Decentralization Of A Network - How To Measure And Evaluate

avatar of @geekgirl
12 min read

We hear words like decentralized, blockchain, and web3 a lot in crypto world. As of late these words have been used in non-crypto world as well as a marketing tool and drive attention to crypto or non-crypto business endeavors. These words and the innovations they represent have deep meanings and revolutionary in nature. However, since they are used to describe anything and everything, overwhelming amount of information makes it impossible to really distinguish the genuine efforts in decentralization and web3 from the ones that are only intended for deception, profiteering, and in some cases scamming. The idea of decentralization is not a new one and the benefits it delivers are massive. Advancement of the internet and web technologies that are capable of connecting people in global scales and create networks with ease, made it possible for the awakening of the concept of decentralization with practical implementation models. The catalyst of the decentralization movement was the creation of the Bitcoin network.

Why Even Bother With Decentralization?

Decentralization removes the influence and power to make certain decisions on behalf of all participants of the network by a centralized entity, be it an individual, a group, a company, or a government. For a very long time, and still today organizations, communities, governments, networks have been operating within centralized frameworks. Centralized systems are not entirely bad. They worked just fine for a very long time. They are very efficient at what they do. The problem centralized system have always had was the abuse of the power by the central governing entities. When these entities have power to make decisions on behalf of the entire networks without any checks, they may take actions that are beneficial for themselves and those who are in their friends or partners circle, but at the same time these actions may cause harm to the other participants of the network.

For example, let's take a look at the US Dollar network. USD is used globally by people, organizations, companies, countries for all kinds of financial transactions. The only authority that has a power to control and influence this network is the Federal Reserve. The actions of the Federal Reserve like adding more dollars to the circulation, setting interest rates, getting involved in stock and bond markets do influence the value of the dollar, economies, and markets. If these actions benefited all participants in the dollar network equally or somewhat equally there wouldn't be an issue. But that is not easy to achieve, especially there is greed and power is involved. So, occasionally we may see some decisions made that only benefit few participants of the network, and rest pay the costs. US government is one of the participants of this network. Most of the time, US government gets the most of the benefits from the Federal Reserve, as it would never say no to print more money and give it as a debt. More supply would create more inflations, and devalue the dollars other participants have in their possessions. That is not just people. These participants are companies, non-profits, countries, etc.

Bitcoin was born at the time of the global financial crisis as an alternative to traditional financial systems, since it became too obvious how in centralized systems few can make decisions that produce not so positive outcomes for the rest, this time in global scales. The system Bitcoin offers is removing the power from central entities to dictate the monetary policy, embedding a monetary policy in a core protocol that can only be changed by the consensus of the network. Bitcoin achieves this by utilizing blockchain and proof-of-work consensus algorithm in an open and permissionless network. So far Bitcoin was able to attract many and build a decentralized network. However, it hasn't fully delivered all of its promises yet. Time will tell if this alternative monetary network is actually a better system compared to the traditional financial global systems. Nevertheless, it already has delivered some benefits like property rights, and ownership that doesn't require a need to trust some entities like banks.

After bitcoins creations many other projects have emerged in a rapid speed, all claiming to be advocates of decentralization. Some try to decentralize financial systems like bitcoin, others attempt to decentralize web, gaming, social networks, etc. They all can't be decentralized, can they? It is understandable that, decentralization is process and might even be very long one. Decentralized networks require centralized catalysts to get them started. That's why usually these catalysts may have too much influence in the network, and the question is how fast this influences diminishes and transfers to the other participants of the network. The core protocol, structure and vision of the network can sometimes suggest if it is actually moving towards decentralization. In some cases, the catalysts are not really interested in decentralization, they just use the term and blockchain tech to make money by deceiving others.

Prerequisites For Decentralized Networks - Open Source, Open Network, Transparency

Decentralized networks do not have to be digital. However, we live in a world that digital technologies enhance our ability to connect, communicate, and build networks. These technologies allow us to create better decentralized networks.

The core code of the network must be open source, so that anybody can review and make decisions for themselves if this is the network they would want to participate in. The network also must be open for anybody to join or leave. Being open network and open source, the network facilitates decentralized participation. Anybody can contribute to the code or participate in the network the way they choose. Openness also provide transparency, and any actions or activities that may harm the network or network participants can be easily discovered and mitigated.

Blockchain technologies are great with providing open source, open network, transiency features for the network. They also come with features like immutability, that transactions can only be added, and not removed, edited or altered in any shape or form. Blockchains also come with timestamps, distributed ledger, and unique private/public keys for all participants. So, the blockchains are great solution for building decentralized networks. The most powerful feature of blockchains is their consensus mechanism that makes the network trustless, without a need to trust any one entity.

However, blockchains are not always open networks and come with open source. Blockchain technologies can also be private with closed source and participation is based on permissions by central controlling entities. This can also be a great tool for some centralized companies, organization, and government agencies. Because the would still provide more efficient ways of doing traditional things, and be more transparent within that private network. In other words, just being a blockchain based network doesn't mean the network is decentralized. It is the openness of the network what helps us to determine if it has any intentions of being decentralized or not.

How Decentralized Is The Network?

Decentralization is a process. It is not a state. It is a journey. I don't even think there is such thing as perfect decentralization. In my view perfect decentralization would be when all participants of the network have equal amount of influence over the network. That is impossible to achieve, since regardless what consensus mechanism is used, the influence of participants always oscillates.

Most decentralized networks if not all start out with centralized catalysts, who are usually the founders and creators of the initial code and network. What makes them different from centralized networks is their intentions and goal of making the network decentralized. The protocol and tools they utilize may determine how fast they can reach the goal of decentralization or being decentralized enough.

I would consider that network has reached decentralization, when initial catalyst have lost most of their influence and the network is at the point of no return to centralization. In other words, it is no longer possible for one or group of people to gain control over the network. There are mechanisms in place that would discourage any ambitions to centralize the network, and any such attempts would do more harm to those who try than the other network participants.

Some would argue that there is always a forking option, if there were any attempts to take over the network. I believe this should only be considered as a last resort, and having this option doesn't necessarily mean the network is decentralized. In fact, if the only mechanism of fighting back centralization attempts is through forking, then the network is most likely not decentralized yet. The network would pass the decentralization threshold if it doesn't require forking as defense mechanism anymore, and can withstand any centralization challenges based on other factors.

The history of the network and how long it has existed matters as well. The longer the network existed without any security breaches, blockchain problems, the more trust it builds in its journey towards better decentralization. Forks that results in new chains, would restart this time clock and determining how reliable the network is.

Following are some of the factors and metrics we may consider when measuring and evaluating if the network is decentralized.

Distribution Of Influence

All participants of the network have an ability to influence it one way or the other. The influence may be significant or negligible. How participants may influence the network may be determined by the consensus mechanism. The question is how easy it is to consolidate this influence to implement changes in the core protocol to benefit the few and harm the other participants of the network. How easy it is for small group of people to dictate the future of the blockchain in a form of core protocol changes.

In a network where the influence is distributed broadly, it would be difficult of even impossible for small group to consolidate their influence to gain power. Let's say there is a bad player in the network, and it would be a good idea if the account of this player is removed from the account. Hypothetically speaking, this action could probably benefit 99% of participants. If some influence holders can combine influence to do something like this, the distribution of the influence may not be wide enough and the network is probably have not reached decentralization yet.

Any action that would harm an account, especially their property would violate the integrity of the network. In a network where influence is widely distributed, such action would take place, because network loosing integrity would put all accounts in danger.

Usually, the founders of the network or their foundation or their companies may have big influence in the network. The network won't reach decentralization until their influence become negligible. Of course this would all depend how much influence is enough to make protocol level changes.

Distribution Of Native Coins

In DPoS and PoS consensus mechanisms, the influence in the network is determined by the stakes. Other consensus mechanisms like PoW may not require stakes or holding of native coins in reaching the consensus. However, in all cases the distribution of native coins play a vital role in determining the decentralization of the network. Since, native coins in the network are valuable properties, large coin holders even in PoW networks may have indirect influence or use their property to acquire more influence.

But of course distribution of native coins plays bigger ole in any stakes based networks. Since native coins can be used to stake and gain influence in the network, we would want to know how widely the native coins are distributed within network participants, and see the likelihood of consolidation of these stakes and influence to take actions that may not be in the best interest of all participants.

Block Production Participation

One of the important participants in blockchain networks are the block producers. Depending on the network and consensus mechanisms implemented, they may be called by different names: block producers, witnesses, miners, validators. These participants keep chain of blocks growing by approving and adding blocks to the chain. They also approve or reject changes to the blockchain software and decide which version of the blockchain software to run. They do play a crucial role in blockchain networks and their participation also can demonstrate how decentralized the network is.

In an decentralized network, we would want to see a healthy competition for the role of a block producer. Since the network offers incentives to produce blocks, usually there is a good competition. Depending on what determines the selection of the block producers or their nodes to produce blocks in the network, this too can be manipulated by those with influence. Ideally, in a decentralized network influence would be shifting all the time, and this would also cause constant changes in block producers.

In an open network, anybody could put forward their candidacy for block production and run block production nodes or miners. Depending on multiple factors like time, skills, and incentives the interest to participate in a network as block producers would fluctuate. In a decentralized network these fluctuations would be healthy for the network. Stagnation of interest in joining the network as block producers is not desirable for a decentralized network.

Node Participation

Miners, witnesses and block producers are only one type of a node in a blockchain network. There are also other nodes blockchains need. Usually, rewards are paid to block producer nodes and there incentive to run these nodes. Other nodes while also play important role may not have any incentives to run them. We also need to consider the participation of these other nodes in a network.

For example in Bitcoin networks, in addition to miners there are also full nodes, light nodes, pruned nodes, etc. Bitcoin nodes broadcast, validate, process and store BTC transactions and blocks. These nodes may not necessarily require expensive equipment as miners, yet still play crucial role in security and decentralization of the network. On Hive, in addition to witness nodes, there are also full and seed nodes. There is no incentive to run full nodes on Hive. But they are very important, and usually are run by top witnesses who receive higher rewards for producing blocks as they produce them more frequently compared to backup witnesses.

But it is not a requirement to be a block producer to run a full Hive node. Some people may run it because they have the resources to do so and want to help the networks, some people run them because they are building apps or other solutions in the network. The amount of active nodes in the network would also help us in understanding how decentralized the network is.

Core Development Participation

Since decentralized networks are based on open source software, everybody is welcome not only review the open source code but also contribute the the core code. Usually the core code is developed and maintained by the core team, founders, or other catalysts. Decentralized network would attract more developers to contribute to the core code and gradually remove the relying in maintaining and developing of the core code by the initial catalyst team. The network will never reach true decentralization if it doesn't attract more developers who have interest, and skills to contribute to the core code. Because this would mean the entire network would rely on the skills, resource, and vision of a centralized entity. This in itself would create influence by a small group of over the network.

While the process of decentralization is gradual, there should be signs of growth in participation in the core development. If the network relies only on the initial catalyst team, and one day the team loses interest and abandons the project, this may lead to the decline of the network unless somebody else steps in. In a decentralized participation, developers joining or leaving the core team would affect the network, because there would always be continuous development that doesn't rely on one individual or a group.

Layer 2 Development Participation

Just like core development is important, layer 2 development activity is also important in decentralized networks. Most networks are built to attract layer 2 development in the network. Decentralized networks provide solutions to create better tools, apps, games, and communities. The true innovation of the decentralized networks is discovered through creativity and building of layer 2 solutions. First, layer 2 developers must see the benefits of building anything on top of the layer 1 blockchain networks. If there is no value of doing so, nobody would even consider. One of the benefits is decentralized networks provide access to the network and its participants.

There are also other benefits like built in account creation systems, native payments systems, and other features the blockchain may offer. Some developers may be interested in smart contracts, others may be interested in data storage, etc. Levels of interest in building apps in a network, and development activity is also provide good metrics to see how decentralized the network is.

Network Participation and Activity

Not everybody is interested to be investors, developers, or block producers in the network. People join networks for various reasons. I tend to think decentralized networks try to offer better alternatives to the traditional centralized systems. This may not be the case all the time. But for the most part the ideas and intentions are there. Any network can grow with continuous increase in participation by ordinary people. A better network should be able to offer ordinary participants to engage in activities they already do in centralized networks, but with more benefits to them comparted to the centralized networks.

For example, many people use bank accounts on daily basis. This is convenient for most but also come with risks. People don't really have full control over their property. By participating in Bitcoin network, people can take full control over their funds and become their own banks. With this ownership comes responsibility to keeping these funds secure as well. There is no customer service like in a bank that would help with retrieving funds if one lost their private keys. They may be gone forever. It is up to individuals to decide if it worth participating in decentralized networks.

Participation in these networks and activity increase or decrease can also show us important information whether the network is heading towards better decentralization or not. Any decline in network participation and activity would suggest lack of interest in the network. Decentralization requires participation. Only by participation and its gradual growth it is possible for a network to achieve decentralization.

I am sure there are other aspects of network to consider when analyzing the levels of decentralization. These were the few I could think of now. If you have any others to add, feel free to share in the comments. What do you think? What makes a network more decentralized?