“You can have it good, fast or cheap. Pick any two.”

This so-called Project Management Triangle refers to the unavoidable trade-off engineers have to make during the development process.

But why should you care? Well, the way each project goes about solving the Blockchain Trilemma will determine whether:

• All token holders have equal access and voting power in the network
• The network can resist attacks that try to shut it down or steal funds from user wallets
• Your favorite project will be able to support mainstream adoption

A blockchain project that can do all three of those things will have found the holy grail of crypto: a lasting solution to the Blockchain Trilemma. In this article, we’ll explore the root of the problem, how Layer 1 and Layer 2 solutions are trying to solve the trilemma, and how five of the top blockchain projects stack up in that regard.

To understand the Blockchain Trilemma, we first have to understand the differences between centralized and decentralized networks.  

A centralized network is a group of interconnected computers (i.e. nodes) that depend on one master node to store and process data. Most of the web apps we use today, like Google, YouTube and Facebook, are built as centralized networks that serve content to end-user devices.

There's a good reason for that too. Centralized networks are:

• Faster because all nodes answer to one master.
• Easier to maintain because only one server needs to be updated.

But on the flipside, centralized networks are also:

• More prone to crashes because they have a single point of failure.
• More vulnerable hacks because attackers only need to compromise one target.

Distributed networks, on the other hand, consist of interconnected nodes that share the workload. These nodes are often spread out across the globe but might still follow some kind of leadership hierarchy. Hence, distributed networks still allow for partial centralization.

In doing so, decentralized networks are:

• More resilient, as one failed node doesn’t bring down the whole system
• More transparent since data is broadcasted for everyone to see
• More corruption-resistant, since attackers can’t manipulate the system without help from a majority

But decentralized networks do have their trade-offs, like:

• They are difficult to maintain because every single node needs to be updated.
• You can’t police bad actors because the system is permissionless.
• Not every node is equal, some have more data and processing capabilities than others.

While there are many unique challenges to running decentralized networks, most of them can be grouped into 3 categories that make up the blockchain trilemma.

The blockchain trilemma is a framework that says that every public blockchain has to sacrifice one of three features: decentralization, security, or scalability. Let’s dive into each.

One key thing is required for decentralization: a large number of independent nodes distributed across the globe. But here’s the kicker. Since every single node has to reach consensus before a transaction is finalized, becoming more decentralized actually slows down a network. And that’s not good for global adoption.

Now to make a network faster, you actually want to reduce the total number or distance between nodes. But that raises the issue of security.

For a distributed network to reach consensus, a majority (51%) of nodes have to agree. So the fewer nodes there are, the easier it becomes for hackers to take control of the network through a 51% attack and print infinite money for themselves (also called the double-spending problem).

A fork of the Ethereum network (ETH) called Ethereum Classic (ETC) actually suffered from 51% attacks multiple times in 2020, costing users millions of dollars. All because the network was too small and thus easy to control.

In addition to that, blockchain security requires each node to have skin in the game. In other words, it needs to cost them something to participate—for instance, computing power or financial collateral. 

But why? Otherwise, malicious actors would carry out Sybil attacks, spamming a network with transactions until it crashes. Solana (SOL), for instance, was knocked offline on September 14, 2021, after hackers flooded the network with up to 400K transactions per second.

Okay so clearly decentralization and security are non-negotiable. But what does that mean for scalability, a blockchain’s ability to maintain or increase speed even as new users join the network?

For blockchains that don’t want to compromise on decentralization, that means slower transactions. Bitcoin, for instance, is only able to process 7 transactions per second (TPS). Blockchains like BNB, on the other hand, choose to be centralized for the sake of blazing-fast transaction speeds.

Scalability is ultimately necessary for blockchains to be able to rival legacy networks like Visa, which can process up to 65K transactions per second. So how are the leading blockchains aiming to scale?

Two of the most common buzzwords in crypto are Layer 1 and Layer 2 solutions. But what do they mean and why are they essential to blockchain scalability?

A Layer 1 solution is an attempt to change the internal rules of a major blockchain to make it faster and more efficient. Examples of layer 1 solutions include:

• Increasing block size so that more transactions can be fit into each block.
• Increasing the rate at which each block is confirmed.
• Breaking transactions into shards that can be processed in parallel. Zilliqa, for instance, does this.
• Transitioning to a less energy-intensive consensus mechanism.

Ethereum’s upgrade to Proof-of-Stake (PoS) is one example of the latter solution. While Proof-of-Work (PoW) requires miners to use substantial computing power, PoS only requires validators to put up enough financial collateral to secure the network.

A Layer 2 solution is a third-party tool that connects to a major blockchain and takes some of the load off. And after layer 2 is done processing, it sends the results back to the main chain. Examples of Layer 2 solutions include the Lightning Network for Bitcoin and Polygon (MATIC) for Ethereum. 

At the end of the day, layer 2 solutions allow major blockchains to have their cake (decentralization, security) and eat it too (scalability). Hell, projects like Ethereum and Cardano (ADA) combine both Layer 1 and 2 solutions to eke out every last drop of network performance.

Measuring how well each project solves the blockchain trilemma is a hard task. Even more so when comparing networks with different consensus mechanisms or definitions of what constitutes a transaction. 

But trying to measure this is still necessary in order to know if a project has true staying power. And to make this more of an apples-to-apples comparison, we’ll only focus on Proof-of-Stake (PoS) smart contract platforms.

Decentralization & Security

These two pillars of the Blockchain Trilemma can be measured by variables like:

• The total value staked in the network
• The percentage of the total coin supply that’s staked
• The requirements to become a validator—the lower the better
• The total number of validators 
• Whether there’s a saturation limit for staking, which incentivizes users to stake to smaller pools

The key variables that determine blockchain scalability are:

• Block size: A block is a bundle of transactions, with size limiting how many transactions can be processed in one epoch.
• Block time: How long it takes validators to verify transactions within one block.
• TPS: The number of transactions that a blockchain can handle per second. In the below table, I don’t consider a blockchain’s theoretical TPS, only the current one.
• Transaction cost: A fee that’s charged to process transactions. Higher fees make a blockchain less usable (but ironically more secure).

Welp, that’s a wrap. In addition to giving you some tools to compare blockchain projects, here are a few takeaways and disclaimers:

• Decentralization, security, and scalability exist on a spectrum, and none of the major blockchains are entirely good or bad. 
• Each component of the blockchain trilemma is affected by an array of variables, and we can’t measure every single one.
• Reliable data isn’t always available, so I relied on metrics that can be verified for the most part. 
• Choosing a different set of metrics from the ones listed above can paint an entirely new picture of each blockchain’s competitiveness.