August 10, 2018
Why the Real Bitcoin Has Big Blocks: Episode 1
Whether you have been involved with Bitcoin and cryptocurrencies for long, or just beginning to get interested: the discussion of what makes Bitcoin work is an ever-present enquiry into any action performed by investors; developers deciding on which platform to work on, venture capitalists deciding which platform will be the most valued one in the future, miners performing Proof-of-Work on the most profitable chain, or users finding out which is the most useful in terms of costs, speed, and security.
It is the decision between Bitcoin Core (BTC) and Bitcoin Cash (BCH), a decision between small blocks and big blocks, and whether to use payment channels via the so-called Lightning Network, or on-chain scalability to reach the billions of potential users on the planet who would benefit from a peer-to-peer cash system that is trustable, inexpensive, and both easy and fast to use.
Satoshi Nakamoto and year-long developers Gavin Andresen and Mike Hearn, who followed into the founder’s footsteps, have both made very clear suggestions on how Bitcoin shall scale, and that Moore’s law would ensure that on-chain scalability would prove to stay the one form of scaling which has brought Bitcoin to where it is now. But that is not even the fundamental reason why I and many others came to believe in bigger blocks and Bitcoin Cash. The fundamental reason is of economic nature. Big blocks are the only way to scale Bitcoin, and any other method of scaling would end in effectively killing the blockchain that is Bitcoin.
In this series I intend to explain why big blocks are necessary to keep the blockchain and vision of Bitcoin alive. It strives to explain the basic but essential economic structure that makes Bitcoin work, and any attempts in trying to reject it would result in rejecting economics altogether.
1. On the Incentive Structure of Bitcoin
Bitcoin is a database that is constantly updated and secured by a consensus mechanism that requires heavy computational power performed by so-called miners or nodes. The process involves solving a hash puzzle and is called Proof-of-Work whereby the miner with the most processing power is most likely to solve the puzzle and add a block onto the blockchain. Each block contains transactions signed by users and relayed by nodes. Because the Proof-of-Work consensus mechanism implies large amounts of energy and thereby costs, miners have two monetary incentives in the form of electronic coins to secure the network through Proof-of-Work:
- The creation of new coins that are allocated to the node adding a block (currently these are 12,5 coins per block)
- The sum of all transaction fees chosen by the users (under big enough blocks these are around a penny or less per transaction)
It is important to realise, though, that there is another implied incentive structure that comes with each node deciding and thereby voting (per computing power) which block to mine on top of, which is not solely dependent on who first solves the hash puzzle, but first and foremost on transactions being valid, meaning that they must be the first of any 2 conflicting ones and signed by their owners.
This has severe consequences for the ‘behaviour’ of nodes, because it is in their mutual interest to at all times be aware of all relayed transactions, lest they include a transaction that is fraudulent or a double spend, as we call it; if a node were to intentionally or unintentionally include a fraudulent transaction, then other nodes will simply not mine on top of such a fraudulent block, as it is in the interest of all nodes to keep the network working and therefore honest, as their future income is at stake.
The result is a network that is highly interconnected, with every node both relaying and including all valid transactions they can possibly get as is the pareto-optimal case to constantly do so. Such a network we call a small world or near-complete graph, and in simple illustrative terms we can imagine it to perform in the following way:
It is important to note that such a process happens within milliseconds, because time is of essence for nodes, as time represents a major cost factor; if one node is not fully up to date, they run risk of being orphaned, which means other nodes decide not to mine on top of its mined block, making its coin rewards worthless.
The security of the Bitcoin blockchain stems from precisely such a network effect: the most robust nodes are financially incentivised to mine faster than anyone else in the network, and financially disincentivised to break the rules of the game, leading to solely including transactions that are valid.
2. Where it Gets Interesting: Scaling Bitcoin
With every block mined every 10 minutes, we get closer and closer to the block-reward halving that occurs about every 4 years—the next one being due in approximately 2020. Miners will have less new coins as a reward and will be more and more reliant on the sum of all transaction fees instead, until miners are exclusively reliant on transaction fees.
There are 3 ways miners will continue to spend vast amounts of energy to secure the network:
a) either the users are willing to pay continuously increasing fees for a transaction, or b) we keep creating new coins for miners and thereby the total supply of coins (which is unlikely to happen given the deeply rooted fixed supply in the Bitcoin protocol), or c) we simply allow more transactions to take place per block, increasing the total transaction-fees amount through massive volumes of trade.
For Bitcoin to stay Bitcoin as a non-inflationary currency for all people, including people with little purchasing power, Bitcoin must scale on-chain—big time. This means the blockchain itself must scale: if we imagine millions of transactions per second happening on the chain, the monetary incentive for miners grows with it—which in turn strengthens the security of the network.
But if we scale Bitcoin not through bigger blocks, but by imposing an artificial block-size limit as is the case with the Bitcoin Core blockchain, and instead through an off-chain layer on top of the blockchain such as the Lightning Network, which by definition would not offer the same on-chain transaction volume, we are left with the options of continuously increasing on-chain fees, or inflation. And the more the Lightning Network is used for off-chain scalability (that is assuming it were to work), the greater the extent of the high fees or the inflationary supply to even cover the energy expenses endured by nodes.
Even if some people decided to provide nodes at their own costs, the incentive for more computationally powerful nodes to mine on a blockchain that scales to billions of users on-chain would by far outweigh the incentive to mine on a network that is limited in its block-size capacity—putting the security of such a network at great risk.
After analysing the fundamental nature of the incentive structure of Bitcoin, it becomes clear that we have the choice between a network that is either getting more and more expensive to use, will have an inflationary supply of coins, or that is no longer secure to use, and a network that scales on-chain—keeping it fast, secure, and inexpensive to use, with all its simplicity that we all once fell in love with over the past 10 years. It is therefore not too far to think that Bitcoin Core is planning to use the Lightning Network as the sole form of scaling, and—as Bitcoin as a settlement layer would merely be a transactional barrier—get rid of the Bitcoin blockchain altogether. But in all those scenarios: which one to you sounds like the true form of Bitcoin once described as ‘Bitcoin: A Peer-to-Peer Electronic Cash System’?
In the next episode I shall look at why big blocks are not a threat (but the solution), and the importance of 0-conf (instant transactions) with 10-min block-confirmation times.