Will large-block Bitcoin lead to centralization?

Chapter 0 Introduction

The debate on capacity expansion has basically reached a consensus on 2M blocks, but two completely different routes have emerged after 2M. Classic advocates using unlimited to dynamically adjust the block size, and this route advocates the use of large blocks. Core advocates the use of small blocks supplemented by the Lightning Network route, and this route advocates the use of small blocks.

The main support for the Core route is that small blocks are conducive to decentralization, while the storage and bandwidth costs brought by large blocks will reduce the number of people running full nodes, leading to centralization.

This article wants to prove that it is not the large blocks, but the small blocks that threaten the decentralization of Bitcoin.

Chapter 1 The biggest driver of centralization is that no one cares

The decentralization of Bitcoin is mainly reflected in three aspects. The first is the decentralization of complete nodes. This is from the perspective of Bitcoin users. The second is the decentralization of computing power. This is from the perspective of Bitcoin mining. The third is the decentralization of development. This is from the perspective of Bitcoin development.

The main reason for discussing the impact of block size on decentralization is that the larger the block, the higher the storage cost of the full node and the bandwidth cost of transmitting data. The higher the cost, the less motivation users will have to run full nodes, which will threaten the decentralization of full nodes. The second reason is that mining pools will go bankrupt due to the high storage and bandwidth costs, especially the orphan block rate caused by large blocks, which will lead to the centralization of computing power. For developers, large and small blocks have little impact.

This is the main reasoning process of the argument that large blocks affect decentralization, namely cost determinism, which holds that the higher the cost, the greater the threat to decentralization. This reasoning has serious flaws, which we will explain in two cases.

From the perspective of users running full nodes, because users are not motivated by profit to run full nodes. On the surface, if the cost is low, users will make a balance between hobbies and expenses. The lower the cost, the more beneficial it is for users to run full nodes. But in real life, users are willing to spend money for their hobbies, and most businesses are also trying their best to find ways to arouse users' interest. Only when they cannot attract users' attention from their interests will businesses use low-price strategies. This is especially true in the current era of abundant materials.

A very simple example is that Lamborghini is so expensive that it cannot threaten its presence around the world. We cannot say that the cost of consuming Lamborghini determines that its consumers will be centralized.

No matter how expensive a sports car is, it will be sold all over the world because it can be used to show off. A domestic car without a brand effect, such as QQ, has a very low purchase cost, but the consumer concentration of QQ is much higher than that of a sports car. We have not seen QQ sold all over the world yet.

From the user's perspective, running a full node requires costs but does not bring any benefits. In order to achieve decentralization, it is more important to arouse user interest rather than reduce its costs.

Litecoin is the third largest digital currency by market capitalization. The centralization of its full node wallet is much more serious than that of Bitcoin, but the storage cost and data transmission bandwidth cost of Litecoin full nodes are much smaller than those of Bitcoin.

Some of Bitcoin users are merchants, such as wallet operators, exchanges, and large companies that accept Bitcoin. In order to provide Bitcoin users with a better payment experience and prevent potential double-spending, they are motivated to run full node wallets as long as Bitcoin has enough users, even if the cost is high. On the contrary, they have no motivation to run a digital currency wallet used by a niche user, such as Litecoin.

The most serious period of Bitcoin centralization was when Satoshi Nakamoto first launched Bitcoin. The computing power of the entire network was 100% centralized, and the coins were 100% centralized. Because at that time, no one cared about Bitcoin except Satoshi Nakamoto. Even at that time, the storage cost and bandwidth cost of running a complete Bitcoin wallet were lower than running QQ.

In order to fully achieve decentralization, we should focus more on how to make more people like Bitcoin instead of worrying about storage costs and bandwidth costs.

Chapter 2: Large blocks are more beneficial to mining pools

Now let’s talk about whether large blocks will lead to centralization from the perspective of mining pools. Theoretically, the larger the blocks, the higher the storage and bandwidth costs, and the higher the orphan block rate, making it more difficult for small mining pools to survive, which will lead to centralization.

First, we rule out the possibility that the increase in the orphan block rate leads to centralization. This can be resolved technically, such as thin blocks, block header first, and other technologies. Secondly, the mining pool can adjust the block size and will not produce blocks that exceed the network's carrying capacity. For example, under a 1M block, because China's export bandwidth is worse than that of Europe and the United States, Chinese mining pools will mine empty blocks and limit blocks to 741KB. Once the bandwidth limitation is resolved, whether through priority transmission or purchasing better broadband, Chinese mining pools will eliminate empty blocks and 741KB.

The larger the block, the more transaction fees can be included, which is good for the mining pool. There is no doubt that as long as the network carrying capacity is large enough, the mining pool is willing to accept larger blocks. There is absolutely no reason to think that larger blocks will harm the mining pool.

But there is another view that small blocks are conducive to increasing transaction fees, which will bring higher profits to mining pools. This view is very mainstream, and many elites in the community hold this view. It is necessary to analyze it carefully.

The price of a commodity is determined by market supply and demand, not by the producer, let alone the production cost. Bitcoin transfer transactions are also a commodity, which is equivalent to users hiring miners and purchasing their bookkeeping services. Users pay a handling fee as compensation. Assuming that a block is limited to a maximum of 7 transactions per second, the bookkeeping service has a supply upper limit. If there is more user demand, when the transfer is congested, that is, when the block is congested, the handling fee will increase as some users want faster bookkeeping services. If the block is not congested, that is, the average waiting time for all users is 10 minutes, there is no benefit in paying more fees, and users have no motivation to pay higher fees.

Some people may say that if we directly increase the default fee of the network, for example, from the current 0.05mbtc/kb, can we increase the overall fee without congestion? No, the default fee for Bitcoin transfers is not mandatory. Even if you increase it to 1BTC/kb, users can still construct a 0.05mbtc/kb fee transfer. Moreover, it is impossible for all wallets on the network to send Bitcoin according to this default value. For example, exchanges and mining pools have to consume a lot of fees themselves. They will definitely try to pay as little as possible on the premise of meeting the waiting time of about 10 minutes.

In addition, there is competition between mining pools. As long as the blocks are not isolated, miners tend to pack more transactions to get more fees. In this case, if you want to maintain the transfer fee of the entire network at a certain high level, it is equivalent to a price alliance. The alliance requires all mining pools to default not to pack transfers below the default fee. Price alliances have never been sustainable. Even in China, there are powerful central agencies such as the Price Administration Bureau and the National Development and Reform Commission, which cannot force a unified price for a certain commodity. For example, the price of gasoline. The National Development and Reform Commission said that it should maintain a high price. But in reality, some gas stations have started to recharge gas cards for 100 yuan and get 5 yuan free. Some gas stations will offer car washes for gas. The apparent price remains unchanged, but the actual price has dropped. Mining pools are the same in competing for fees. The current mining pool income mainly comes from block rewards, and the competition for fees is not obvious. When the block reward is halved, there will be competition between mining pools to grab accounting services. When mining pools compete with each other, it is more difficult for fees to rise.

Therefore, relying on small blocks to increase the handling fee is equivalent to launching a DDOS attack on the Bitcoin network, filling up the blocks, forcing users to increase the handling fee and obtain faster accounting services. That is equivalent to drinking poison to quench thirst. If the blocks are more congested, users will inevitably leave Bitcoin. In the case of block congestion, merchants will leave, because they accept Bitcoin payments and find that there are a large number of accounts waiting to be received. At this time, the risk of accepting Bitcoin payments is very high. They will refuse to accept Bitcoin. A large number of dark web users will also have to leave, because what they need is timely payment. Once some users leave, the block congestion will be relieved, and users can use less handling fees to purchase accounting services. Then comes the next cycle, the block is congested again due to the large number of users, and the handling fee has to be increased. Users will give up Bitcoin again due to congestion, and the handling fee will therefore decrease.

It is impossible for miners to continue to receive high transfer fees by only providing a very poor accounting service. On the contrary, large blocks can accommodate more transactions, and the mining pool will have higher profits.

We also need to analyze the benefits of mining pools from another perspective, which is the price of Bitcoin. Undoubtedly, the higher the price of Bitcoin, the higher the benefits of mining pools. However, the factors that affect the price increase are too complicated, and we cannot infer the price increase or decrease from the block size and the number of transactions contained in a single block. So let’s not discuss this for now.

In the long run, small blocks are not conducive to the interests of mining pools, while large blocks are more conducive to collecting more handling fees. Naturally, the former will eliminate more small mining pools, while the latter is conducive to the decentralization of mining pools.

There are still many voices in the community who hope that the Lightning Network can solve the problem of limited payment quantity. At the same time, they believe that the Lightning Network can deposit transaction fees into the hands of mining pools. Well, let's infer whether the Lightning Network can help decentralization.

Chapter 3 Lightning Network with Small Blocks Will Be a Nightmare

At present, the Lightning Network has not been realized, it is just a concept product. So we don’t know its specific operation, but from the white paper we can know that the Lightning Network is indeed a very good concept, even comparable to the Bitcoin white paper.

(First of all, since the Lightning Network has not yet been implemented, I have only read its white paper, and it is the Chinese version. So the following understanding may be wrong.)

The Lightning Network is essentially an offchain wallet with 100% reserve. The most likely scenario is that it needs a centralized company to operate it. A company uses this protocol to develop a wallet that settles with the main chain. Users can use the company's products to make lightning payments. Another possibility is that it is similar to a decentralized wallet, like Bitcoin, where users who use this wallet can open lightning payment channels. The result of the evolution of the Lightning Network may be both.

Regardless of the former or the latter, considering that the Lightning Network is more efficient when operated by a company, the company can launch marketing, provide customer service, and even subsidize users to use it, so the future evolution is more inclined to company operation. For example, the current CEO of Haobtc, Xingkong, has repeatedly claimed that he will launch a Lightning Network service.

The advantages of the Lightning Network are very clear in the white paper. If it is really implemented, it can solve a large number of old user transaction problems. That is, the Lightning Network will accumulate a large number of user transactions. These transactions will share the pressure on the main chain, making the main chain blocks much smaller. In addition, with the advantages of companies operating the Lightning Network, some large companies will be formed in the market, and these large companies will operate a large amount of Lightning Network transaction data. This means that the operation of the Lightning Network will be centralized.

Once a big company operating the Lightning Network encounters trouble, such as being hacked by the government, the Bitcoin network will face a peak in transaction transfers. The Bitcoin that was originally deposited in the Lightning Network of this company will be sent to the main chain in a very short time. This is equivalent to a DDOS attack on the Bitcoin network. Given that the Lightning Network itself can technically accept a huge amount of transactions, if a company operating the Lightning Network really goes bankrupt, the main chain of the small block is likely to face a DDOS that far exceeds its carrying capacity.

In fact, this situation is just like the collapse of an offchain wallet merchant in the existing Bitcoin network. If an offchain wallet merchant collapses and faces liquidation, the main chain will face a large number of transactions in a short period of time. Now the median size of Bitcoin blocks has reached 0.8M. If a large offchain wallet merchant collapses, the block will face congestion.

Therefore, if the main chain is not strong enough and the capacity to carry transactions is not large enough, any fluctuations in the lightning network and offchain wallets on the main chain will pose a strong threat to the main chain. If any lightning network or offchain wallet with a large transaction volume faces a run in a short period of time, the small block main chain may not be able to bear it.

How the Lightning Network guarantees the miners’ income is also a problem. The white paper does not explain it clearly at present. But it is vaguely explained that the Lightning Network can deposit transaction fees to the miners on the main chain, just like how many transactions are recorded on the Lightning Network, and the transaction fees will be settled with the miners when the Lightning channel is closed. I don’t know if this is feasible. If it is feasible, it will add a large amount of transaction data to the main chain to record transactions within the Lightning Network. This is different from the offchain wallet. The main chain does not need to care about the content transfer data of the offchain wallet. From this perspective, it is even more unfeasible to combine small blocks with the Lightning Network.

In addition, the handling fee of the Lightning Network still needs to be adjusted by market supply and demand, but according to the white paper, the Lightning Network can expand the transaction volume by thousands of times, which means that the handling fee should be very cheap. Whether this economic calculation can bring benefits to miners on the main chain needs to be verified. If the computing power on the main chain cannot get enough mining fees, after multiple halvings, the computing power will have to be reduced. The security of the Bitcoin blockchain will face challenges.

Therefore, even if the Lightning Network is implemented, we still need large blocks to deal with the potential risk of the Lightning Network being squeezed. If the small-block Bitcoin leaves the Lightning Network and the sidechain system in the future, or if they have problems, the main chain will have difficulty surviving. Large blocks do not have this risk.

Chapter 4 Small blocks coupled with Lightning Network and sidechain coins will lead to mainchain centralization

Let's assume that the Bitcoin ecosystem has evolved to this point, with small blocks on the main chain and very high transfer fees. The Lightning Network and sidechain coins have taken away a large number of user transactions. Under this premise, can Bitcoin remain decentralized?

It is conceivable that there will be fewer ordinary consumer users running full nodes. This is because users are paying more attention to the Lightning Network and sidechain coins. Whether for profit or for showing off, the Bitcoin full node itself is not enough to attract ordinary users to run it.

Merchants only need to run a lightning network or a sidechain coin wallet, and there is no need for them to run a wallet with expensive transaction fees. This will further reduce the number of running Bitcoin full nodes.

In this case, only mining pools have interests and must run full nodes. Both Bitcoin computing power and full nodes will face serious centralization.

If it is a large block, the future is much brighter. The transaction fees of large blocks can still keep the transaction attractive to new and old users. As long as there is attention, running a Bitcoin full node can be "sentimental", and there will definitely be enough people motivated to run a Bitcoin full node, which is conducive to decentralization.

Moreover, the low transaction fees and sufficient user attention keep it attractive to merchants, which is naturally conducive to the decentralization of the complete node.

Chapter 5 Summary

1. The obstacle to decentralization is not cost, and marginal increases in storage and bandwidth costs are not enough to threaten Bitcoin's decentralization.

2. The biggest obstacle to decentralization comes from the lack of attention. When users' attention is attracted to the Lightning Network and side chains, the Bitcoin main chain becomes a database that only a small number of merchants and mining pools pay attention to, which brings about terrible centralization.

3. Large blocks are more beneficial to the interests of miners. As long as the blocks are not so large that the orphan block rate increases, miners will not worry about the marginal increase in storage and bandwidth costs. Instead, they will be more concerned about whether they can package one more transaction and charge more fees.

4. The operation of the Lightning Network will bring centralization, which is not a problem in itself. But the problem is that when the centralized operator of the Lightning Network goes bankrupt, it will deal a fatal blow to the small block Bitcoin. We need a large block Bitcoin main chain to bear this risk.

5. Rather than worrying about rising storage and bandwidth costs threatening decentralization, we should be more worried that the Bitcoin main chain will become something no one cares about, and that running a full node will become something without sentiment.

Thanks for reading.

If you find the article useful, please give me some Bitcoin to encourage me to continue writing.

Author: tan90d (Weibo @LightningHSL WeChat tan90d)
My BTC address: 14mhzjkJ71oMAMkKu3dy98dnUpkyQBHL1r
Source: Babbitt Information