How to measure the degree of decentralization of a blockchain?

The concept of blockchain network decentralization has generated, generates, and will surely continue to generate much controversy in the crypto world. It has frequently been used as a weapon between proponents and detractors of various projects, often without solid arguments or reliable data.

But beyond futile debates, decentralization is a very relevant aspect of a blockchain network. Let's try to shed some light on the matter.

Ethereum's main proponent and one of the leading figures in Blockchain technology, Vitalik Buterin, articulated years ago what we now know as the Blockchain Trilemma (or Scalability Trilemma), where he cites the 3 main attributes of a network:

Security, Scalability, and Decentralization. The trilemma highlights the challenge developers face in the arduous task of designing or implementing improvements in these types of networks: they must prioritize two attributes at the expense of one of the three.

Setting aside scalability and security (which would warrant two more articles), decentralization was one of the main reasons Satoshi Nakamoto created Bitcoin: to design a financial ecosystem that would allow person-to-person payments, without a central entity governing it, without intermediaries, without censorship, accessible to everyone from anywhere in the world… and all of this in an environment of anonymity and, therefore, low trust.

The concept of decentralization refers to the network's governance model, which is not vested in an institution or a small group of people, but rather in the users themselves, empowering them and avoiding risks of monopolies or oligopolies. It democratizes the system. But it's not black and white; there's a wide range of gray.

Therefore, the question is: How do we measure the degree of decentralization of a blockchain?

What types of blockchain networks exist?

In practice, no completely decentralized network exists, nor is it easy to quantify its degree of decentralization. It is worth remembering at this point that there are three types of networks classified according to the dispersion of governance among their users:

Centralized, where one person or a small group of people have control over it (current financial system).

Decentralized: where control of the network is distributed with varying degrees of success among a large number of its users who democratically govern it (Bitcoin, Ethereum…).

Distributed: or 100% decentralized, only described in theory as it is a utopian concept, as we will see during the development of this article. We will mention it later (some sources also call it “disseminated”).

Escaping centralization and achieving the highest possible degree of decentralization is one of Bitcoin's main reasons for existence (if not the main one).

A network that would escape the control of governments and corporations, with its governance falling to its users (people's empowerment, if we want to call it that), who would operate anonymously, and where all events would be recorded transparently and immutably, visible to all.

However: the fact that a blockchain network is centralized does not, in all cases, imply that it is worse than a decentralized one. There are occasions when a centralized network offers certain advantages. Often, it is advisable to analyze what problem we want to solve and what type of network is best suited, but this is not the topic at hand.

But, what is the purpose of measuring decentralization in a network?

Quantitatively understanding the degree of decentralization first allows us to determine the margin for improvement for future network updates. Secondly, and no less importantly, we can be aware of a project's decentralization scope to consider it when making decisions about participation, investment, developing our proposal, etc.

How can we measure the extent of a network's decentralization?

Before addressing this question, it's worth noting that there is no perfect method or absolute truth. What we can do is try to get closer to an objective result, using certain parameters for which we can find information by researching publicly available sources. And if we don't find this information about the project, we can start to suspect that it lacks the optimal degree of decentralization.

1. Decentralization of miners (and validators)

Blockchain networks are inherently secure due to the encryption they are built upon, but they are not immune to malicious attacks aiming to seize control for self-gain. In fact, it can sometimes be easier than it seems. At this point, the line between the attributes of security and decentralization may become blurred.

Depending on the network's consensus protocol type, there are different roles within it, including those responsible for verifying the authenticity of transactions and confirming the latest block (or ledger) so that the network can progress smoothly.

They are a key component, and as can be deduced, this grants them significant power over the network. We must ensure that taking control is extremely difficult.

In the case of Bitcoin's Proof of Work, miners play an important role, potentially gaining control if they acquired, for example through association, enough computational power (hashrate) to dominate the network.

Based on this premise, it's reasonable to assume that the greater the number of miners and their geographical dispersion, the more difficult it will be for them to conspire and associate with each other to try and take control of the network.

We can also ask ourselves other questions about this parameter:

Is there freedom to be a miner or validator? Does the network pose any obstacles? Is very high technical knowledge required? Is a high hardware investment needed? Do you need to own network tokens to mine/validate? Are there pools that concentrate high computational power, thereby centralizing power?

If we can only adopt this role with someone's permission or under certain exclusive circumstances, we are, in the long term, favoring positions of relative power to the detriment of decentralization.

Another aspect to highlight, as we've mentioned, is the geographical distribution of miners/validators, and let's look at an example:

Let's imagine a network with a million users and tens of thousands of miners/validators. Prima facie, a network with a good miner ratio, but what would happen if they all turned out to be operating from the Community of Madrid and the regional government decided to ban mining? Or if an event of any kind occurred that caused a power outage or any other accident?

We not only need to see if the ratio of miners/validators to users is high, but also assess their global distribution to minimize legislative, political, environmental, and other risks.

Let's consider another real example: we are just weeks away from one of the most anticipated events in the crypto world, where Ethereum will make its definitive transition from Proof of Work to Proof of Stake (the Merge).

When this becomes a reality, anyone who wants to be an Ethereum network validator will face no obstacles, except that they must own at least 32 ETH (which, at the time of writing this article, is a not insignificant 48,000 euros).

In our decentralization classification, how would this fact score? Does it make it a more or less decentralized network?

2. Decentralization of nodes

We won't elaborate much on this section, as the reasoning is very similar to the previous case. Nodes play a very important role in expanding the network, enabling its adoption, and managing transactions within it.

The more nodes and the more distributed they are, the better. However, this comes at the expense of transaction speed.

3. Developer Decentralization

It's clear that technology is advancing very quickly, sometimes faster than society itself. This requires networks to be constantly updated. One of the undeniable strengths of a blockchain project is having a solid community of developers with the right skills and technical knowledge.

The vast majority of networks have web repositories and forums where documentation can be stored, code can be viewed, and users can stay up-to-date on new proposals, discussions, etc.

Any modification or improvement to a network requires very high technical knowledge, the publication of papers, arduous technical discussions, and months or even years of testing on the testnet before seeing the light of day.

We apply the same criteria as in the previous two cases: having a large community of properly trained developers distributed globally increases the network's decentralization.

If, however, the community is small, we run the risk that the network won't be updated or that proposals will come from the same direction, from a group of like-minded individuals, leading to a loss of creativity and depth in the proposals.

4. Exchange Decentralization.

While this aspect may not be among the most important ones mentioned here, we should still consider it.

If we want to exchange fiat money for tokens or vice versa, we must use an exchange. A key aspect of decentralization implies that any user of the network can access it at any time without issues.

If we're thinking about Bitcoin, advanced users can interact directly with the network, although they will always initially need the services of an exchange. Users with less experience or lower digital literacy (the vast majority) will need an exchange platform to interact with the network.

It comes back to the same point: for a project to enjoy good decentralization, it must be listed on reliable, solvent exchanges in countries across the globe that are up-to-date with relevant regulations, ensuring liquidity and 24/7 operational capability.

5. Supply Decentralization:

Supply refers to the total amount of cryptocurrencies or tokens issued and expected to be issued by the network itself. In the case of cryptocurrencies, it's obvious why it's important to know how they are distributed.

If a single entity holds a substantial amount, they could control its price at will. Furthermore, tokens can also grant us voting power within the network, the ability to be a validator, etc.

When a significant percentage of the total tokens are held by the network's creators, this grants them considerable power, and we can assert that the network loses points on the decentralization scale.

For this reason, it's crucial to thoroughly research the project, whether by reading its whitepaper, Roadmap, current holdings of major wallets, etc.

Conclusion

Considering the ideal scenario (in terms of decentralization), we would find at least that:

All its users would be miners (validators) and nodes, we would have the availability and money to buy and maintain the necessary hardware, we would be equitably distributed across the globe, we would have a vote to participate in decision-making, we would have the time and technical knowledge to propose, validate, test, and develop improvements, the supply would be "homogeneously" distributed among us, the network would be accessible 24/7 from anywhere in the world, and we could exchange tokens for fiat money at any time without legal or liquidity issues.

As you can imagine, fulfilling all these conditions in practice is impossible. Therefore, when we refer to a decentralized network, we must bear in mind that, in one way or another, there is always some degree of centralization within it. It is our responsibility to carry out the necessary due diligence to gather the required information.

Now that you've made it this far, you probably have new questions. In addition to the parameters explained here: What other aspects of a network can be considered when calculating the degree of decentralization?

In this regard, and as recommended reading, I encourage you to read the article by Balaji S. Srnivasan, who, as early as 2017, explored the minimum Nakamoto coefficient as a quantitative measure, including the client (software) as another metric to consider.

By Guillermo Subirá.

I want to know more!