisions but on the other hand you cant just deny them the vote
I’m told that if you think you are likely to vomit soon, drinking lots of water (but not like, extreme amounts, just large amounts) can make it much less unpleasant when it happens. Just some advice my dad gave me.
@@drdca8263 :-) Sounds as if a father (who was drunk in his younger times at every party) gave his now teenage child good advices how to deal with alcohol.
animal friend I can see how it would seem like that, but my understanding is that my dad has always disliked the taste of alcohol enough to prevent being drunk more than a handful of times (a single time ?)
I think Borda Count, although not perfect, would probably be the best option. Convert the rankings into points, so a first choice gets 3 points, a second choice 2 and a third choice one, and the winner is the one with the highest overall score. It therefore elects the overall most popular/least unpopular option. In this example the winner would be North, which on the map does look like the best compromise. Admittedly two of the methods demonstrated also make North the winner in this example, but the problems with both are discussed.
@@jamesabestos2800 The main problem with Condorcet, as mentioned in the video, is that it doesn't always produce a winner. Aside from that, I agree, it's a good system.
@@stirlingblackwood STAR Voting does seem like a reasonable option as long as it successfully overcomes this problem in Range Voting where people vote tactically by only using the maximum and minimum scores. I'd need to see some real life figures to be able to assess that.
My favorite voting system is Black's System, which is kind of like a combination of Borda Count and Condorcet. It puts each candidate head to head, and if there's a candidate who's able to beat every other one in a one on one match (and is thus a Condorcet winner), that candidate wins automatically. Otherwise, it plays by the same rules as the Borda Count system.
I would like to mention though that there are several solution to that paradoxical a>b>c>a in the Condorcet voting, my favourite of which is the Schulze method. It mathematically finds the candidate that would make people the most happy, even in a case of a circle as mentioned. Its downfall is that it's more prone to draws than most other voting methods, but these are still quite rare.
Condorcet is the best one listed. The fact that there are sometimes circular ties is not a flaw of the voting system. Any election can result in ties. Condorcet just makes those ties apparent instead of choosing an unrepresentative candidate in those cases. That's a good thing.
STAR (score then automatic runoff) voting is essentially score voting with a Condorcet-like runoff phase. It does a better job minimising tactical voting incentives.
An active refusal to vote with a nuanced understanding of the position is a refusal to give your consent to the government. DO not blame the victim. If you look at Martin Luther King Jr, he would not have supported voting as a means to make your voice heard. He supported civil disobedience to accomplish that end. Particularly in his later years. Organized and sustained social movements accomplish things voting has always managed to suppress.
I'm shocked to see no discussion here of range voting, in which voters give each option a score, say from 0 to 10. Key advantage: the theorem mentioned (Arrow's impossibility theorem), which proves that "all voting systems violate certain common sense fairness intuitions" applies only to "ordinal" systems, where your either pick a favourite or put options into preference order. Range voting is a Cardinal system, so this theorem doesn't apply. To see how it works, suppose all the Martian voters give 10 to the closest location, 7 to the next-closest, 4 to the third, and 0 to the furthest away. The most "central" location, North, wins - and this seems fair, since no-one scores it zero - i.e., it isn't any voter's least-favourite option (the same is true for South, but there are more people in North). If someone were to propose an "optimally central" location between the 4 Mars bases, to minimise the average airport travel-distance per Martian citizen, I would expect this proposal to beat North in a Range Vote, just as it should. In most other voting systems, such an obviously-fairest solution would almost certainly fail. It's worth mentioning that Star voting, cited as an example of voting systems to which the Arrow theorem applies, is actually a modified version of range (or "score") voting. The author really ought to know this!
I was thinking the same thing. The fact that they mention Condorcet methods when Condorcet doesn’t even always produce a winner (let alone a “fair” winner) is a shame. They should’ve known that Arrow’s Theorem doesn’t apply to *all* voting systems (tho a stronger voting theorem does state that all voting systems, including Range and STAR, can be subject to strategic voting).
@@ClementinesmWTF That's right. However, the negative impact of "rational" strategic voting in range-vote elections is quite limited: it may sometimes produce a theoretically "less than ideal" result, but it can't lead to the disastrous result of making a winner of a candidate most voters would honestly rate very negatively. Plurality voting and the various ordinal systems can all do that.
I like this and I feel that this as close as to getting to an ideal decision. On the other hand, This voting method may result in selection of candidate which no voter wants. Voters would feel they were deceived and would ask for re-election.
@@anniekallen4472 I'm afraid I disagree. Warren Smith, the leading expert on range voting, has a critique of star voting on his website, which I find convincing: rangevoting.org/StarVoting.html. Basically, star voting sacrifices many advantages of simple range voting, supposedly to sell itself to skeptics of new voting systems. The only theoretical advantage of Star is when there is a much higher level of strategic voting than is ever actually observed in real-world practical studies of score voting. That said, Star is better than all the ranked-choice options, as well as FPTP.
New Zealand has the best voting system according to me. MMP (Mixed member proportional). Every citizen gets 2 votes - one for the party, another for the candidate.
This system isn’t really a voting system like the ones in this video since it uses systems like first past the post to determine candidate winners and party seats are made proportional to the votes they receive. This seems to be less of a way to determine winners and more of a way to allocate power among the parties and individuals that do win.
@ In both Australia and New Zealand there are many parties but only 3 or 4 regularly win seats, the other minor parties are often single issue parties somewhat like lobbies. The competition for government boils down to 2 major partys, centre-left vs centre-right. If the voting system changed Americans would likely see the Greens and Libertarians grow and maybe a split between the Tea Party and GOP. But the major votes would still be between the Democrats and Republicans. From an outsiders view, one of the weirdest things about American politics is the 2 major partys aren't left and right, but centre-right and right-wing.
Yep - and then we ruin it by having a 5% minimum vote requirement (even though there are 120 seats so 0.83% is enough for a seat), plus the media devote 95+% of exposure to two parties and pretty much ignore the other fourteen. It's sad to see the potential of MMP go to such waste.
@@almightyhydra The vote Minimum is of course arbritrary( 4% in Austria, 5% in Germany 10 % in Turkey). In the Weimar Republic, and other countries this does not exist and often lead to scattered parliaments, blockades, chaos and coalitions of many Parties, that can't achieve anything and break up easily.
Rank by Point based system (Position Voting), some sporting event uses it like F1. 1st place - 4 points 2nd place - 3 points 3rd place - 2 points 4th place - 1 point Max total = 400 (1st) + 300 (2nd) + 200 (3rd) + 100 (4th) = 1000 points If using 2:13 as example (voting for one answer) then final results for: W: (4 x 42) + (1 x 15) + (1 x 26) + (1 x 17) = 226 S: (2 x 42) + (4 x 15) + (3 x 26) + (3 x 17) = 273 N: (3 x 42) + (2 x 15) + (4 x 26) + (2 x 17) = *294* E: (1 x 42) + (3 x 15) + (2 x 26) + (4 x 17) = 207 (If voting fore share number of representatives) then it will be distributed with 22.6%(W) : 27.3%(S) : 29.4% (N) : 20.7%(E) ratio of politicians on each location (out of 1000 points) IF all for site get 25% equally (or 250 points) then it simply means equal number of representatives on each site, but this become unlikely to happen if more than 2 political flavor existed or in a population size of a "large" country.
In any singular instance, you can point to a deficiency, but that is letting the perfect be the enemy of the good. What you have to look at is the system over numerous votes, and pick the best from there. Something like sortion, for instance, has a greater chance for the majority to win any particular vote, but isn't always guaranteed. But statistically, the will win the majority of elections, with minority views also wining a proportion of the time.
While there is no perfect voting system, some systems (e.g. STAR and STV, both not mentioned here) are way more representative and closer to the preference of most people. Probably the worst is plurality (first-past-the-post). CGP Grey has a good series about them. STV is great for maximum representation in parlimantary elections while STAR is great for single winner elections.
@@CTimmerman They are not quite the same. Instant Runoff Voting is for electing a single winner, whereas Single Transferable Vote is for electing multiple winners.
It's pretty clear that instant runoff has the best balance of fairness and simplicity. Perfect? No, that's impossible. But still excellent. I will be voting for it on the Massachusetts ballot this November and look forward to using it in future elections.
IRV is pretty good, but you might want to take a look at this page: en.wikipedia.org/wiki/Template:Comparison_of_Schulze_to_preferential_voting_systems IRV fails a bunch of criteria that I would personally consider important. For starters, IRV isn't monotonic, meaning that in some cases, ranking a candidate higher can prevent them from getting elected, and ranking them lower can get them elected where they otherwise wouldn't have been. It's also not a Condorcet method. Condorcet isn't a voting system, it's also a criterion. A voting method can be called a Condorcet method if it elects the Condorcet winner if there it one. It's where there _isn't_ one that they diverge. I believe Ranked Pairs is pretty good. It takes every possible matchup and ranks them by the percentage with which that matchup was won. Then you go down the list, "locking in" each candidate's relative ranking. When a result would cause things you become circular, you flip it. Since you start with the greatest victories, the people whose preferences are ignored (i.e. when a result is flipped) it ignores the least possible amount of people. Still, it seems like you've thought about this a lot, and ultimately you should go with the system you think is best (for example IRV has later-no-help and later-no-harm which many others don't). After all, what's voting for if not choosing the best out of a set of people's preferences? I wish you the best of luck this autumn.
@@Gamesaucer In the system where I live ranked pairs sounds like horror to vote on. I feel like that works better when the number of pairs is still quite limited.
@@hkr667 What do you mean? You'd still just cast a single ballot. You'd just have to rank the candidates, and everything else would be extracted from that, same as other ranked voting methods. And in fact, since Ranked Pairs allows putting candidates on equal ranks, you can just, say, vote for the top 5 and leave the rest unranked. They'd collectively be your least favourite choice as a result. What you could even do, if for example things are organised into parties, but each member of each party can be elected, is a system where you _rank_ the parties, but give a single _preference vote_ to your preferred candidate within that party.
@@hkr667 E.g. Say that there are 6 candidates (i.e. 15 pairs) and you rank them in this order: A > B > F > D > E > C. Since A is at the top of this list, this implies that not only (A > B) but also (A > F), (A > D), (A > E) and (A > C). Similarly, C is at the bottom so it's implied that (B > C), (F > C), and (D > C). The same logic applies to (B > D), (B > E), and (F > E).
@Gamesaucer @@Stratelier Yes, but in the video they said you could have a rock-paper-scissors situation. A > B, B > C, C > A. I can't really imagine a situation where it would happen though. But I can imagine ties - A > B, B = C.
Like in this map, the correct space port is the one with the shortest length from each base to the port, added together and then multiplied with their population as a percentage of the total population. Then the most amount of people travel the least amount of distance. Somehow we need to define a map of the political spectrum, maybe using these voting systems with their different strengths and weaknesses in combination to define where people are located within the map. A politician must then, before anything else, get a thorough check for his position on the map. This could help clarify where they stand on certain issues. Huge questionnaires would have to be conducted, scaling what people think is most important. The size of the issue on the map could be its importance, the length and direction from the origin could be how much it matters for a specific party and which party the issue matters for. Most definitely doesn’t work in practice, but could be fun to test the idea
Just take the rankings from 1:22 and assign points for each vote. For example: 5 points for someone's first choice, 3 for their second, 1 for their third, and 0 for last. Doing this results in North winning, South finishing 2nd, West finishing 3rd, and East finishing last. Almost all reasonable point values (eg. 3 for 1st, 2 for 2nd, 1 for 3rd, 0 for 4th) end in the same result. I was only able to get a different winner once I made 1st place votes worth 4x the point total of 2nd place votes, which would obviously be absurd.
That system is called Borda Count, and its main failure is that it can be gamed through candidate cloning. For example, in a multi-ideology race, the one that runs the most candidates will likely win.
3:56 called non-transitive preferences. Logically though, preferences should be transitive. So condorcet should work 95% of the time. In the 5% case, just use a different voting system to break the tie.
I've seen convincing arguments that one of the best methods of determining the Condorcet winner in a single round of voting and best protection against favorite betrayal and other vote-splitting side effects, is STAR. Basically the optimal combination of IRV and approval/score. Not saying it's perfect (since there is no perfect), but since FPTP/plurality is basically the WORST, i'd say it's a vast improvement.
Because all voting systems will likely have a flaw, the question then is “which is the most fair way to pick someone or something?” We can all agree that minority rule is unfair (which is when you win with less than 50%). But instant runoff is probably the antidote for this because based on preference there will always be someone that gets a majority of support. In US elections this is how you can help ensure that whoever is elected has a majority of support. ME already does it and it works.
@@10gamer64 depends what you name as justice. US is like a union of 51 countries, so having 51 elections in each country and select 1 liked in the most countries is reasonable from this side.
@@dmitrizaslavski8480 Why would it be too complicated? Not at all, and I d'ont see why it would be for not engaged people the same as plurality, they will be able to choose whomever they want
That might actually be the best. In our case it would result in 126 points for W, 194 for N, 107 for E and 173 for S, leading to North being selected as the winner, which seems like the best and fairist outcome.
I have I think a very good idea. You make a vote where people vote with 5 different systems, so that each criteria (at the end of the video) is filled only once. After that, you look at the city that has the most victories. If there is a tie, the most second, most third... if the tie is perfect, second round.
Agreed. Also, while Condorcet can "cycle," Approval approximates to Condorcet without cycling. And it is *much* easier for the voter than any ranking system!
@@WG55 And approval voting's failure of the Condorcet criterion is illusory because it is based on a slightly different value system. When necessary, it will, by design, select candidates who may have fewer true supporters, but who have a higher share of voters who find them at least acceptable.
Agreed, Approval voting and STAR voting are the best voting methods I've found. Approval does a really solid job at being fair/ accurate while being dead-simple, and STAR voting does a knock-out job while still being more simple than IRV. electionscience.github.io/vse-sim/VSEbasic/
@@darbyl3872 Yeah, that's one of the things I love about STAR voting. It's very flexible and the voter can use it however they want: Like an Approval ballot, like a ranked ballot, like a scoring ballot, even FPTP if they want.
You should add STAR Voting to your video. No one has a video on something big that is wrong with STAR. You did a nice job choosing key weaknesses in the systems you covered.
The two-party system exists in both the US and the UK and has thus almost a monopoly in the English-speaking world. But look no further than the EU and you'll find a great number of systems that we who are blessed to live in the EU consider standard: a multi-party democracy, where the parliament is proportional to the vote.
For Condorcet, if you do get a loop like that, there are ways of resolving it, either baked into the system, or using a secondary system to pick from that loop (known as the Smith set). What you can also consider is what kind of tendencies a voting system has. FPTP, especially if it is a regularly held vote with the same or similar candidates, such as typical elections, has a tendency to result in effectively a two-party system because it creates situations in which only two parties/candidates have a reasonable chance to win. Instant Runoff has a tendency of the centre-squeeze, that is, centrists often receive fewer first place votes and are squeezed out, and split over the more extreme candidates, even though most would probably consider them a reasonable option (albeit not necessarily their favourite) and prefer it over the other extreme. Condorcet has kind of the opposite effect, and is more likely to elect moderates who may not be many people's first choice, but who are deemed reasonable by most, and preferable by a (varying) majority to every other candidate. Two round systems can kind of go either way depending on how people (tactically) vote: if a centrist and an extremist go to the second round, the centrist usually wins by getting the votes from the other end's extreme (unless there is something going on where people prefer any anti-establishment extremist over an establishment centrist), whilst if two extremists go to the 2nd round, then the centrists are squeezed out and forced to pick one of the extremes. For picking something like a president, having a more centrist uniting factor may be preferable, and thus Condorcet would be preferable, but for picking local representatives for a national level (e.g. UK constituencies), it may be less ideal as it would cause mainly centrists to end up representing the population, rather than being more representative of the varying views in the country (but then again, electing local representatives for a national level is generally bad at being representative of the differing views that exist within society; you need something more proportional for that).
In the instant runoff scenario, if you add the votes up like points, awarding no points to the least favorable position and three points to the most favourable position, two points to the second most favorable, and 1 point to the third most favorable (per voter) the North base comes out on top.
I’ve finally decided something, there isn’t any one voting system that can be applied for every situation, in this case approval voting makes the most sense (voting for all candidates you’re ok with winning), but in political elections with more then two candidates instant runoff is superior
Nice explanation of Arrow's impossibility theorem! In Kenneth Arrow's words, "Most systems are not going to work badly all of the time. All I proved is that all can work badly at times."
Also in Kenneth Arrow's words from a 2012 podcast: "Well, I’m a little inclined to think that score systems where you categorize in maybe three or four classes probably (in spite of what I said about manipulation) is probably the best."
I'm not sure if this system also exists, but a voting system that seems alright to me would be using the ranking that we saw for the "Instant runoff voting" system, but instead of counting the votes in the way that was shown, instead, each voter's first place vote would be worth 4 points, their second-place vote would be worth 3 points, third-place would be worth 2 points and last place would be worth 1. Then, we would add up all the points for each place. If this method is used, then East actually ends up with the least amount of points and North ends up with the most points.
STAR (score then automatic runoff) voting is essentially what you described mixed with Condorcet voting, and generally produces the least disliked result.
@@stirlingblackwood Oh, that's great to know! Thanks for informing me. Do you have the source that shows that it generally produces the least disliked result? I've love to read it.
I like ranking and the weighting votes. So for this example, each colonist’s first choice would receive 4 points, second 3 points, third 2 points, and first 1 point. You add up each choice’s total points, then the highest is the winner. So: West= (42*4) + (15*1) + (26*1) + (17*1) = 226 East= 207 North= 294 South= 273 So North would win, which makes sense because it is the most central to all the colonies. It’s not a perfect system, but it seems better than the electoral college at least.
4:41 very interesting. Are there any sources online where I can see how they proved it? :) Great topic with not too much info available online, so I am very happy it got posted :D I would love to see more about that and related stuff!
Lol I am surprised that my first attempt, searching for "prove that votes with more than two cannot be fair", immidiately got me to the wiki page about "Arrow's impossibility theorem" that states exactly the statement they mentioned in the video. I remember when I looked up things about this topic some years ago I did not find much; and now this is exactly the stuff I wanted to find! :D
Look up the videos about Arrow's theorem on the Infinite Series channel, they explain pretty well its proof (sometimes wikipedia can be way too deep and technical in math stuff)
STAR voting works well here, giving the Condorcet result while only asking voters to 5-star-rank the options. It also puts North against West in the Automatic Runoff round, quite reasonably consolidating the "eastern bloc" vote for the higher-population North.
Approval or STAR voting has been shown to be the most optimal system at maintaining the voters' preference and eliminate the need to vote against your favorite pick.
For the Condorcet method, there's actually a good solution. You take the portion of votes N won against S, E, and W and get the geometric mean, the "volume" of support for N over all others. Do the same with each of the other candidates and you have a metric to rank them to see which is most preferred, without a circular outcome.
In Denmark we have a proportionalt election system which reflects the political distribution in the population fairly. If 40% of the population votes for party A then it will have 40% of the parliament's votes. The leader of the government becomes the one who can gather the most members of parliament behind him (which is not necessarily a majority of all members of parliament!). Currently, the parliament has 11 different parties. This electoral system promotes compromise and cooperation between the parties since a single party rarely (never) has an absolute majority. Perfect? Hardly. But it reflects the views of the population. The underlying mathematical distribution model is a bit complicated, but as a voter you only have to deal with two things in practice, namely: A) Will the party I am voting for obtain a minimum of 2% of the votes (if not, the party is not represented in parliament). B) Which party / candidate will I vote for.
Instant runoff works pretty well because usually there are 2 viable options that get the majority of the votes, so every single person can vote both on who they most want, as well as who they prefer out of the 2 big options(mostly the 2 major parties)
Equal Vote Coalition has a RUclips video called "How Does STAR Voting Work?" which walks through the basics of STAR Voting. From the voter perspective, it feels very similar to ranked choice voting, but the tabulation does a better job at picking a winner that represents the will of the overall electorate.
Instant runoff guards well against the spoiler effect, e.g. a third party candidate coming in from the fringes and stealing votes from a more moderate candidate. They will typically get eliminated first, and their votes redistributed as if they had never run. However, it doesn't guard well at all against a centrist candidate being squeezed from either side. Centrists will also be quick to depart the race. In the end, you still have the same two-party system that Plurality incentivizes.
Aproval voting (vote on one or more that you aprove) is the best. It would go like this: W: 42 N: 26+42+15=83 S: 15+26+17=57 E: 17+15=32 North wins, wich is the fairest.
Raise your hand if you want to go to Mars! *No one does anything.* Raise your hand if you want to go on the moon! *Still, everybody isn’t raising their hand.* Raise your hand if you want to go to Jupiter! *Pure silence..* _Okay, we’re sending everybody to the sun..._
I propose discussion. Why do you want the space port where you do? Because it benefits you? Well then, that's no better a reason than having it benefit someone else. If there truly is no benefit above that of convenience, then I propose it be equidistant between East and West. Yes, it benefits North and South far more, but neither East nor West - in this scenario - is favoured over the other; they needn't feel forgotten and outcast. What if either East or West had only a single citizen? Would matters change? No; not If you expect the vacant spaces to fill, in time. This scenario can be likened to that of our present; we all need to shop, and many of us travel to work; everything can't always be equidistant to everyone - but we can meet the matter with thought, discussion, and a empathetic care for others as well as ourselves.
As a Tasmanian Hare-Clark with the Robison Rotation is the best voting system. So multi member districts (7 members per district is ideal) using ranked choice (so rank the options then last place gets eliminated) and then for the ballot it's self rotate around each party's place on the ballot and also rotate each candidates position in their party's list of candidates. This was everyone has a say you still have a local member and internal party politics doesn't affect candidate placement.
FPTP was a British Colonial export, e.g. overlay list of all countries British colonized (past & present) vs FPTP countries (some have moved on like Canada & Australia). Therefore FPTP are one of the oldest, not great for solving multiple choice question and usually end up having 2 major flavors close to the wire of 49:51%
One element of instant runoff voting not shown is that someone can simply refuse to cast a vote fo tor one of the options. For example, vote 1 for north, 2 for south, but refuse to put a 3 or 4 if they think east and west are completely against their interests.
Participation is a really weird criteria when you think about it. For example, under instant-runoff, _the order in which candidates are eliminated matters._ E.g. if there are just 3 candidates and the initial vote tallies show A > B > C, the second choices from A and B voters are _never actually used._ And if the race between B and C is close, it actually gives tactical incentive for some C voters to NOT EVEN VOTE, because if enough C voters participate such that C > B, then B is the party whose second-choices (possibly for A) get counted. I personally like one of the points methods, such as (points) = (1 / rank) e.g. 1st = +1, 2nd = +0.5, 3rd = +0.33, etc, and the highest final tally wins.
This is a very good point. One observation I would make is that the ranked ballot methods all have the same flaw: they quickly become unmanageable the more candidates you have. e.g. if you have 20 candidates, it is a huge mental chore to keep track of which ones should be, say, 15th and 16th. A large field of candidates would be the norm if we didn't have to worry about spoilers and clones. Ranged ballot systems, e.g. STAR Voting, don't have this problem. You simply go down the list and assign each one a score within the allowed range (usually 0-5 or 0-10), following whatever criteria make sense to you, without the need to keep all your entries unique.
Wow, thank you for making this video! I enjoy your added perspective on the subject from that of CGPGrey's. Personally, this is the type of discourse I want to see on the subject of voting. What is voting?? And how can different social structure's accomplish it? And what is it's relationship to differing social structures, such as democracy and fascism?? Yes, fascist governments have held their own elections. Context matters! This is far more helpful to our country than the bipartisanship harassment policy where you, the victim, get blamed for not giving them your consent by voting. Especially considering how much your get "punched in the face" for not voting, you must have a very developed and nuanced argument for why to continue to not vote. Which only serves to represent how broken the system truly is if a significant portion of the populace do not have sufficient faith in their government in order to vote in elections despite all the effort pushing them into doing so. It's important to remember that the citizens of a government are the victims of a failed state, To hold them personally responsible is disingenuous at best.
2:38 but would the South base think the same and vote for East? so the South base will have all the East vote, while the East base will have all the South vote, and thus both of them will still be removed? also, how about assigning a numeric value to the rank of instant runoff (like 4 for 1st place, 3 for 2nd, 2 for 3rd, and 1 for 4th) multiply with the number of voters who choose that option, and then pick the highest result? As such, the North base will win with 294 votes, South base come second with 273, West base with 226 and East base 207 - seem fair to me.
To be honest I prefer instant runoff voting and of course the first is the least two choices for the second because there is no way you can invent a system that going to make everybody comes of it a winner. But what you can do is invent a system that gives the candidate more than one chance with the people and that is instant runoff voting. for example there are 6 candidate and one of them has never been chosen as the first but he has been chosen as the second and the third by all the voters. This candidate will be the winner and with or without the voters knowledge they actually agreed on a middle ground.
CGP Grey has taken over Ted Ed
Expecting to see this comment as soon as i saw the title
I approve of this coup
i wonder how he change his voice
Came here to say that :)
@@merrittanimation7721 The rule for rulers.
One thing not mentioned in this video is that plurality voting also "incentifizes lying about your preferences" not just multiple rounds. This happens (kind of) in the US, where people might want to vote for a third party (ex. green party, libertarian party, etc.) might end up voting for the democratic/republican candidate because they think that the third party they support is unlikely to win.
For example, South and East base residents could agree to vote for North to keep West base from winning
And it gets worse - not only does supporting a third party not help them win, but it also draws votes away from the major party you otherwise would have supported. Too often, this leads to your worst option getting elected. I like to describe Plurality as mathematically driving us to a two-party system, and this is why.
The only way you're going to change it is to stop being short sighted and vote for the best candidate regardless of electability.
ted-ed:Which voting system is the best?
Me: I don't know, that's why I'm here
I think the best is Distributed Voting (DV); here you find a description of the method (see Procedure):
electowiki.org/wiki/Distributed_Voting
@@robbe1534 Oh that's really cool and interesting, I didn't know that existed. However, wouldn't that (over time) incentivize voters to assign all 100 points to their candidate of choice since, if they don't, someone who does assign all points to their candidate of choice will inevitably overshadow their first choice
This depends on how you normalize the scores. Personally I prefer a simple free scoring system. Its more intuitive and still allows better expression than any of the ranked methods shown on the video.
@@jacobandrews2663 DV eliminates the worst candidate by redistributing the points, so if you give all 100 points to a candidate who is eliminated among the first, the vote becomes irrelevant. To avoid this it's better to distribute the points.
Also, you consider that using range [0,10] (then converted to 100 point votes), it's difficult to accumulate points.
@@pepemotion DV can use votes with range [0,10] (then converted to a 100 point vote), appearing from the voters' point of view, as a free scoring system.
Highlander method: voting to eliminate the worst, until theres only one.
Yes the purge
But that would be a very long process
Whichever way will be fine as long as it is not China's way
China's way = the non-elected govt makes all the decisions forever for the country with the claim that these are the people's unanimous decisions.
for me it’s also unfair. like if everybody knows you’re a threat then they would join forces to eliminate you first.
*whistles* do mi re fa~~ 🕊️
0:13 I love how you guys put the sound of a person breathing in a space suit and the sound of a space ship. Just gave the video a special touch 💝
Radical Change 140 likes with only one reply?
183 likes with two replies?
206 like with three replies?
217 with only four?
Wow , I didn't even realized that😅
I actually took a whole class centered around the mathematics surrounding politics. We spent weeks just on voting systems
In which class?, game theory?
Different voting systems should be in basic civics classes.
@@Knightmessenger yes! Its scary how clueless we are with voting
Fascinating! I'd like to learn more myself
Did you vote which one was your favorite?
Maybe it’s the friends we made along the way
No, NO, NO WE CANNOT KEEP THIS MEME, JOKE, WHATEVER THE F*CK IT IS!
Adam Abdi why not?
😂
The friends we made along; the way home
Nah
All too often we laud "popular vote" as a majority system when in reality it's most often a plurality in most countries and this video very well illustrates how that system can be flawed and other systems in comparison. Kudos!
Yes,
USA for example, has most of the population in East and West coasts. So if they had a popular vote system, everything would be unfair to the farmers in the vast land in the middle.
Current system they have is worse, but popular vote doesn't solve everything.
what do you mean? most countries use a two rounds system, not a straight up plurality.
@@phs125 However, there are far less farmers in the interior of the country than there are people who live on the coasts. Does it really make sense to divide people up by where they live? No. The votes of the people on the coast should count just as much as anywhere else, and the election should be about the policies the candidates espouse, because not everyone who lives on the coast is going to vote in one way and not everyone who lives in the heartland will vote the same way.
Almost no countries use plurality voting. They use proportional voting.
A plurality and proportional voting system aren’t the same
I participate in condorcet voting everyday when playing rock paper scissors
It's the only true way of making any sort of decision.
And choosing a starter Pokemon!
Well said
@@Warrka4 There's a similar one. One side picks odd, the other even, then like RPS both show their numbers at the same time, the total sum of the numbers will decide if even wins or odd.
JonatasAdoM oh yea odd or eve. That's basically like cricket though so it doesn't have the same system as rock paper scissor cause it in voles more of statistics than this.
0:00 Intro
0:28 Graphical Modelling
0:40 1. FPTP/plurality
1:12 2. IRV/instant runoff
2:16 3. Two-Round runoff
2:46 Strategic/Tactile voting
3:01 4. Pairwise/Condorcet method
3:50 Cyclic Dep Tie (non-transitivity)
4:04 Reflection
4:14 Other methods (names)
4:41 Arrow's Impossibility Theorem (hint)
4:55 End
What is a horse’s primary concern when voting?
A stable economy.
booo!
This guy!
@Krishna Laxkar 🔫 you got one minute to understand the joke, or you're getting the Glock.
Thank you ^_^
That said, when asked if it was in favor of the bill, it said neigh
Let's give a big hand to Ted-ed for there immense hard work for making education videos on all aspects of life.
Yea I don't really dig TED Talks but TED-Ed is the bomb
*their
Crash Course too!
These channels are revolutionizing education!
"Which voting system is the best?"
"Yes"
I think a better answer would be "no"
bad comment
yes, but actually no
Yesn't
Give people the illusion of choice to placate them, then do what ever you want anyway. Problem solved.
Ugh. I love it when he says “not necessarily.” Iconic.
*Meanwhile Putin and kim Jong unn :* Uh huh
Xi Jinping :" Hmmm..."
America and China: hmmmm
Britain: this is not the vote you were looking for
dont understand first of all russia is a democracy and nk is a dictatorship
@@moustafakhattab8142 No, Russia is an oligarchy led at the moment by Putin. The basilar freedoms are limited, (not eliminated) and it's impossible to candidate aganist Putin (so he can win everytime)
*_“Many forms of Government have been tried, and will be tried in this world of sin and woe. No one pretends that democracy is perfect or all-wise..._*
*_Democracy is the worst form of Government except for all the others.”_*
- Winston Churchill, 1947
Direct democracy > representative democracy
Feudalism has been proven to be pretty stable, just saying.
@@molybdaen11 lol. But didn't the serfs usually overthrow the king's?
@@daotheeternalnamelessbeyon8778 Anarchism, and especially anarcho-capatalism is contradictory.
@@samuelforesta It happened from time to time, usually in areas like china where the king/emperor grown to powerful for external treats and forgot where his power came from.
At one point the people have enough and replace the emeror with one of they own.
This new emperor family will behave for a few generations until they forget the lesson as well and the circle continued.
In europe uprisings were rare because most of the time the ruling familys had to defend themself from competition and needed the support of the people.
I know CGP Grey is gonna like this episode of TED-Ed
Taco?
Nah, CGP Grey's videos were significantly better. This one doesn't even mention MMP, which is undeniably the fairest voting system in practice.
@@LeagueUnionSevens You mean STV not MMP
@@stonferen584 No, I mean a parliamentary system elected under MMP, which allows for broader representation than STV.
Look at CGP Grey's videos on STV and MMP to see what I mean.
@@LeagueUnionSevens And how does it allow for more broad representation?
I'd choose North no matter where I'd live, it being centermost.
Unfortunately, most people (at least from the ones I know) wouldn’t
Ah, a centrist
That idea is called compromising. Everyone doesnt necessarily benefit but it's the anwer that satisfies most
No
Just imagine going from Brazil to china each time you have an important work.
No one would.
Even if it was North Korea?
This video was brought to you by CGP Grey
Hey what is CGP Grey? I am not familiar with that term.
@@bestrongcourageous2932 it's a youtube channel known for their videos on voting systems
@@me_myselfand_i2099 and other very interesting videos
@@Zeragamba well he's best known for voting systems which is why I mentioned those videos. He does make other interesting videos as well
CGP Grey's videos were significantly better. This one doesn't even mention MMP, which is undeniably the fairest voting system in practice.
I live in Switzerland and most people love how the president is chosen. We vote for the party we want and then the party’s with the most votes can get on to seven on the seven seats available. No one has more than 2). Then all 7 elected have the same power and collaborate. They are each president for one year but have no additional power.
I also love „Volksabstimmung“ since it really gives the power to the people and not just the parliament
I can imagine US parties splitting in the same way monopolies "don't exist" while all being owned by mega corps. That is, Dems or Reps just form subsidiary parties that are legally distinct but the same rotation of people with the same financial interests. Our political parties are very good at subverting systems meant to benefit everyone, though.
Seems pretty fair. They also apply direct democracy methods. Switzerland has a system closer to being called "democracy" than anyone else. Their methods indeed have similarities with original democracy in ancient Greece.
The Swiss system is a close to a real democracy as it gets but that works better with a small, affluent population. But it also means that sometimes it takes too long a time to reach decisions and of course its success depends on voters being well informed and interested in politics.
@@TheFreakyable the problem with the people deciding is that it is easy that many do missinformed decisions but on the other hand you cant just deny them the vote
Random recommendation. Read the title as “Which VOMITING system is the best”...still clicked on it
I’m told that if you think you are likely to vomit soon, drinking lots of water (but not like, extreme amounts, just large amounts) can make it much less unpleasant when it happens.
Just some advice my dad gave me.
@@drdca8263 thank you!
@@drdca8263 :-) Sounds as if a father (who was drunk in his younger times at every party) gave his now teenage child good advices how to deal with alcohol.
animal friend I can see how it would seem like that, but my understanding is that my dad has always disliked the taste of alcohol enough to prevent being drunk more than a handful of times (a single time ?)
@@drdca8263 Ok. That's why I put a smiley at the beginning.
Meanwhile in Pakistan: *The military will decide your fate*
Egypt: first time?
Oof
The USA: Everytime. No matter how you vote, we still go to war with someone to smash their nose.
@@fisherfriendman Nah Republicans after the war generally became more lenient
Cambodia: no matter how you vote it will some how be the ruling party
I think Borda Count, although not perfect, would probably be the best option. Convert the rankings into points, so a first choice gets 3 points, a second choice 2 and a third choice one, and the winner is the one with the highest overall score. It therefore elects the overall most popular/least unpopular option. In this example the winner would be North, which on the map does look like the best compromise.
Admittedly two of the methods demonstrated also make North the winner in this example, but the problems with both are discussed.
I prefer condocet
@@jamesabestos2800 The main problem with Condorcet, as mentioned in the video, is that it doesn't always produce a winner. Aside from that, I agree, it's a good system.
STAR (score then automatic runoff) voting would actually accomplish this even better than Borda count.
@@stirlingblackwood STAR Voting does seem like a reasonable option as long as it successfully overcomes this problem in Range Voting where people vote tactically by only using the maximum and minimum scores. I'd need to see some real life figures to be able to assess that.
My favorite voting system is Black's System, which is kind of like a combination of Borda Count and Condorcet. It puts each candidate head to head, and if there's a candidate who's able to beat every other one in a one on one match (and is thus a Condorcet winner), that candidate wins automatically. Otherwise, it plays by the same rules as the Borda Count system.
I would like to mention though that there are several solution to that paradoxical a>b>c>a in the Condorcet voting, my favourite of which is the Schulze method. It mathematically finds the candidate that would make people the most happy, even in a case of a circle as mentioned. Its downfall is that it's more prone to draws than most other voting methods, but these are still quite rare.
Short answer: Not the one used in Argentina.
Kam which one is used in Argentina
If you know, you know ahre
Argentina actually uses one of the best electoral systems ever created: proportional representation for congress and ballotage for the president.
What voting system is used in Argentina?
Condorcet is the best one listed. The fact that there are sometimes circular ties is not a flaw of the voting system. Any election can result in ties. Condorcet just makes those ties apparent instead of choosing an unrepresentative candidate in those cases. That's a good thing.
STAR (score then automatic runoff) voting is essentially score voting with a Condorcet-like runoff phase. It does a better job minimising tactical voting incentives.
@@stirlingblackwood Yes, STAR is very good too
according to simulations it is as good as ranked robin (described in the video) but easier to understand and calculate@@stirlingblackwood
“Our lives begin to end the day we become silent about things that matter.”
― Martin Luther King Jr.
What is the connection between this quote from Martin Luther king and this video
@@tsarplague1228 "hmm, what's voting?"
@@revspikejonez Not really the question on what is voting, but what is a fair system
@@tsarplague1228 the people who don't vote.
An active refusal to vote with a nuanced understanding of the position is a refusal to give your consent to the government. DO not blame the victim. If you look at Martin Luther King Jr, he would not have supported voting as a means to make your voice heard. He supported civil disobedience to accomplish that end. Particularly in his later years. Organized and sustained social movements accomplish things voting has always managed to suppress.
I'm shocked to see no discussion here of range voting, in which voters give each option a score, say from 0 to 10. Key advantage: the theorem mentioned (Arrow's impossibility theorem), which proves that "all voting systems violate certain common sense fairness intuitions" applies only to "ordinal" systems, where your either pick a favourite or put options into preference order. Range voting is a Cardinal system, so this theorem doesn't apply. To see how it works, suppose all the Martian voters give 10 to the closest location, 7 to the next-closest, 4 to the third, and 0 to the furthest away. The most "central" location, North, wins - and this seems fair, since no-one scores it zero - i.e., it isn't any voter's least-favourite option (the same is true for South, but there are more people in North).
If someone were to propose an "optimally central" location between the 4 Mars bases, to minimise the average airport travel-distance per Martian citizen, I would expect this proposal to beat North in a Range Vote, just as it should. In most other voting systems, such an obviously-fairest solution would almost certainly fail.
It's worth mentioning that Star voting, cited as an example of voting systems to which the Arrow theorem applies, is actually a modified version of range (or "score") voting. The author really ought to know this!
I was thinking the same thing. The fact that they mention Condorcet methods when Condorcet doesn’t even always produce a winner (let alone a “fair” winner) is a shame. They should’ve known that Arrow’s Theorem doesn’t apply to *all* voting systems (tho a stronger voting theorem does state that all voting systems, including Range and STAR, can be subject to strategic voting).
@@ClementinesmWTF That's right. However, the negative impact of "rational" strategic voting in range-vote elections is quite limited: it may sometimes produce a theoretically "less than ideal" result, but it can't lead to the disastrous result of making a winner of a candidate most voters would honestly rate very negatively. Plurality voting and the various ordinal systems can all do that.
I like this and I feel that this as close as to getting to an ideal decision.
On the other hand, This voting method may result in selection of candidate which no voter wants. Voters would feel they were deceived and would ask for re-election.
I thought the same thing! No voting method is perfect, but STAR voting goes a lot farther than other methods. www.starvoting.us/criteria
@@anniekallen4472 I'm afraid I disagree. Warren Smith, the leading expert on range voting, has a critique of star voting on his website, which I find convincing: rangevoting.org/StarVoting.html. Basically, star voting sacrifices many advantages of simple range voting, supposedly to sell itself to skeptics of new voting systems. The only theoretical advantage of Star is when there is a much higher level of strategic voting than is ever actually observed in real-world practical studies of score voting. That said, Star is better than all the ranked-choice options, as well as FPTP.
New Zealand has the best voting system according to me. MMP (Mixed member proportional). Every citizen gets 2 votes - one for the party, another for the candidate.
@ Direct democracy is not possible in big democracies
This system isn’t really a voting system like the ones in this video since it uses systems like first past the post to determine candidate winners and party seats are made proportional to the votes they receive.
This seems to be less of a way to determine winners and more of a way to allocate power among the parties and individuals that do win.
@ In both Australia and New Zealand there are many parties but only 3 or 4 regularly win seats, the other minor parties are often single issue parties somewhat like lobbies. The competition for government boils down to 2 major partys, centre-left vs centre-right. If the voting system changed Americans would likely see the Greens and Libertarians grow and maybe a split between the Tea Party and GOP. But the major votes would still be between the Democrats and Republicans. From an outsiders view, one of the weirdest things about American politics is the 2 major partys aren't left and right, but centre-right and right-wing.
Yep - and then we ruin it by having a 5% minimum vote requirement (even though there are 120 seats so 0.83% is enough for a seat), plus the media devote 95+% of exposure to two parties and pretty much ignore the other fourteen. It's sad to see the potential of MMP go to such waste.
@@almightyhydra The vote Minimum is of course arbritrary( 4% in Austria, 5% in Germany 10 % in Turkey). In the Weimar Republic, and other countries this does not exist and often lead to scattered parliaments, blockades, chaos and coalitions of many Parties, that can't achieve anything and break up easily.
Rank by Point based system (Position Voting), some sporting event uses it like F1.
1st place - 4 points
2nd place - 3 points
3rd place - 2 points
4th place - 1 point
Max total = 400 (1st) + 300 (2nd) + 200 (3rd) + 100 (4th) = 1000 points
If using 2:13 as example (voting for one answer) then final results for:
W: (4 x 42) + (1 x 15) + (1 x 26) + (1 x 17) = 226
S: (2 x 42) + (4 x 15) + (3 x 26) + (3 x 17) = 273
N: (3 x 42) + (2 x 15) + (4 x 26) + (2 x 17) = *294*
E: (1 x 42) + (3 x 15) + (2 x 26) + (4 x 17) = 207
(If voting fore share number of representatives) then it will be distributed with 22.6%(W) : 27.3%(S) : 29.4% (N) : 20.7%(E) ratio of politicians on each location (out of 1000 points)
IF all for site get 25% equally (or 250 points) then it simply means equal number of representatives on each site, but this become unlikely to happen if more than 2 political flavor existed or in a population size of a "large" country.
That's called Borda, and its main weakness is that running multiple candidates increases your odds of winning.
Summary:
No voting system is fair
Fairness can not be defined
There's NO perfect way possible
Well nothing is perfect in this world. If it were perfect, it's major flaw would be it's too perfect.
@@jugemujugemugokonosurikire4735 how would a perfectly fair voting system for everyone (if it would exist) be a flaw ?
@@jugemujugemugokonosurikire4735 I would like to know as well.
In any singular instance, you can point to a deficiency, but that is letting the perfect be the enemy of the good. What you have to look at is the system over numerous votes, and pick the best from there.
Something like sortion, for instance, has a greater chance for the majority to win any particular vote, but isn't always guaranteed. But statistically, the will win the majority of elections, with minority views also wining a proportion of the time.
Conclusion: Democracy is a mistake
While there is no perfect voting system, some systems (e.g. STAR and STV, both not mentioned here) are way more representative and closer to the preference of most people. Probably the worst is plurality (first-past-the-post).
CGP Grey has a good series about them.
STV is great for maximum representation in parlimantary elections while STAR is great for single winner elections.
Single Transferable Vote = Instant Runoff Voting covered at 1:11.
@@CTimmerman They are not quite the same. Instant Runoff Voting is for electing a single winner, whereas Single Transferable Vote is for electing multiple winners.
It's pretty clear that instant runoff has the best balance of fairness and simplicity. Perfect? No, that's impossible. But still excellent. I will be voting for it on the Massachusetts ballot this November and look forward to using it in future elections.
IRV is pretty good, but you might want to take a look at this page: en.wikipedia.org/wiki/Template:Comparison_of_Schulze_to_preferential_voting_systems
IRV fails a bunch of criteria that I would personally consider important. For starters, IRV isn't monotonic, meaning that in some cases, ranking a candidate higher can prevent them from getting elected, and ranking them lower can get them elected where they otherwise wouldn't have been. It's also not a Condorcet method. Condorcet isn't a voting system, it's also a criterion. A voting method can be called a Condorcet method if it elects the Condorcet winner if there it one. It's where there _isn't_ one that they diverge.
I believe Ranked Pairs is pretty good. It takes every possible matchup and ranks them by the percentage with which that matchup was won. Then you go down the list, "locking in" each candidate's relative ranking. When a result would cause things you become circular, you flip it. Since you start with the greatest victories, the people whose preferences are ignored (i.e. when a result is flipped) it ignores the least possible amount of people.
Still, it seems like you've thought about this a lot, and ultimately you should go with the system you think is best (for example IRV has later-no-help and later-no-harm which many others don't). After all, what's voting for if not choosing the best out of a set of people's preferences? I wish you the best of luck this autumn.
@@Gamesaucer In the system where I live ranked pairs sounds like horror to vote on. I feel like that works better when the number of pairs is still quite limited.
@@hkr667 What do you mean? You'd still just cast a single ballot. You'd just have to rank the candidates, and everything else would be extracted from that, same as other ranked voting methods. And in fact, since Ranked Pairs allows putting candidates on equal ranks, you can just, say, vote for the top 5 and leave the rest unranked. They'd collectively be your least favourite choice as a result.
What you could even do, if for example things are organised into parties, but each member of each party can be elected, is a system where you _rank_ the parties, but give a single _preference vote_ to your preferred candidate within that party.
@@hkr667 E.g. Say that there are 6 candidates (i.e. 15 pairs) and you rank them in this order: A > B > F > D > E > C. Since A is at the top of this list, this implies that not only (A > B) but also (A > F), (A > D), (A > E) and (A > C). Similarly, C is at the bottom so it's implied that (B > C), (F > C), and (D > C). The same logic applies to (B > D), (B > E), and (F > E).
@Gamesaucer @@Stratelier Yes, but in the video they said you could have a rock-paper-scissors situation. A > B, B > C, C > A. I can't really imagine a situation where it would happen though. But I can imagine ties - A > B, B = C.
"It doesn't matter who votes, it matters who counts the votes" - Stalin
I feel like Trump would have been a good friend with him lmao
It doesn't matter how many could vote, what matters are the one who will vote
@@ryanstarlight8018 lol Stalin and Trump would do great until they talk politics
Stalin was a sick man
um...Stalin was a tyrant....
Kim jong un : *nods knowledgeably*
Lmao🤣🤣
Loll
For those who are interested about the mathematical proof mentioned at 4:41, look up Kenneth Arrow’s impossibility theorem
Interesting video. Greetings from Spain. Lots of love for everyone. Have a good day!
Pehn yuva bondiya
You too ^^!
Love from somaliland too...wow this is really funny,how we communicate this easily
Sos tan agradable, gracias. Igualmente para vos ytyu familia.
Like in this map, the correct space port is the one with the shortest length from each base to the port, added together and then multiplied with their population as a percentage of the total population. Then the most amount of people travel the least amount of distance.
Somehow we need to define a map of the political spectrum, maybe using these voting systems with their different strengths and weaknesses in combination to define where people are located within the map.
A politician must then, before anything else, get a thorough check for his position on the map. This could help clarify where they stand on certain issues.
Huge questionnaires would have to be conducted, scaling what people think is most important. The size of the issue on the map could be its importance, the length and direction from the origin could be how much it matters for a specific party and which party the issue matters for. Most definitely doesn’t work in practice, but could be fun to test the idea
Just take the rankings from 1:22 and assign points for each vote. For example: 5 points for someone's first choice, 3 for their second, 1 for their third, and 0 for last. Doing this results in North winning, South finishing 2nd, West finishing 3rd, and East finishing last. Almost all reasonable point values (eg. 3 for 1st, 2 for 2nd, 1 for 3rd, 0 for 4th) end in the same result. I was only able to get a different winner once I made 1st place votes worth 4x the point total of 2nd place votes, which would obviously be absurd.
That system is called Borda Count, and its main failure is that it can be gamed through candidate cloning. For example, in a multi-ideology race, the one that runs the most candidates will likely win.
@@tomandnic77 number of candidates per party can be capped
3:56 called non-transitive preferences. Logically though, preferences should be transitive. So condorcet should work 95% of the time. In the 5% case, just use a different voting system to break the tie.
i was actually googling some of this stuff on my own the other day, always nice to see a ted video about it
I love the starting soundtrack of their videos . It's kind of a relaxing sound. I would like to put it on and go to sleep.
Anakin my allegiance is to democracy
I've seen convincing arguments that one of the best methods of determining the Condorcet winner in a single round of voting and best protection against favorite betrayal and other vote-splitting side effects, is STAR. Basically the optimal combination of IRV and approval/score. Not saying it's perfect (since there is no perfect), but since FPTP/plurality is basically the WORST, i'd say it's a vast improvement.
Because all voting systems will likely have a flaw, the question then is “which is the most fair way to pick someone or something?”
We can all agree that minority rule is unfair (which is when you win with less than 50%). But instant runoff is probably the antidote for this because based on preference there will always be someone that gets a majority of support. In US elections this is how you can help ensure that whoever is elected has a majority of support. ME already does it and it works.
The US system sucks
The best system is actually the Condorcet system because the winner win against each other candidates
@@maten146 too complicated, so for not engaged people it will be the same as plurality.
@@10gamer64 depends what you name as justice. US is like a union of 51 countries, so having 51 elections in each country and select 1 liked in the most countries is reasonable from this side.
@@dmitrizaslavski8480 Why would it be too complicated?
Not at all, and I d'ont see why it would be for not engaged people the same as plurality, they will be able to choose whomever they want
I literally read "which *vomitting* system is the best?"
Mount Rushmore of RUclips learning:
TED, CGP Grey, Crash Course & Vsauce.
Veritasium too😊
And kurzgesagt!
Philosophy Tube and Contrapoints.
@JoJo Yawson My guy
Let's not forget Extra Credits
1:14 maybe we can give points. The top priority will get 3 points then 2 then 1 and last will get a 0 . Candidate with highest point will win.
That might actually be the best. In our case it would result in 126 points for W, 194 for N, 107 for E and 173 for S, leading to North being selected as the winner, which seems like the best and fairist outcome.
That's the Borda count
That's called Borda, and its main weakness is that running multiple candidates increases your odds of winning.
Absolutely in love with these videos
I have I think a very good idea. You make a vote where people vote with 5 different systems, so that each criteria (at the end of the video) is filled only once. After that, you look at the city that has the most victories. If there is a tie, the most second, most third... if the tie is perfect, second round.
that is way too complicated for the normal person
That would take too long in practice. Maybe even years
Shame they didn't include Approval Voting- letting people vote for more than one option; winner is who has the most approval.
Agreed. Also, while Condorcet can "cycle," Approval approximates to Condorcet without cycling. And it is *much* easier for the voter than any ranking system!
@@WG55 And approval voting's failure of the Condorcet criterion is illusory because it is based on a slightly different value system. When necessary, it will, by design, select candidates who may have fewer true supporters, but who have a higher share of voters who find them at least acceptable.
I'd upgade that by giving one vote that has power factor of two and any number of secondary votes with power factor of 1.
Agreed, Approval voting and STAR voting are the best voting methods I've found. Approval does a really solid job at being fair/ accurate while being dead-simple, and STAR voting does a knock-out job while still being more simple than IRV. electionscience.github.io/vse-sim/VSEbasic/
@@darbyl3872 Yeah, that's one of the things I love about STAR voting. It's very flexible and the voter can use it however they want: Like an Approval ballot, like a ranked ballot, like a scoring ballot, even FPTP if they want.
You should add STAR Voting to your video. No one has a video on something big that is wrong with STAR. You did a nice job choosing key weaknesses in the systems you covered.
The two-party system exists in both the US and the UK and has thus almost a monopoly in the English-speaking world. But look no further than the EU and you'll find a great number of systems that we who are blessed to live in the EU consider standard: a multi-party democracy, where the parliament is proportional to the vote.
And then there's Hungary :/
@@Debre. based
proportional systems often result in hung parliaments and unstable coalitions...not great as far as I am concerned..
@@Debre. based hungary does it again
"Democracy is so overrated."
Frank Underwood
Erdogan supporter spotted
@@angrybirdo lol true, Islamism ruined Turkey
i agree, it only stays because of the social stigma of other methods.
but in reality it's very unefficient and not always fair
Alex it stays because it’s still the most efficient and fair of any other known system
"The age of consent is so overrated" Kevin Spacey.
Proportional Ranked-Choice Voting seems to be the best voting system yet imo
Instant runoff seems the best. The "bad result" in the video seems very constructed.
For Condorcet, if you do get a loop like that, there are ways of resolving it, either baked into the system, or using a secondary system to pick from that loop (known as the Smith set).
What you can also consider is what kind of tendencies a voting system has.
FPTP, especially if it is a regularly held vote with the same or similar candidates, such as typical elections, has a tendency to result in effectively a two-party system because it creates situations in which only two parties/candidates have a reasonable chance to win.
Instant Runoff has a tendency of the centre-squeeze, that is, centrists often receive fewer first place votes and are squeezed out, and split over the more extreme candidates, even though most would probably consider them a reasonable option (albeit not necessarily their favourite) and prefer it over the other extreme.
Condorcet has kind of the opposite effect, and is more likely to elect moderates who may not be many people's first choice, but who are deemed reasonable by most, and preferable by a (varying) majority to every other candidate.
Two round systems can kind of go either way depending on how people (tactically) vote: if a centrist and an extremist go to the second round, the centrist usually wins by getting the votes from the other end's extreme (unless there is something going on where people prefer any anti-establishment extremist over an establishment centrist), whilst if two extremists go to the 2nd round, then the centrists are squeezed out and forced to pick one of the extremes.
For picking something like a president, having a more centrist uniting factor may be preferable, and thus Condorcet would be preferable, but for picking local representatives for a national level (e.g. UK constituencies), it may be less ideal as it would cause mainly centrists to end up representing the population, rather than being more representative of the varying views in the country (but then again, electing local representatives for a national level is generally bad at being representative of the differing views that exist within society; you need something more proportional for that).
One of the best videos I've seen on this topic,good job TED-Ed
In the instant runoff scenario, if you add the votes up like points, awarding no points to the least favorable position and three points to the most favourable position, two points to the second most favorable, and 1 point to the third most favorable (per voter) the North base comes out on top.
The system youre thinking of is borda count
I love watching TED-Ed animated videos!
One could mix the instant runoff voting but count it the condorcet way. That way the problem a>b>c-Problem would be solved
This animation is super well made and so satisfying to watch. I love this channel!!
I’ve finally decided something, there isn’t any one voting system that can be applied for every situation, in this case approval voting makes the most sense (voting for all candidates you’re ok with winning), but in political elections with more then two candidates instant runoff is superior
Actually majority judgement is the best solution
Ted-Ed is coming out with all of these relevant videos! Keep it up!
Nice explanation of Arrow's impossibility theorem! In Kenneth Arrow's words, "Most systems are not going to work badly all of the time. All I proved is that all can work badly at times."
Also in Kenneth Arrow's words from a 2012 podcast: "Well, I’m a little inclined to think that score systems where you categorize in maybe three or four classes probably (in spite of what I said about manipulation) is probably the best."
just use all the systems at once then pick the candidate which wins the most times
Lots of money and time I guess
Idame Cantagile lol yeah people are lazy
I'm not sure if this system also exists, but a voting system that seems alright to me would be using the ranking that we saw for the "Instant runoff voting" system, but instead of counting the votes in the way that was shown, instead, each voter's first place vote would be worth 4 points, their second-place vote would be worth 3 points, third-place would be worth 2 points and last place would be worth 1. Then, we would add up all the points for each place. If this method is used, then East actually ends up with the least amount of points and North ends up with the most points.
STAR (score then automatic runoff) voting is essentially what you described mixed with Condorcet voting, and generally produces the least disliked result.
@@stirlingblackwood Oh, that's great to know! Thanks for informing me. Do you have the source that shows that it generally produces the least disliked result? I've love to read it.
Your videos are worth watching! Thank you Ted-ed for teaching me!
I like ranking and the weighting votes. So for this example, each colonist’s first choice would receive 4 points, second 3 points, third 2 points, and first 1 point. You add up each choice’s total points, then the highest is the winner. So:
West= (42*4) + (15*1) + (26*1) + (17*1) = 226
East= 207
North= 294
South= 273
So North would win, which makes sense because it is the most central to all the colonies.
It’s not a perfect system, but it seems better than the electoral college at least.
4:41 very interesting. Are there any sources online where I can see how they proved it? :)
Great topic with not too much info available online, so I am very happy it got posted :D I would love to see more about that and related stuff!
Lol I am surprised that my first attempt, searching for "prove that votes with more than two cannot be fair", immidiately got me to the wiki page about "Arrow's impossibility theorem" that states exactly the statement they mentioned in the video. I remember when I looked up things about this topic some years ago I did not find much; and now this is exactly the stuff I wanted to find! :D
Bro check out CGP grey channel he has Videos on his channel about voting system going into depth.... even how can system be systematically manipulated
Look up the videos about Arrow's theorem on the Infinite Series channel, they explain pretty well its proof (sometimes wikipedia can be way too deep and technical in math stuff)
STAR voting works well here, giving the Condorcet result while only asking voters to 5-star-rank the options. It also puts North against West in the Automatic Runoff round, quite reasonably consolidating the "eastern bloc" vote for the higher-population North.
Approval or STAR voting has been shown to be the most optimal system at maintaining the voters' preference and eliminate the need to vote against your favorite pick.
Wow, such an interesting video! Especially the presumptions make the whole thing so complex.
~Crewmate~
There are 0 *good voting systems* among us
One Of the Best Videos Ever!
Queen Lion is still contemplating each method…
For the Condorcet method, there's actually a good solution. You take the portion of votes N won against S, E, and W and get the geometric mean, the "volume" of support for N over all others. Do the same with each of the other candidates and you have a metric to rank them to see which is most preferred, without a circular outcome.
2:55 IM LOOKING AT YOU BRIAN KEMP
In Denmark we have a proportionalt election system which reflects the political distribution in the population fairly. If 40% of the population votes for party A then it will have 40% of the parliament's votes. The leader of the government becomes the one who can gather the most members of parliament behind him (which is not necessarily a majority of all members of parliament!). Currently, the parliament has 11 different parties.
This electoral system promotes compromise and cooperation between the parties since a single party rarely (never) has an absolute majority. Perfect? Hardly. But it reflects the views of the population. The underlying mathematical distribution model is a bit complicated, but as a voter you only have to deal with two things in practice, namely:
A) Will the party I am voting for obtain a minimum of 2% of the votes (if not, the party is not represented in parliament).
B) Which party / candidate will I vote for.
Instant runoff works pretty well because usually there are 2 viable options that get the majority of the votes, so every single person can vote both on who they most want, as well as who they prefer out of the 2 big options(mostly the 2 major parties)
But it falls apart whenever there are 3 or more viable options. That's why I prefer STAR voting.
@@anniekallen4472 How so and what is STAR voting? Single Transferable Vote / Instant Runoff Voting is simply a ranked list of approved parties.
Equal Vote Coalition has a RUclips video called "How Does STAR Voting Work?" which walks through the basics of STAR Voting. From the voter perspective, it feels very similar to ranked choice voting, but the tabulation does a better job at picking a winner that represents the will of the overall electorate.
Instant runoff guards well against the spoiler effect, e.g. a third party candidate coming in from the fringes and stealing votes from a more moderate candidate. They will typically get eliminated first, and their votes redistributed as if they had never run. However, it doesn't guard well at all against a centrist candidate being squeezed from either side. Centrists will also be quick to depart the race. In the end, you still have the same two-party system that Plurality incentivizes.
Best edu channel on RUclips 👑👑👑
I think STAR voting has solved the riddle
Star gets the most affected by tactical voting by far
@@oyunoynayanbiri that is up to the voter if they don't want their vote to go further
In this riddle it is by far the better method to use
Aproval voting (vote on one or more that you aprove) is the best. It would go like this:
W: 42
N: 26+42+15=83
S: 15+26+17=57
E: 17+15=32
North wins, wich is the fairest.
Raise your hand if you want to go to Mars! *No one does anything.*
Raise your hand if you want to go on the moon! *Still, everybody isn’t raising their hand.*
Raise your hand if you want to go to Jupiter! *Pure silence..*
_Okay, we’re sending everybody to the sun..._
I propose discussion. Why do you want the space port where you do? Because it benefits you? Well then, that's no better a reason than having it benefit someone else. If there truly is no benefit above that of convenience, then I propose it be equidistant between East and West. Yes, it benefits North and South far more, but neither East nor West - in this scenario - is favoured over the other; they needn't feel forgotten and outcast. What if either East or West had only a single citizen? Would matters change? No; not If you expect the vacant spaces to fill, in time. This scenario can be likened to that of our present; we all need to shop, and many of us travel to work; everything can't always be equidistant to everyone - but we can meet the matter with thought, discussion, and a empathetic care for others as well as ourselves.
Spoilers: it's not first past the post
As a Tasmanian Hare-Clark with the Robison Rotation is the best voting system. So multi member districts (7 members per district is ideal) using ranked choice (so rank the options then last place gets eliminated) and then for the ballot it's self rotate around each party's place on the ballot and also rotate each candidates position in their party's list of candidates. This was everyone has a say you still have a local member and internal party politics doesn't affect candidate placement.
RUclips: Video was posted 1 minute ago
Comments: Posted seven minutes ago
Me: ???
this format became unfunny in 2017
Man the sound effects in this one are simply gold
Approval voting. Simple, less gaming, best! Vote for all approved candidates.
FPTP was a British Colonial export, e.g. overlay list of all countries British colonized (past & present) vs FPTP countries (some have moved on like Canada & Australia).
Therefore FPTP are one of the oldest, not great for solving multiple choice question and usually end up having 2 major flavors close to the wire of 49:51%
Whole world : CORONA VIRUS ! CORONA! CORONA!
Ted ed : Which voting system is the best?
They already made a video about the virus.
From what I see from riots I can say US needs a new president. Which happens with voting.
One element of instant runoff voting not shown is that someone can simply refuse to cast a vote fo tor one of the options. For example, vote 1 for north, 2 for south, but refuse to put a 3 or 4 if they think east and west are completely against their interests.
Which voting system is the best
The one where leader elects himself
@Teringventje it was a joke
Participation is a really weird criteria when you think about it. For example, under instant-runoff, _the order in which candidates are eliminated matters._ E.g. if there are just 3 candidates and the initial vote tallies show A > B > C, the second choices from A and B voters are _never actually used._ And if the race between B and C is close, it actually gives tactical incentive for some C voters to NOT EVEN VOTE, because if enough C voters participate such that C > B, then B is the party whose second-choices (possibly for A) get counted.
I personally like one of the points methods, such as (points) = (1 / rank) e.g. 1st = +1, 2nd = +0.5, 3rd = +0.33, etc, and the highest final tally wins.
This is a very good point.
One observation I would make is that the ranked ballot methods all have the same flaw: they quickly become unmanageable the more candidates you have. e.g. if you have 20 candidates, it is a huge mental chore to keep track of which ones should be, say, 15th and 16th. A large field of candidates would be the norm if we didn't have to worry about spoilers and clones. Ranged ballot systems, e.g. STAR Voting, don't have this problem. You simply go down the list and assign each one a score within the allowed range (usually 0-5 or 0-10), following whatever criteria make sense to you, without the need to keep all your entries unique.
Wow, thank you for making this video! I enjoy your added perspective on the subject from that of CGPGrey's. Personally, this is the type of discourse I want to see on the subject of voting. What is voting?? And how can different social structure's accomplish it? And what is it's relationship to differing social structures, such as democracy and fascism?? Yes, fascist governments have held their own elections. Context matters!
This is far more helpful to our country than the bipartisanship harassment policy where you, the victim, get blamed for not giving them your consent by voting. Especially considering how much your get "punched in the face" for not voting, you must have a very developed and nuanced argument for why to continue to not vote. Which only serves to represent how broken the system truly is if a significant portion of the populace do not have sufficient faith in their government in order to vote in elections despite all the effort pushing them into doing so. It's important to remember that the citizens of a government are the victims of a failed state, To hold them personally responsible is disingenuous at best.
2:38 but would the South base think the same and vote for East? so the South base will have all the East vote, while the East base will have all the South vote, and thus both of them will still be removed?
also, how about assigning a numeric value to the rank of instant runoff (like 4 for 1st place, 3 for 2nd, 2 for 3rd, and 1 for 4th) multiply with the number of voters who choose that option, and then pick the highest result? As such, the North base will win with 294 votes, South base come second with 273, West base with 226 and East base 207 - seem fair to me.
SINGLE TRANSFERABLE VOTE -
I agree that or mmp
To be honest I prefer instant runoff voting and of course the first is the least two choices for the second because there is no way you can invent a system that going to make everybody comes of it a winner. But what you can do is invent a system that gives the candidate more than one chance with the people and that is instant runoff voting. for example there are 6 candidate and one of them has never been chosen as the first but he has been chosen as the second and the third by all the voters. This candidate will be the winner and with or without the voters knowledge they actually agreed on a middle ground.
Nine seconds ago?? Ok this is my chance....hi
Hi
Hi
hi
Hi
Hi