Staking is so hot, but inflation accelerates centralization

Inflation has a direct impact on the security of any blockchain, the distribution of tokens, and the design of incentives. When we talked specifically about EOS, we believe that the expression of the inflation debate is too simple. Inflation can lead to good results, bad outcomes, or little impact, depending on where it flows and how it is used.

The purpose of this article is to examine the problem of EOS inflation to confirm that EOS inflation without direct verifiers (consensus participants) may have significant benefits, and that inflation given only to these participants may have significant negative effects. We hope that this discussion will help promote community talk about this issue and would like to receive feedback from voters and discuss it further.

1. The centralization effect of cryptocurrency

Characteristics of cryptocurrencies: Micro-economies with different attributes that are easy to analyze and use mathematics, are usually predefined and irreversible. For example, Bitcoin will only issue 21 million before it stops inflation. This makes the analysis very unique – we can analyze these closed economic systems accurately and very accurately, and look at attributes such as currency circulation speed and distribution as a whole in the case of all information disclosure. In this article, we will attempt to analyze the changes in the centralization characteristics.

There are many contributing factors to the cryptocurrency system that can generate power concentration – except for what we usually think of, such as wealth accumulation. Taking Bitcoin as an example, we have observed that miners focus on selecting large mining pools. In PoS (and DPoS), we see a similar phenomenon; decentralization can be limited by these capable participants due to a small number of potential consensus participants with comparable technical capabilities. The debate surrounding Bitcoin has been ongoing.

Proponents of the “community block” believe that the large block will always give the new miners too much data burden, and the “big block” supporters believe that technology is improving, and this burden will decrease with time and technological progress. In fact, the right balance point may be somewhere in between.

The investment in the start-up phase and the accumulation of long-term inputs will undoubtedly increase the centralization of Bitcoin, as new miners cannot afford to verify the system and reduce the number of potential consensus. However, it should be careful to find a suitable balance between the technical cost of not burdening the new miners and the available systems that will not drive away the new miners.

In the PoW system, current rewards do not affect future rewards, as the ability to implement PoW is an external factor. This is why the PoW system is “de-licensed” – their consensus mechanism is not a closed system and requires external factors (in the form of electricity and 'work').

For DPoS systems, their reward structure is quite different from PoW. In many DPoS systems, the supply of tokens is not fixed, but inflation. This means that current rewards actually affect future rewards. If we use the incentives that generate more influence on the consensus to give participants who are responsible for the consensus, then the degree of decentralization of the system will be affected in the long run.

2. Inflation reward

In order to understand this problem, we need to take a step back and see more clearly how the attributes of inflation work in a closed system. For example, suppose there are two people, Alice and Bob, each of whom has a currency of 100 units. If Alice and Bob each get 100 more units, then now everyone has 200 units of currency. Is there any change? In a closed system, we didn't do anything – Alice originally had 100 of the 200 units (50%) and now has 200 of the 400 units (the percentage is still 50%).

Except for the number of units of account, the same level of inflation did not change. In terms of the “% of money supply” reference, inflation has completely changed the percentage. If we give Alice 100, but not Bob, then Bob still has only his original 100, but Alice has 200. We used "inflation" to dilute Bob's value – his holdings changed from 1/2 to 1/3 now, and Alice's share changed from 1/2 to 2/3.

What needs to be noted here is that we must understand inflation as it changes the proportion of total supply. If inflation makes one profit, there will be another party that is damaged.

So how does this actually affect the blockchain system? In DPoS, the control of consensus is determined by the number of votes. Simply put, those who hold the most will actually control the consensus. However, what we reward for those who are consensus participants is to use inflation to dilute the part of the rewards of those who do not control the consensus.

Imagine that Alice, Bob, and Charlie both have a unit of currency with a total supply of three. If they elect Alice and she increases the supply by one and rewards herself, she now has two votes (compared to Bob and Charlie). It is impossible for Alice to vote for himself now to be considered invalid. When the consensus is controlled in a closed system, ultimately, the control of the current system rapidly develops into centralized control in the future.

We can call this the form of "asymptotic convergence", in which the supply that is concentrated by a participant (as part of the entire issued quantity) increases asymptotically in a closed system (ie , it gradually gets closer and closer to the entire issued volume). Although we only use inflation as an example, this effect can occur in the case of fixed or variable fees and variable inflation, but the calculations will change slightly.

3. Consensus progressive concentration

When the supply of contemporary coins is important to consensus, and inflation is given to those who are currently leading the consensus, then inflation over time may lead to the centralization of the system. This is the Asymptotic Consensus Convergence.

Of course, real-world systems are much more complicated than the simple examples we have. Consensus rewards are not always all picked up. To enhance understanding, let's use physics to make an analogy. Suppose we have a box in the middle of the warehouse. If the box is continuously applied in a single direction, the box will move in that direction. If this force is offset, the box will not move. In this analogy, the progressive concentration of consensus is that fixed forces are imposed on the currently selected consensus participants. This force can be offset, but this needs to be done voluntarily. If consensus participants always sell all their rewards, then this force will be completely offset.

However, as long as there is a consensus participant to receive the reward, the box will move a little bit. The more this practice, the more the box moves. This strength can be defined based on rewards:

For example, if we assume a 1% annual reward (achieved with inflation), then this will produce the following curve:

(You can plot this curve by typing "(1.01^y -1)/(1.01^y) for y=0 to 100" at https://www.wolframalpha.com/input/)

From the chart we see that the percentage of control over the system (that is, the control over the new supply compared to the new supply) has changed from zero to less than 10% in 10 years, 100 years later About 60%. This means that at this time, the maximum rate at which the box can move in our analogy is that number in the worst case (only through this force).

There is a very important reason why we use force to make an analogy: offsetting this thrust is only a direct result of delay, and will not prevent future effects – in the example, consider a person who is constantly kicking a box with a kick: Doing so may temporarily stop them, but this does not stop them from constantly moving the boxes.

Although this doesn't sound like today's problems, these numbers should be worrying. In EOS, the number of votes to the head node is currently only about 12% of the total supply (note: when translating this article, the tokens voted for the top 10 nodes accounted for 17% of the total supply). However, voting in EOS has a number 30 that multiplies: if a group of 30 people agrees, they can collectively use this number to increase their votes (by voting).

In this case, a group can gain control over a period of up to five months, as long as it gets a different supply by voting for a different group (and hopes that the new group will voluntarily and unselfishly oppose it). Concentration of rights can consolidate your position. (Note: Each account can vote for 30 nodes)

In fact, when considering the multiple competing collectives of a single dominant control that competes for consensus, the numbers become very complex. However, through some efforts, it can be proved that the group with the greatest rights (especially the group that receives the most votes from the current dominant group) will eventually win. For those who are not in the largest collective of rights, this is caught in a prisoner's dilemma – if you choose to join the group, you may get a "reward." If you refuse, and they can unite with others, then you can't get anything.

4, let power concentrate

This is confirmed by the fact that if you are concerned that participants in the concentration of power have stabilized their position due to inflation. However, we should not only consider how this problem exists, but also how to solve it.

There are some extreme measures that can be taken. For example, the inflation incentives for consensus participants are completely eliminated. This does solve the problem, but it does not motivate the node to work for the network from the start. Another attempt is to separate the attributes of the token: voting tokens (supply is fixed) and utility tokens (supply is inflation).

In theory, the rewards of consensus participants can only be paid using utility tokens, so they do not accumulate more voting rights over time. However, this does not actually work! Because as long as they can exchange these tokens in the market, this method can be invalidated.

However, the complete elimination of voting rights is not the only solution. In fact, if there is a second source of inflation to join, an equal but opposite force, then concentration of rights cannot be achieved (or even eventually decentralized)!

In the above equation, the reward is the amount of supply that the consensus participant receives relative to the total supply. If we increase the total supply in another way but don't give consensus participants, then their control over the system will not increase. Take a look at the previous examples of Alice, Bob, and Charlie: Suppose everyone has a unit of currency, and Alice gets an extra one as before.

But for now, suppose Charlie also gets an extra unit, so that Alice's extra voting rights are offset. For other purposes, this will use additional forms of inflation. Like all inflation, there are always winners and losers: in this case, Bob's voting rights are “diluted” and become losers – but perhaps he is the least involved or not involved in the system.

However, finding a good reason to introduce a second source to offset inflation is more difficult. In the initial inflation problem, if we only consider the percentage of ownership rather than the arbitrary number of units, then inflation is only a tool to sacrifice one side to make the other side favorable to adjust each other. Therefore, we need to choose a winner (select new rewards for those who are not part of the consensus participants) and new losers. You might think of an obvious solution – we can pick those who are not BP to be winners! But this will be a precise offset, which is equivalent to not giving BP a reward at the beginning.

If you have been following this logic, then you are fully aware of what kind of solution will be followed: Worker Proposal. In fact, the solution to this problem is actually the second source of inflation, which is led by the holders to non-block participants. This may include funding applications, services, grants and scholarships, as well as other costs that may arise to increase the value of the entire network.

It is important to strongly encourage block producers not to double the use of these funds, because if they do, then this approach has no practical significance. With this second source of inflation, the power of centralization will be offset, and in fact it may become a force to promote decentralization. If inflation to the community is greater than inflation that only gives consensus participants, then although consensus participants may still be rewarded for their efforts, their status is no longer the most attractive factor in obtaining returns.

The worker proposal system can work in a variety of ways. There is a very simple way to copy the structure of the block producer, but only pay the wages by weight and not because of the number of blocks. This second group will be selected from the non-BP, not involved in the control of the consensus, and can get their own inflation – determined by the holder of the currency vote.

This is great for organizations that are technically weak or not good at governance, and then they can spend time on what they are good at; whether it's community outreach, marketing, scholarships or other EOS business projects. In fact, through this structure (or similar structure), consensus participants are more responsible for community decision-making because more voting rights are transferred to the community over time.

5. Summary

In this article, we highlight three key points to consider when using inflation in a closed consensus system such as DPOS.

1. Inflation itself is just a tool for generating winners and losers based on the size of ownership of total supply.

2. If we give those who have more influence on the consensus to those who are responsible for the consensus, then we will eventually give those who currently have power more power.

3. It is important to prevent the consolidation of power. One of the best ways to avoid this is to take appropriate constraints – ensuring that a single party or individual group is not the only recipient of the award.

As the EOS community continues to discuss the pros and cons of various configurations of EOS inflation, it is important to remember that changes to it may have long-term and second-order effects. Although directly reducing the amount of inflation seems to be beneficial to the value of the token, doing so should ensure that the long-term health of the network is achieved. We believe that short-term operations can create changes in market prices, but long-term development can drive real value creation. To achieve this, we must seriously consider how changes to important parameters can affect the centralization of the network.

-END-

Translator's profile: Chuan , the author of the Blockchain Learning Society.

Disclaimer: This article is the author's independent point of view, does not represent the position of the blockchain study society, and does not constitute any investment advice or advice.

Original link: https://blog.greymass.com/eos/@greymass/inflation-centralization-and-dpos