The Blockchain Trilemma: Scalability, Decentralization, and Security

What Is the Blockchain Trilemma?

The Blockchain Trilemma is a dependency problem when trying to achieve absolute security, decentralization, and scalability. It claims that trying to improve two of these variables will negatively affect the third one. The blockchain trilemma is why there are so many blockchains, each taking a different approach to its solution.

Cryptocurrency mass adoption is every investor’s dream. But it’s not as simple as just getting more buyers. In fact, that’s the biggest reason we haven’t seen higher prices and adoption (only about 300M people use it).

The biggest blockchains are “suffering from success.” They invented crypto payments, smart contracts, and NFTs. But Bitcoin and Ethereum weren’t designed for today’s user volume. They’re now dealing for years with the blockchain trilemma, which will likely remain unsolved for a long while.

The blockchain trilemma is a development choice. There are three choices, each with its pros and cons. No matter which one you choose, some will love it and others will hate it.

Why is it a big deal?

Because you put at risk a lot of money from millions of people worldwide. Balance isn’t feasible because not having at least two can be catastrophic. If you chase all three, you’ll achieve none, or at best, have a mediocre network that nobody uses:   

• Without absolute security, any minor bug is an opportunity to breach billions of dollars.
• Without decentralization, cryptocurrencies wouldn’t exist in the first place. It’s the essence of blockchain.
• Without scalability, your network adoption is limited. If left unsolved, user growth will make it slow, expensive, and less attractive than other competitive chains. For every new user that joins, you lose another one.

You could look for the solution in a fourth variable, although not recommended. More variables add complexity, and complexity often worsens problems. While there might be a theoretical solution, the practical decision is to choose two at the cost of one.

Why Are There So Many Blockchains?  

Depending on which ones you prioritize, you get a different network type. That’s why there are so many blockchains, and why new ones will keep appearing. While looking for a hypothetical solution, developers have many variations to explore.

You first choose which one to leave behind and then which one to specialize in from the two. This creates at least 27 combinations. And because different solutions can solve the same problem, there might be several projects for every option.

Cryptocurrencies are, by definition, public blockchains. Their strengths and weaknesses are decentralization and scalability.

Curiously, the blockchain that struggled the most is the one that discovered it. In 2018, the co-founder Vitalik Buterin coined the Blockchain Trilemma to explain the challenges Ethereum faces in development. Seemingly, today’s closest solution to the blockchain trilemma is a hybrid blockchain called Algorand.

But to find possible solutions, we first have to know how the three variables interact.

Blockchain Trilemma: Security  

Blockchain security has little to do with traditional security. Financial institutions improve security to increase their trust, whereas blockchains remove trust altogether. Instead, we trust in a hard-coded system. 

Consensus mechanisms work as security systems that allow users to trade with each other directly. In this new approach, there are no 3rd-parties or regulators. In a way, every network node is responsible, and the bigger the network, the better the security.

Scalability is pointless without security because it only puts more people at risk. Decentralization often goes hand in hand, but without secure consensus mechanisms, the number of nodes doesn’t matter. 

How Decentralization and Scalability Are “Bad” For Security

Decentralized networks replace trust with fault-tolerant mechanisms. But if those systems have design flaws, everybody using them is at risk. You won’t know it until it’s too late, and because there’s no authority, no one is responsible for your losses.

Scalability also limits security because it often comes at the cost of centralization. You might think the network is slower when there are fewer validators than users. The reality is the opposite because fewer nodes will reach consensus faster.

Reducing nodes is like putting your eggs in one basket. It also means it takes less work to attack or control the network. The largest blockchains had the most incidents when they were small.

Blockchain Trilemma: Decentralization

Decentralization is the lack of a central authority. Like the Internet, decentralized blockchains are run by everyone, and no one alone can control them. They use a consensus mechanism that’s designed to prevent centralization.

The ideal decentralized network would have thousands of nodes and evenly share control. Decentralization is synonymous with permissionless, which means there’s no entry barrier. Anyone worldwide can join within minutes.

Without decentralization, everyone’s security depends on a single point of failure (SPOF). You have to hope there are no security flaws (no system is perfect), and also that the authority controls it as intended.

How Security and Scalability Are “Bad” For Decentralization

A blockchain is decentralized as long as nobody can control 51% of the network. Depending on the consensus model, it might mean having the most computing power or the most tokens at stake. It’s easier to beat the majority rule when there are fewer people.

Some networks limit how many validators there can be because it’s efficient. Fewer validators don’t have to share as many network fees, and they also reach faster agreements as to what blocks to accept. So they don’t allow new validators, or they raise the minimum requirements a lot.

Decentralization is also compromised when security matters more. For example, centralized exchanges have to comply with regulations and impose terms. They can’t afford to risk money on high-threat-level users, so they add verification requirements. 

If networks can control the number of users (security) or validators (scalability), they’re centralized.

Blockchain Trilemma: Scalability  

Scalability is the ability to maintain or increase efficiency while the network expands. A blockchain is scalable when it always maintains high transaction speed and low costs. Potentially, it means there is no limit to how many users or trading volume it can handle.

Scalable cryptocurrencies gain value because more people use them. Scalable utility tokens also gain value because it speeds up the development of new features. But scalability alone can’t bring users to the network.

How Security and Decentralization Are “Bad” For Scalability

Large blockchains are the most affected by scalability. By definition, scalability is the ability to adapt or control performance while growing. Decentralized networks don’t allow any direct control.

Secure networks also tend to lose efficiency:

• There are more users waiting to confirm their transaction blocks
• There are more validators, which slows down block time in certain consensus mechanisms.
• If there are too many or too few validators, network fees skyrocket.

Because scalability affects how many people can efficiently use it, the lack of it prevents true decentralization. So what’s the solution?

Blockchain Trilemma: Practical Solutions  

When it comes to public blockchains, decentralization is paramount. It happens to be closely tied to security, which is why the large blockchains are good at both. Scalability isn’t really an issue until the network gains adoption.

Therefore, scalability solutions are the simplest answer to the blockchain trilemma. Not only a solution but a reality. At least four scalability features are already in practice on different blockchains:

Interoperability  

In blockchains, interoperability is synonymous with communication. It’s the ability to connect incompatible networks to share data and infrastructure. This connection allows blockchains to focus on their strengths while covering each other’s weaknesses.

By default, most blockchains aren’t compatible because they use different consensus algorithms and code. To link them, we use relay chains like Polkadot and decentralized oracles like ChainLink. Not only does it solve the trilemma, but it also makes the whole blockchain technology more decentralized. AKA more user-focused.

Interoperability solutions are often called Layer 0. Layer 0 refers to all technologies that allow connecting Layer 1s (which are blockchains like Ethereum). For example, the Internet is another interoperable Layer 0.

Layer-2 Blockchains

Layer-2 blockchains are extensions of Layer-1 blockchains. They inherit the same infrastructure while improving performance. For example, Arbitrum and Optimism are faster networks (L2s) that inherit Ethereum’s security (L1).

Layer-2s and 1s work together, one specialized in scalability and the other in architecture. On Ethereum, L2s are used for a scalability strategy called “roll-ups.” They improve speed and reduce costs over ten times by processing L1 transactions on L2, and then sending them back as an L1 block.

When L2s have higher speed and capacity, you can complete more transactions within the same block (and network cost).

Sometimes, L2s develop their own ecosystems, which divides liquidity from the L1. When blockchains create independent infrastructure rather than inheriting it, they’re called sidechains.

Sidechains

The most popular Ethereum sidechain is Polygon Network with a block time of ~2.3 seconds, $0.01 to $0.20 on network fees, and an independent dApp ecosystem. Unlike L2s, it can run regardless of what happens to Ethereum. But security won’t be as high because it’s a smaller network with a different infrastructure, roadmap, and team.

What makes sidechains relevant is compatibility. Polygon is Ethereum-Virtual-Machine compatible (EVM), which allows developers to easily build or migrate from Ethereum. While Ethereum and Polygon ecosystems consist of different tokens, more and more projects start supporting both.

Shards

Sharding is a load-sharing method that divides a database into partitions. It improves scalability because validators will only have to verify blocks from their shard. They can still read all other portions, but they only verify their own.

While shards are easier to attack than standalone blockchains, attackers first need to know who the validator is (which isn’t easy due to randomness). 

One of Ethereum’s future updates is to divide it into 64 sharded blockchains. This will reduce hardware requirements so more validators can participate. These new networks will help Ethereum to share data load and reach 100,000+ transactions per second (TPS).

PoS-Based Consensus Mechanisms

Many intricacies of the blockchain trilemma can be avoided if you start with the right consensus model. Public blockchains are immutable, so there’s not much you can do to improve scalability once they’re running. Still, big blockchains can switch models if everyone agrees, just like Ethereum switched from proof-of-work (PoW) to proof-of-stake.

PoW was the default mechanism for the first cryptocurrencies developed before 2015. It’s still the most secure because a network based on computing power is very expensive to attack. It’s scalable in the sense that decentralization improves security (but it’s slow and expensive).

A hundred times cheaper is the proof-of-stake mechanism. Almost every modern blockchain uses some variation of PoS for this reason. To “break” a proof-of-stake blockchain, you have to beat 51% of the network in locked holdings, staking length, luck, and voting majority. Unfortunately, too many protocols prioritize holdings, which leads to centralization/ manipulation.

The most successful blockchains don’t use PoS but variations. PolkaDot uses nominated PoS, Algorand uses pure PoS, and Solana uses a hybrid between PoS and Proof-of-history (PoH). These result in lightspeed transactions and unlimited scalability.

Will the Blockchain Trilemma Be Solved?

While there are already solutions to the blockchain trilemma, it will take many years before a blockchain solves it alone. So far, none of them has achieved the three absolutes. The closest ones are PoS hybrids, which tend to give too much weight to token quantity.

Without enough new users joining the network, it tends to centralize. Some will argue that large holders aren’t a problem because they have the most to lose on bad decisions. So while it’s “safe” in practice, it doesn’t solve the trilemma. It’s recommended to increase the minimum number of validators or set a cap to the weighted validator lottery. Sharding and pure PoS are good examples.

It may not seem like it, but Ethereum is becoming more scalable every year. Once the Pulsechain fork is live, load sharing will allow Ethereum to save even further in costs and speed. Still, efficiency is relative to user growth.

It took 40+ years for fiat currencies to reach billions of people, 30 for the Internet, and ~25 for digital banking. Blockchain has been around for ~10 years, and it still doesn’t scale enough with 300M users. It’s why solving the blockchain trilemma is a huge step for mass adoption.