What voting system should the AssangeDAO adopt? Simple token voting, quadratic voting, weighted voting

This is a follow on to Zylo’s 11 speed proposals.

The voting system currently used for the initial setup is “Basic voting”, which is a single choice voting with three choices: for, against, or abstain. It implements a basic token based quorum voting, pursuant to which a minimum of 3,470,307,901 $Justice tokens equivalent to 20% of the $Justice token supply are necessary for a proposal to pass. Once this threshold has been met, whichever decision has more votes wins. If a proposal does not reach the quorum threshold, a proposal fails. This system was chosen by the core team as a temporary measure until a more thought through voting system is agreed.

There may be more suitable methods for a governance system. Snapshot offers several voting systems (see more details on the voting systems):

  1. Basic voting
    Single choice voting with three choices: for, against and abstain

  2. Single choice voting
    Each voter may select only one choice.

  3. Approval voting:
    Each voter may select any number of choices.

  4. Quadratic voting:
    Each voter may spread voting power across any number of choices. Results are calculated quadratically.

  5. Ranked choice voting:
    Each voter may select and rank any number of choices. Results are calculated by instant-runoff counting method.

  6. Weight voting:
    Each voter may spread voting power across any number of choices.

The current simple token based voting system is plutocratic: it gives persons with the most tokens the most votes. Plutocracy is a paradigm considered widely inferior to democracy (Plutocracy - Wikipedia). The current voting system also provides an attack vector to the DAO through vote buying, whereby the DAO may be taken over by a group or person unaligned with the mission by a party borrowing a sufficient number of $Justice tokens for the simple act of voting.

I propose to adopt a Quadratic Voting system - at least for matters that are of great importance. Quadratic voting is a mathematically elegant way to reduce the influence of whales. It thus allows for more democratic decision making, by favouring the voting power of the “poor and many” over the “rich and many”. Quadratic Voting reflects the intensity of people’s preferences in collective decisions. It greatly mitigates tyranny-of-the-majority and factional control problems.

Accordingly, my suggestion would be the following:

For any action to be implemented by the AssangeDAO, a AIP must be prepared and must pass through the following process, as amended from time to time and publicized on the AssangeDAO Forum:

Draft Proposal:

  1. Creation of a draft proposal using the AIP Template and its discussion in the AssangeDAO Forum, with the discussion open for a minimum of 7 days. Where a proposal requires any additional on-chain interaction by AssangeDAO, the required code must be provided prior to the draft proposal moving into Active Proposal stage.
  2. A simple poll is conducted on the forum to take a temperature check as to whether the Proposal is likely going to pass.
  3. Where a proposal is vetoed by the Assange family unit, it will not be processed. [This aspect was already decided by a previous AIP.]

Active Proposal:
Where a proposal meets the requirements of the Draft Proposal stage, a Moderator (known as “Author” on Snapshot - to be separately voted on) will assign an AIP number and move the proposal into active voting on Snapshot.

An AIP is accepted if, within the Voting Period:

  • the [quadratic voting] result in favour of the AIP is higher than the quadratic voting result against; and
  • The quorum threshold is 3,470,307,901 $Justice tokens (equivalent to 20% of the token supply).
6 Likes

Thank you for a clear and simple explanation of the different voting methods.

I agree 100%.

isn’t it just possible, to instead spread out the tokens among many many wallets, in which case the quadratic system wouldn’t work anyways?

1 Like

can we use ML techniques for this?

I googled an article about Quadratic Voting: What Is Quadratic Voting?. Quadratic Voting is a method of… | by Shaan Ray | Towards Data Science

I agree that this method of voting makes the “poor” voices louder.

The blockchain is traceable, and any token transfer operation can be checked. And a lot of bot voting can be monitored. It is also recommended to have a set of anti-cheat mechanisms to deal with related issues.

谢谢您的以上解释,但是我有一个关键问题请教:

请问,您研究过 Quadratic Voting的数学模型吗,这是一个二次方方程。

因为每个项目的token数量级是不同的,对投票的结果影响不同。如果一个项目的token数量特别地大,使用Quadratic Voting,作恶的成本就会变得非常小。而我们的token总量是170亿个。

总结,如果提案里建议使用Quadratic Voting,那么请详细地列出Quadratic Voting的合理数学参数和相关的仿真结果。

如果只是单单列出一个Quadratic Voting的概念,对于促进投票公正性是没有实质性帮助的,是没有太大意义的。

Thank you for the above explanation, but I have a key question for you:

Please, have you studied the mathematical model of Quadratic Voting, which is a quadratic equation?

Because each project has a different order of magnitude of token, which affects the outcome of the vote differently. If a project has a particularly large number of tokens, the cost of doing evil becomes very small with Quadratic Voting. And total number of our tokens is 17 billion.

To summarize, if Quadratic Voting is proposed in the proposal, then please list in detail the reasonable mathematical parameters of Quadratic Voting and the related simulation results.

It does not make much sense to just list a Quadratic Voting concept alone, which is not substantially helpful to promote fairness in voting.

3 Likes

The premise of quadratic voting to ensure true fairness is: kyc
At present, the DID infrastructure in the blockchain industry is not perfect. Therefore, the introduction of quadratic voting is actually a disguised way to guide people to participate in voting in a decentralized and multi-account way, which will lead to unforeseen large numbers of fraudulent behaviors.
二次方投票要确保真正公正公平的前提是:kyc
当前,区块链行业的DID基础设施并不完善,因此,贸然引入二次方投票,其实是变相引导人们去采用分散、多账户方式参与投票,会导致不可预见的大量做假行为。

1 Like

The Snapshot proposal site should be a digital model feature that does not support secondary voting and is limited to the functionality of the tool site. Assangedao will need to develop its own Snapshot proposal site first, which will cost a lot of money. Also in the crypto world: the power to hold coins is more important. It is not the power of the poor that you say, otherwise it is unfair to the large holders of money itself! They are also not motivated to have more justice, and the purpose of Assangedao is to encourage users to have more justice

Yes, this is a problem in systems without KYC and would require monitoring of how persons try to game the system, f we care enough as SshW suggested. I wonder however whether persons would make the effort to split their holdings into very small portions to defraud/ game the system.

While I am a mathematician I am not the most technical person: I felt the voting system module on snapshot would allow us to do quadratic voting. Could you expand on why you think that would not be sufficient?

Could you please expand your point? What would you consider as “doing evil”?

“Doing evil” means spreading tokens to many wallets, each requiring very few coins, in such a way that the same tokens can get a very large number of votes.

I have calculated that in Quadratic Voting way:

Scenario A: If an address has 100 coins (worth 0.07u), then it will get about 14 votes.

Scenario B: If an address has 10 million coins (worth 7000u), then you get about 4600 votes.

If I divide 10 million coins which worth $7,000 into 100,000 wallets, then with 100 coins per wallet, I get 1.4 million votes, which is 304 times more than the number of votes I get in Scenario B.

Because the total amount of jusitc is 17 billion, which is extremely large, the number of 10 million is only 0.058% of the total and can not be considered as a whale. $7,000 is also what a lot of people can afford. When it is compared with a more extreme case, which is to divide 8 coins per address, then using 7000u you can get 5 million votes, so the cost of doing evil is very low.

The Gas fee is also a cost that is not accounted for, as it can be achieved through tools that can transferr tokens to many addresses at once.

For our project, the problem with using Quadratic Voting is that the number of coins is too large.

I sincerely hope you can understand my explanation, thank you.

作恶的意思是把tokens分散到很多钱包,每个钱包只需要很少的币,以这样的方式,同样的tokens就可以获得非常多的票数。

我计算过,在Quadratic Voting的方式下,

情况A: 如果一个地址100个币(价值0.07u),那么就能得到大约14张票。

情况B: 如果一个地址1000万个币(价值7000u),那么就能得到大约4600张票。

如果我把价值7000美元的1000万个币分到10万个钱包,那么每个钱包100个币,我就可以获得140万张票,是情况B得到票数的304倍。

因为jusitc的总量为170亿,这是极其大的数字,1000万个的数量只占了总量的0.058%,并不能认为是鲸鱼。7000美元也是一个很多人能够承担。如果比较更极端的方式,就是每个地址分8个币,那么使用7000u就可以得到500万选票,所以说作恶的成本非常低。

Gas费也是成本之一,没有计算在内,因为可以通过工具可以实现一次传输到很多个地址。

对于我们这个项目而言,使用Quadratic Voting的问题在于币的数量太庞大了。

我真心希望您能理解我的解释,谢谢。

7 Likes

I wonder why you have this question, coz you should know who could be the potential enemy of this DAO,who has money and methods.

我想知道你为什么问这个问题,你应该知道这个DAO的潜在的敌人是谁, 这个敌人有钱有办法。

2 Likes

Is there a better way to reduce whale impact?

What you describe is a risk, the cost of implementing it is not only the gas fee, but also the operation. It is difficult for a single user to use a large-scale wallet to vote manually, and it can almost only be done by machine, but machine voting can be found through traffic monitoring. In addition, the diversion from one wallet to countless wallets is recorded on the blockchain. By this method, cheating can be found.

In any case, this is a big improvement over direct voting. Quadratic voting gives a more equal voice to the rich and the poor, and it costs more for the rich to have the same voice as a direct vote (cheating will be detected instead). Quadratic voting does not completely solve those problems, but it does alleviate many of the drawbacks of direct voting.
Quadratic voting is more democratic than direct voting and does not create new problems.

你所说的是一种风险,但是要实施它的成本除了gas fee,还有就是在操作上。单一用户使用大规模钱包投票很难手动操作完成,几乎只能用机器完成,但是通过流量监控可以发现机器投票。另外从一个钱包分流到无数个钱包在区块链上都是有记录可查,通过这种办法作弊是能被查到。

无论如何,这比起直接投票都是一个大的进步。直接投票比的是谁股份多,Quadratic voting让富人和穷人的话语权更平均了,富人想要和直接投票一样的话语权需要更多的成本了(作弊反而会被检测到)。Quadratic voting虽然不能完全解决那些问题,但是要缓解很多直接投票的弊端。

Quadratic voting比直接投票更民主,而且不会产生新的问题。

1 Like

What if the rich doesn’t care about the gas fee since they’re rich?
With Quadratic voting, They can lead the voting result completely deviate from the will of the community.

3 Likes

This question is the comparison of Quadratic voting and direct voting, and the problems that arise in Quadratic voting are bound to arise in direct voting. Direct voting is very obvious to “lead the voting result completely deviate from the will of the community”. But Quadratic voting alleviates that.

It must be clear that “lead the voting result completely deviate from the will of the community” is not a problem of Quadratic voting, but a problem of the current mechanism of how much equity is used to vote.

We need to make things better and fairer with limited options.

The 10million votes from 10 accounts and 10million votes from 1000 accounts are the same. The quadratic doesn’t change the nature of things, it’s just giving the bad guys a chance to ruin our DAO.

3 Likes