Hello everyone, I’m Watermelon (well, the ID is not this), and I’m here again, this time about the idea of ​​a community governance dapp, the same way, no funds and no team, just think about it, if you want to have a deeper chat, you can Private me.

(1) The origin of the idea

It has been three months since I entered the circle of blockchain without knowing it. In the past three months, as a novice, I have slowly grown into a rookie. In the process, I have come into contact with a lot of blockchain concepts and novel application ideas. I have been looking forward to decentralized facilities ever since I watched "Out of Control".

In the past three months, I have tried many applications, such as DeFi, chain games, tools, DAOs, meme ecology, NFT issuance in transactions, etc. Thanks to my smart brain, I have encountered many scams, but it’s nothing. Too much loss. In front of these blowout dapps, I even feel that time is not enough. There are several different applications in the same field, and the official website looks pretty good after a scan. Although I may only have one or two left between them until the end, this scene, which is quite similar to the battle of a hundred groups in the blue ocean, is really exciting.

But I always feel that something is missing.

In the past three months, from Xiaobai to a rookie, I can really feel the confusion in the blockchain field. Behind the anonymous address is a bunch of opportunities for wealth and freedom, as well as the monsters and ghosts who use the opportunity for wealth and freedom to make you zero. Many people can't keep up with the rapid development of the blockchain at all. It may be that which coin invested today made 100%, and was cut off by new tricks the next day, or fell in a pool of blood.

What is missing?

This afternoon, I was drinking coffee to study economic law knowledge (in the past month, my CPA exam has been basically cold because of the blockchain), and suddenly my brain flashed: What is lacking in the blockchain is not something similar to Douban, Rotten Tomatoes, etc., are the communities that rate each project?

Regarding the audit and objective data of blockchain projects, there have been many projects in progress, but the objective data and code review sometimes cannot directly reflect the experience of a dapp, the user's feelings and confidence in it, etc. Moreover, the data-only theory is completely wrong. If it is correct, the investment field will not flourish, just quantify it directly.

Moreover, Douban's mechanism for adding entries to communities and scoring community ratings is actually DAO!

After having this idea, I thought about some mechanisms and searched and found that there is really no team working in this field, so I was even more excited. Until later I found an item: the dapplist

Found that I thought about it and went together. On the one hand, I am happy, because there are people who are actually developing projects on such a track that few people think of and it looks good; on the other hand, I am also discouraged. Although I have no funds and no team, I think I have found a good opportunity... …

Okay, let’s talk about some of my specific designs. It is different from him, but not much different. It's a coincidence indeed, the idea is very similar ==

(2) Project pain points

First, clarify the needs. Will this be a demand pain point?

I think so, for the following reasons:

1. The blockchain field is still in the early stages of development. The existing DeFi and chain game metaverse concepts have broad landing prospects and have high development potential in the future. The public is still unfamiliar with blockchain, even in China, many people talk about blockchain as a scourge. Even with higher education, such as many of my classmates, top universities, masters and Ph.Ds, they still don’t recognize or just don’t want to spend time to understand the blockchain.

To understand the blockchain, it may not be enough to introduce the concept directly from the book. I think the way out lies in some out-of-land projects, such as the recent NFT, which may be a certain chain tour in the future. These projects may make more people want to understand. At this time, they will find that there are only various articles that make people picky. Each article is news or analysis of different depths. They will need to be similar to Wikipedia. Things to do a basic understanding.

Don’t forget, Wikipedia is also a community-driven project.

1. So why join the scoring system? As mentioned in 1, if he wants to join after he understands, he may face the dilemma you and I are facing at this stage, that is, he doesn't know what to do. In fact, what hinders the development of blockchain at this stage is the security concerns of many outsiders, but there is really no tool that can tell the outsiders how to use this thing, how to use it more securely, and which projects Good experience etc.

Imagine we want to learn about a movie, a book, what would we do? At least I will open Douban and take a look at the score. A high score may not be appropriate, but a low score is definitely not worth watching. This is my experience over the past few years. So in fact, the same is true for things here. A high score is not necessarily good, but a low score is definitely not worth using.

And under each comment, there will be some user experience, scoring reasons, etc., and new users will probably have a general impression of this dapp after reading it. From this perspective, it may even develop into the first stop of the blockchain, a tutorial that connects the encrypted world with the real world.

1. For residents of the crypto world, is this demand real?

Imagine the ways for veterans of the crypto world to acquire new projects: community groups, Twitter.

It can be said that these are the two latest sources of project news.

However, there are too many news from the community group and various types of news on Twitter. It is hard to say that news of a new project can be discovered accidentally. There are also some dapps, such as dappradar, from which new projects can be seen, but users can only see a small amount of objective data, and these objective data do not play any role in judging the quality of the project.

But if my vision becomes a reality, they will see new projects in the categorized areas of their concern, and there will be early user ratings and project information creators (I will call this type of user "founder" later, The discoverer is like the official website link and basic introduction added by Columbus to discover the new world.

So all of a sudden, he obtained a vague outline of the project. If this is what she is interested in, click on the link to explore it and do it in one go. At the same time, if the early dozens of reviews are all one-star, and the experience is poor, there is no need to waste time to get in.

In summary, I think there are demand pain points.

(3) Project logic

So next, what is the project logic?

First of all, it is important to make it clear that this is not a data-driven project, but a community-driven project.

Many people say that the data is objective, but they all overlooked one point. Whether the data is objective depends on whether the person using the data has the ability to interpret it objectively. If it is a user who sees the plunge and thinks it is a scam scam, the data does not actually have value to him.

Post a link, there is a familiar story about sushiswap : A brief history of sushiswap

The price of the currency is impermanent. If there is a community under the sushiswap project after the crash of sushi, you can post a post under the project to discuss your views on the founders' selling tokens. What do you think will happen? At least people who are not aware of it can learn more about the relevant information, so they don't have to think about the data table, how can a core project plummet?

So where is the community-driven value?

It lies in the flow of information.

I believe many people know that inside information of listed companies in the stock market must not be arbitrarily leaked, because any important inside information can make a roller coaster of the stock price of a listed company. Information can play a huge role, but in the world of blockchain, information seems to be too decentralized just like the blockchain itself. We can only look for clues from various WeChat groups and telegram groups. Especially the extremely unfriendly telegram group. The truly valuable information may be hidden in thousands of messages. Will you go through them one by one?

And community-driven means that the more likes a post is, the easier it is to be seen by users, and the information flow is higher.

The familiar Douban, Baidu Tieba, Station B, and NetEase Cloud Music Review Areas are actually community-driven products. The platform does not provide content. The platform is just a repository of content, and users are the ones who actually produce the content.

So, how do we judge the value of the information provided by users?

Because it is very likely that someone will tell lies, guide public opinion, etc.

As I will talk about later, through some effective incentive mechanisms to design a token model, user behavior can be appropriately guided. At the same time, we cannot ignore a basic principle, which is the law of large numbers.

When the number of users is large enough, you have to believe that the information that is collectively held in front of you must be the most valuable.

This is the power of decentralization.

In summary, the core of the project is to use community drive to present a mirror. In this mirror is the true reflection of all blockchain ecology to the heart of users. Each user is both a producer and a user of ideas. As a user, he obtained convenient and quick sources of information and obtained true reflections of other users' use of the project; as a producer, he provided his own real ideas and obtained tokens according to the token incentive mechanism, thereby realizing material incentives Of access.

Around this core, communities begin to be established, and more content will be absorbed, which will then become the foundation of the content culture of the entire blockchain ecology.

(4) Token mechanism design

Having said so much, the more I talked, the more I talked.

Let me talk about my thoughts on token design.

First, clarify the problems that may be encountered.

1. Like Douban and other websites, there are problems such as deliberately negative reviews, paying for reviews, and malicious control of reviews.
2. Users deliberately spread false news and guide and manipulate the market.
3. The token model is unrealistic, causing false incentives.
4. Balance the incentive relationship between early users and latecomers (at the same time it also involves the issue of how to expand the community)
5. Token circulation mechanism

Let's solve them one by one below.

1.1. Think about it, why are there deliberate negative reviews?

It's nothing more than unpleasant. Dissatisfaction comes from several sources, such as losing money, not getting satisfactory expectations, being treated unfairly, and so on. When solving this kind of problem, many people feel that it cannot be solved. In fact, this kind of problem is really tricky, because you can't control the ideas of others, and there will always be people who feel uncomfortable with tens of thousands of users on a project.

You can't control other people's thoughts, but you can guide their behavior.

For example, I set several factors, such as α and β. These factors will affect the user's ability to mine project tokens.

α can be set as: the degree to which user ratings run counter to mainstream ratings. For a project, it is normal for a person to feel different from most people. The model will not do anything, but will make a record of this difference. But if a user's evaluation of all items differs greatly from others, I am afraid it is not accidental. The model will synthesize these data to give alpha weights, which may be simplified to an exponential or power function relationship: a few large differences are not recorded, but if there are always large differences, the model identification may be arbitrarily evaluated, so reduce it. α.

β can be set as: the degree to which each rating is liked or disliked. Users will also write a paragraph of their own feelings when scoring, and the feelings will be recognized or disapproved by more users, which are recorded as β1 and β2 respectively, and some models are fitted to get β to reflect that the comment is under the item Community impact factor.

...

These indicators can be combined to form a weighting factor for user ratings. If the system determines that users are arbitrarily rated, it will reduce its factor according to the model, resulting in 1 his rating is useless and 2 mining power is reduced. This is the punishment of the community for the residents who affect the order.

This is a two-way incentive mechanism for users to leave real ideas. Think about it, what will happen to a user if he maliciously differs in a bad review?

First of all, malicious bad reviews will be detected by most community users, and then click on it. At least the beta factor will be very low. At the same time, because it is a malicious bad review, there is a big gap with mainstream reviews, and the alpha factor is also very low, which leads to 1 user with poor maliciousness. Comment on nothing to use 2 mining revenue decreased. A fool will make a malicious bad review!

So the question is, what if a group of people gave the same view?

In response to this question, what I want to say is that the community respects the choices of users. If there are enough bad reviews, there will indeed be an irreparable impact. 1 Influencing item ratings 2 Influencing a small number of users who left objective and true evaluations, the entire rating mechanism is disrupted.

To solve this problem, you can imitate Solana's idea and add a historical proof mechanism. That is: I believe that gold will always shine, and the anchoring of project ratings will be revised step by step over time, and all users who are approached by the historical trend line will receive the corresponding impact factor compensation.

For example, when sushi plummeted, many users would definitely rush to comment on their dissatisfaction with the founder of sushi and leave negative reviews. This is true from the time point at that time, but the person who gave a good review at that time may be judged by the model as a malicious bad review and reduce the weight. Sushi has slowly improved in a year, and the score has risen. People who gave bad reviews did not have to be punished because that was their truest thought at the time, and it was because of those things that actually happened. A group of people give bad reviews (otherwise there will not be a large group of people, just individual), so they should not be punished. Instead, compensation should be made to the innocent people who were implicated at the time. This is the historical proof mechanism, and the degree of compensation is determined according to the degree of historical trend line approximation.

Because history always chooses the right person.

1.2 Spend money to buy reviews. As long as community users have a common incentive goal, objective and true reviews can bring more benefits. Who will collect money to make inappropriate comments and affect their own mining factors? Moreover, decentralized buying, can investors afford it? There is no need to afford it. With so many anonymous addresses, it is better to buy a scientist to write a script.

As for the script, that's a matter of the code boss.

1.3 Malicious control and evaluation. It involves the maintenance of community order, which will be discussed later.

1. Users deliberately spread false news to manipulate the market. History has proved that the mechanism can be solved very well. With the introduction of an oracle machine, those who were judged to be false in the later stage directly cut the mining capacity by more than half. As for information mining, it’s up to you "policers" to talk about this role later.

Some people say that people's purpose is not to mine or manipulate currency prices?

This involves the scoring model. All of the aforementioned are user ratings, but the displayed ratings of an item also include model judgments. For example, the score that has not been judged by the model is displayed in yellow to remind the user that this score is still user data, and the score after the model judgement is blue, which reminds the user that this score has a higher credibility.

So how to make model judgments?

It still depends on community autonomy (because the model really can't judge you to manipulate currency prices). For the yellow star project, if there is behavior of manipulating the currency price, the model will automatically discharge the suspicious scoring addresses within the relevant time. The remaining addresses will be voted according to the information provided by the "policer". After the voting, a model judgment on the project will be formed. Rate the main page, and the stars start to turn red/blue at the same time. Red means this guy is a habitual offender.

Someone asked again, if there was no before, doesn't it mean that you won't manipulate the market in the future?

This is not our business. Because manipulating the market itself has nothing to do with the project, it is the team behind it, it may be the project party, or it may be a large company. The color of the score is used as a reminder, not a reference factor. The change in the color of the score is just a trivial support provided by the community. No one can predict the real trader.

Of course, everything will be reflected in the comments through the new ideas of the community.

1. The token mechanism causes false incentives. It is also necessary to use the power of the community to establish the identity of a "scientist" scientist, so that users in the community who have the ability to build models can help optimize the model, and at the same time, improve its incentives through some measures.

There are also "policer" and "founder" who also have special identities. The former are the main users who maintain the order of the community. They have many responsibilities. The exercise of their own responsibilities will also leave behind the data that can be reflected in the incentive mechanism through the supervision of community users. The latter is the adder of the new project, and the same is true. For example, the founder will get one hundred millionth of the mining revenue of the new project (very high, persistent, and can set half-year production reduction measures), but the information error will reduce the revenue; At the same time, after the community has formed a certain scale, it will vote to make certain changes to it.

The founder actually does the least, but this is actually an incentive for early users. At the same time, in order to squeeze the founders out, it can be set that 50% of the reward will be released after the score exceeds 100, and the rest will be released after 1,000. In this way, the founder does not simply work on the wiki, but becomes the propagandist and builder in the early stage of the project.

1. How to expand the community while balancing the incentives of users at all stages?

It can be seen from 3 that the early expansion task falls on the founders, after all, their motivation is the highest. But what if the founder creates a play item and leaves? It is possible to establish a competition mechanism for the identity of the founder. For example, any user can recommend the entry 100 days before the entry is set to become a co-founder, and finally determine the identity of the founder based on the number of effective users obtained by the competitor. Of course, the original founder will definitely enjoy a bonus of about 1.1. No. 1 after all.

The problem is again. There are few early users, the ratings will not reflect anything at all, and will mislead people. What should I do?

It's very simple. It's the principle of game theory. I keep the scores under cover, and do not publish score data for less than 1,000 scores. So no one knows how many points this item is. At the same time, in order to encourage the rapid breakthrough of thousands, a lottery is set up at the same time. The 1000th rated user will get a small part of the project's token mining revenue (early creation until the total revenue pool of 1,000 ratings), and the remaining part will be distributed to the rest in proportion 999, the sooner the more.

After 1000 scores, the basic scores can be seen, and the community is slowly getting up. Behind is the entry of more users, giving early ratings likes/steps to support the flow of information...

How to balance the incentives of early and late users?

It's very simple. The user's mining ability is never a constant value, but dynamically changes according to his community contribution value, such as the previous α factor, β factor, etc. Early user rewards have been given through early incentive distribution. For example, for an early user, the mining ability is 0 initially, 10 in the early stage, and 25 in the later period. The late users who contributed the same are only 15 because after all. Early users; but those with greater contributions, such as a comment that resonates with more people, may have a mining capacity of 30.

This setting has an advantage, because everyone is out of mind, and will not give you a thumbs up because you are a giant whale or one or the other to make you mine more. In addition, every comment user is completely anonymous, and no one knows who posted the comment except themselves, so as to avoid fan economy.

Here, everyone is equal.

totally decentralized

1. Token circulation.

All the questions boil down to the last basic question, how to distribute the tokens? How to release? How to dig? How much to dig?

First, should a deflationary model be used? Or unlimited supply?

The deflationary model needs to be adopted. Because no one will maintain a community because of some constantly depreciating tokens, even if they are willing, the material incentives must keep up. I never think that power generation with love can last forever.

However, the usual deflationary model will cause uneven distribution of incentives for users in the early and late stages. How to solve it?

Then use the high combustion mechanism.

What support is burning?

Since there are N projects, each project has a fixed total of m (in order to maintain fairness, whether it is uniswap or an unknown small project, it is the same m), so the total number of project tokens is Nm. At this time, because many projects have a small number of ratings, it is easier to dig for large projects, and it is easy to form false incentives, allowing more useless people to score randomly (anyway, scoring does not require money), so a regular release mechanism is set up, namely :

Reach 1000 scores to unlock 10%m, reach 10,000 scores to unlock 50%m, etc. Just an example, meaning that the unreached ones will never be unlocked, in fact it is an invisible reverse burning mechanism.

The problem is, people will still comment on those small projects. Because according to the existing mechanism, can't it be enough to rush over to unlock it? This is the best way to maximize benefits.

This involves the allocation of mining capacity. The mining power of a user is the total value. Every time he scores a project, the mining power will be divided equally among all the projects that have been rated. In this way, you may not as well not pay for the benefits of small projects.

So the question is again, no one has commented on the small project? Will affect the mining of my big projects.

So there are a series of settings such as founder. The incentive mechanism is to hope that through a series of balances and complex settings, people will forget about mining, focus on evaluating and playing in the community, and the big difference in mining is not bad. What really affects the income is the total amount, that is, the weighting factor, and this requires Your score is enough to be recognized, and the anchoring of material incentives returns to the content.

The distribution of tokens and how to release are also fun, and how to mine and how many are mixed in it. After designing this set, I found that, hey, it is actually applicable to all content-based communities. The blockchain version of Douban is not necessarily a dapp rating. Of course, because dapp scores are more demanding and ecological, it will be easier to build a complete ecology in the later stage, so...

Okay, I'm done typing, my eyes hurt, I went to take a shower at my dreamer