Explaining Friend.tech's Economic Model: Game Theory, Expected Value, and Demand Curves

1. How to make a good Ponzi social networking product

Friend.tech’s economic model looks very simple: (1) the price of the key increases with volume (2) each transaction charges a 10% processing fee, which is shared equally between the agreement and the key issuer (3) points will be distributed to users over the next 6 months

The best way to understand the economic model is to assume the role of the project party, [If I were to design the economic model, how should I do it?] The starting point is that we want to make a SocialFi product, but past experience and current market liquidity make it difficult for us to be optimistic about this, so we hope to make a product with certain Ponzi attributes to complete a cold start.

(1) What does (S^2) /16000 mean

Ponzi’s core is to make money for early entrants. If you imagine all users queuing up to enter one by one, consider that S can only take integers, so you should use differences and sums. As can be seen, DeltaP/DeltaS increases linearly, which ensures that as the number of keys increases, the price of the key will increase, and the rate at which the key price will also increase (that is, the faster it rises).

Obviously, this is a very indirect and efficient Ponzi curve. Every entrant will push up the price, and the driving magnitude will get higher and higher. As for 16000, it is also very understandable. We need a parameter so that S and P have a matching relationship in line with the market. As shown in the table below, if the value is smaller, the P curve will be too steep and the price will fluctuate too much; if the value is larger, the price curve will be flatter, not enough for Ponzi, and 16000 is a compromise choice. The small volume carrying capacity is also in line with the current market liquidity situation.

(2) How does the economy circulate

Optimists will understand Friend.tech as a social platform, and pessimists will understand Friend.tech as a gambling platform. But what these two understandings have in common is that they all have three roles: 1) FT platform 2) key issuer 3) user. The only act that generates profit is the user’s transaction (a precondition for simultaneous use or holding of a transaction).

The question then becomes: How do you entice users to buy? According to the understanding of social platforms, the issuer of Key is a service provider (no matter what this service is), and the platform provides basic services; according to the understanding of gambling platforms, the issuer of Key is a giant horse, responsible for recruiting users.

This division model is also very simple and efficient. A 50% share is equivalent to purchasing KOL services. We can see that quite a few KOLs have accepted this. The reason why it is necessary to adopt the Ponzi model is that the cold start problem was also solved here. In the early days, the services provided by Key issuers were bound to be uneven and unstable. At this time, speculative demand can play a role as an alternative to a certain extent.

(3) Point airdrop

There isn’t much to talk about with points. Their role is to further stimulate demand and confuse users’ speculative needs, usage needs, and investment needs.

2. How much is the wear and tear of the transaction?

Objectively speaking, Friend.tech’s economic model and narrative are very beautiful, but after experiencing it, I still decided not to run my Room because it was a very draining negative sum game.

First, let’s ask a question: How much does Key’s transaction cost? 10% is clearly a wrong answer.

Let’s imagine a situation where you enter this market with 1.1E, and since you also need to pay a 10% handling fee for the purchase, you can only buy 1 key worth 1E. At this time, your Rooms Value is 1E. However, no matter when you sell, you will need to pay an additional 10% processing fee, so the real realizable value of your position = 0.9E. From the moment you buy, the 10 %手续费就已经无法避免，只不过Friend.tech会延期向你收取。事实上从你买入起已经亏损1-（0.9/1.1)=19.2% you sell needs to increase by 22% to pay back.

** 19.2% is not difficult to calculate, but unfortunately this is Friend.tech’s second biggest eyesore method. **

To understand this, we need to first clarify the relationship between “book value (BV)” and “expected value (EV)”. Let’s imagine that all key buyers are speculators (we will discuss other types of users later):

(1) Zhang San, Li Si, and Wang Wu raised funds to buy a cow, a duck, and an egg. Make an agreement that whoever quits first takes the cow, the second one takes the duck, and the last person takes the eggs.

(2) Zhang San/Li Si/Wang Wu all felt that they had a cow’s claim, but in fact their claims were all equal, and there were 6 types of results. The three of them took 6 different results: 1) cow, duck, duck, egg, 2) cow, egg, duck, 3) egg, cow, egg, duck, 3) egg, cow, egg, 6) duck, egg, cow

(3) Since the probability of the six types is equal, what Zhang San actually has = 2/6 cows +2/6 eggs +2/6 ducks is 1/3 cow, 1/3 duck, 1/3 egg.

Through this case, we can see that although Zhang San, Li Si, and Wang Wu all feel they have the right to claim from a cow, there is only one cow, so this is just an illusion. The value of the rights they actually have should be equal to the mathematical expectation (EV) of their claims. If another new player, Zhao Liu, joins this game at this time, he needs to provide a house. When he joins, the house is divided into four parts, and Zhang San, Li, and the Four Kings get 1/4 of each.

• As can be seen, every new entrant will have their EV diluted by previous owners. This is the core of Friend.tech: * (1) Confusing EV with book value to create an illusion of wealth (2) After using the participants’ EVs, the previous users provided profit.

Friend.tech’s trading model is that the pot pool is the only counterparty, so the capital available for trading is TVL in the pot pool, which will make a difference. For example, when the number of keys is 40, the price of the key is 0.1E, and the total market value = 400.1E = 4E, and at this time TVL = Sigma P = 1.38E. Understanding this, we can draw a curve of the relationship between book value (BV) and EV (yellow line chart in the chart).

**As can be seen, when the number of keys exceeds about 20, the EV/BV basically stabilizes at 30 %附近，无限接近30%. **There are two pieces of information implied here:

(1) If you buy in the part where the curve is leveling off, in addition to paying a forward processing fee of 10 %手续费+10%, you will immediately lose about 70% of your EV.

(2) The room value shown by FT is too optimistic. Based on the principle of prudence, it would be more scientific to measure the value of the Key you hold using Room Value* ~ 30% (EV).

** This also explains why everyone seems to have made at least 2-3 times their book return over the past period. **

3. Where is the end of growth?

Next, consider the question of “going back to the original”. Here we still assume that all users joined the game for profit. If we use book value as the evaluation criteria, it is not difficult to pay back. If we buy immediately at 5E, it only takes 27 new buyers to pay back the money.

** However, from the perspective of EVs, it is almost impossible to buy a high-priced key to return money. If you buy a 1E key, 115 new buyers will also be required to achieve EV return. **As protocol data and number of users grow, we will naturally consider book value as a return factor, but once growth stops or declines, this measurement becomes very unreliable.

At the same time, regardless of whether it is calculated based on book value or EV, there is the same problem. The higher the purchase price, the number of new buyers required to pay back will also increase. However, after all, there is a limit to growth. If the growth limit is N, then buyers after the N-M number will not be able to pay back their money. In this way, rational players will not buy after N-M, and since this information can be obtained by everyone, no one in the N-M to N range buys, and rational players will not buy after N-M-L. In this cycle, the final equilibrium price will continue to move down.

In fact, this situation is one of the most classic cases in game theory — “The 2/3 Game,” so if you don’t think this process is easy to understand, you can also directly read the 2/3 game related instructions, or “Block K [Beauty Vote]” from the second season of the Japanese drama “Intruders in Today’s Country.” :)

** To put it more directly, after the net inflow of capital slows down, high-value keys first become unprofitable, and speculators will go after lower-priced keys. The cycle continues, and the top price of individual keys (especially new keys) will continue to move downward. Under normal circumstances, this downward shift would not be a big problem, but another problem is that Friend.tech bots are very rampant. Bots will monopolize Key’s low price range in the new market, so after hitting a new equilibrium price drop, they will directly enter the bot’s arbitrage range, and users’ EVs will be further encroached upon. **

4. Is (3, 3) really reliable?

**The next question to discuss is (3, 3) is it reliable? The answer is unreliable. **A few reasons:

(1) (3, 3) is non-reciprocal most of the time. For example, if you buy a 3E key and your own key price is 0.1E, your buying behavior will contribute 0.15E to the other party’s handling fee, while the other party will only contribute 0.005E to you. (2) The (3, 3) model involving multiple people is extremely unstable. If only 2 people reach the same price, then (3, 3) is stable. It’s a bit like exchanging protons with each other during the Warring States period. If you kill my protons, I’ll also kill your protons. But as soon as the number of people increases, (3, 3) becomes very unstable.

This is another classic model in game theory — the evolutionary game model. The derivation of evolutionary game models is complicated and tedious. Simply put, when there are enough people, there will always be people who try to steal the profit because it is profitable. If A’s rush causes B to suffer losses, then B’s motivation to run and lock in profits/avoid losses will increase, and C, D, and E will also doubt each other. After all, EVs are far lower than BV, and the only Nash equilibrium after the chain of suspicion is formed is (-3, -3).

It should be noted that in the past, it seemed very stable over time (3, 3), but this is only because during the upward cycle, it is easy for people to ignore EV exploitation, and the “-3 trend” is very low, and -3 will become more frequent after growth stops or a downward trend appears) The above only targets multiple people between strangers (3, 3). If you are friends in real life or reach an agreement (3, 3), this kind of (3, 3) will stabilize a lot, because choosing the -3 strategy also requires additional reputational damage.

5. Will it be profitable to use points?

First, let me explain that currently circulating credit estimates are all based on estimated FDV. When formulating one’s own strategy, real EV = return estimated according to FDV * probability of actual airdrop* (1-wear rate) (e.g. linear unlocking required, price falling short of expectations, etc.)

Judging from my own experience and that of other friends, the current points have 2 characteristics: 1) In the end, most users’ points are only related to position value, and there is a snapshot point in time before points are issued, and only the position value at that point in time is taken. 2) As mentioned earlier, the book value of KEY is about 3 times TVL, so when you calculate the total amount of capital invested, you need to calculate TVL*3 as the credit base for all users.

After understanding all of Friend.tech’s mechanisms, if you still want to pay points, the best strategy is to buy and hold your own small key at once. This practice can prevent EVs from being exploited and reduce processing fees by 5%. However, it is important to note that the least wear method is if you buy your own small size in the current position and sell it 6 months later, so your total opportunity cost = total investment capital *0.905, or loss of 9.5% of the principal amount. But you’d better not have any more deals for the next 6 months to avoid additional wear and tear.

6. Where is the future of FT

All of the above discussions are based on the assumption that all participants are speculators, but this is not the case. Many group owners have already begun to provide differentiated services through Room, and these real “services” are the key for Friend.tech to get rid of Ponzi.

Still, using the case of Zhang San, Li Si, and Wang Wu who raised funds to buy a cow, a duck, and an egg, it is explained that if the situation changes: Zhang San promises that he will finally quit the game, then Li Si and Wang Wu’s EVs will change, from 1/3 cow +1/3 duck +1/3 egg to 1/2 cow +1/2 duck, and EV will increase significantly; if Li Si also promises to finally quit, then Wang Wu’s EV will become a complete cow.

**The core of this change is that utility consumers will change the “homogenization of claims” situation, thereby increasing the EV of remaining participants. **In actual Friend.tech, it appears in two categories:

1) Issuer-owned, binding 33 and passive holders (such as ETFs)

2) Users who need to use and hold keys, such as those who want to connect with Key issuers through Room, obtain Alpha information through Room, enjoy Real World benefits, and enjoy the redistribution of potential airdrops.

** Key’s equity will determine utility value, and will also determine the stability of Key chips, making it an inferior claim; speculative demand will only bring homogenized priority claims, and will be more affected and more unstable by price fluctuations. What is certain is that Key will clearly diverge in the future, and it will be difficult to maintain 33 and the pure hype Key. **

7. High pump+Bot is killing this game

After considering utility requirements, Friend.tech’s business model has the potential to break out of Ponzi, but I still decided last week to sell all of my keys and stop running my Room because the FT official’s high margins and bots are killing the game.

On the one hand, there is a four-fold difference between the (5 %+5%) *2 = 20 %的手续费，即使是另一个我们属性的高摩擦市场Opensea也仅收取单边2.5% royalties and 2.5% handling fees charged by Friend.tech. According to the data, Friend.tech’s TVL is currently about 36 million US dollars, and the processing fee has reached an astonishing 24 million US dollars; of this, the handling fee charged by the agreement has reached 12 million US dollars.

According to our previous algorithm, 36 million US dollars corresponds to a total market value of about 110 million US dollars. It is not an exaggeration to use this calculation. However, even if net user withdrawals and bot net withdrawals are not taken into account, the least friction situation is 48 million US dollars entering this market. After less than 2 months of trading, 12 million US dollars of this amount already belongs to Friend.tech, or 25%.

At the same time, when these keys are sold, they will be charged another 10 %，这部分事实上已经产生只是会被递延收取。此外，按照1.1亿美元的Key总市值计算，这些Key如果每天换手率达到5%, so Friend.tech will withdraw 5, 110 million US dollars, 10 %=1650万美元，约占TVL的45% every month. %*30% Everyone’s net deposit will continue to flow far and wide to Friend.tech.

The statement that “high processing fees are charged to encourage Hold” also seems untenable at present. Encouraging Hold does not require 10% tax on buyers, and judging from recent updates (adding a web version, adding a watch list) and credit rules (room value preconditions are buying), Friend.tech doesn’t seem to have substantially encouraged HOLD’s behavior. After all, who can refuse to keep getting real deal revenue?

However, one last point that needs to be clarified is that Friend.tech’s product design, economic model, and operation strategy are all excellent and worth learning from. Social is also one of the definitive directions of Web3. If Friend.tech can reduce pumping to a relatively reasonable level (or continue to use most of it for building rather than buying a luxury home) and solve the flood of bots, I think I’ll be one of its most loyal users.

Additionally. My own Key has no value, and Room won’t operate. If you want to connect with me or have questions to discuss, my Twitter DM is open to everyone (and you don’t need to have a key). If you think my content is valuable to you, welcome to the NFTs of my Mirror (https://mirror.xyz/lokiz.eth) Mint articles (including this one); they are limited (maybe useful some day in the future) and inexpensive (0.001-0.01E, or free).

Disclaimer：

1. This article is reprinted from [mirror]. All copyrights belong to the original author [Loki’s Blockchain Notes]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.