Matthew Zietzke: Fundamental errors of BTC and Lightning Network

Getting your Trinity Audio player ready...

We already know that BTC itself cannot scale and that Lightning Network does not solve BTC’s technical and economic problems. However, only few are able to really pin down the underlying flaws of BTC and Lightning Network. Educator and consultant Matthew Zietzke is one of them, and that is why we caught up with him to hear about the fundamental errors in BTC and Lightning Network.

Hi, Matthew! You have been active in Bitcoin SV for some time and BSVers do appreciate the information they get from your Twitter activity. Today we want to hear about a Twitter thread that you recently posted:

On the subject of whether or not LN "works just fine" if mining becomes consolidated into a mono/duopoly (as is inevitable in a hash-competition-only mining market,) consider what happens when "Miner-Extractable Value" comes into play…https://t.co/5vFt3QmfIi

— Matthew Zietzke (@MZietzke) April 19, 2021

First of all, kindly introduce us to the idea of Lightning Network—from a BTC (!) perspective.

Matthew Zietzke: First, thank you for reaching out to discuss this with me!

I’d like to answer this question by first saying that I will do my best to give a fair steel man position for what most BTC supporters, in my experience, believe. With that in mind, it is inevitable that my characterization is unfit for at least some BTC supporter’s beliefs on how the system works. I do believe, however, that my account here is at least a good characterization of the general BTC narrative.

Before we can understand why BTC supporters are excited about the Lightning Network and why they believe it is necessary, we have to understand the small blocker ideology.

To a BTC supporter, the Bitcoin network is “decentralized” at least in part because every user can (whether they choose to or not) validate the entire history of events on chain from the genesis block to the latest UTXO set (the latest block) on their own hardware, without prohibitive expenses, to ensure that the network is abiding by their preferred rule set.

To them, if the costs to “run your own node” become prohibitive to a sufficiently large number of people, the network will not remain “decentralized.” This is because they (correctly) identify that higher costs of validating the chain will increase barriers to entry for both producers of the chain (miners), and consumers of the chain (users and businesses) who wish to validate that the chain is following their preferred rule set.

BTC supporters generally believe that high barriers to entry are the biggest factor in determining how “centralized” or “decentralized” the system is.

This begs the question, what is “decentralization”? I have yet to encounter a definition of decentralization proposed by BTC supporters, influencers, or those recommended as experts that is either consistently applied or, in my opinion, satisfactory to match how they use the term. While I use the terms “centralization” and “decentralization” to refer to relative distributions of market power and market share, I have encountered very few who have accepted this definition.

Regardless, BTC supporters believe that “decentralization” is required to prevent anyone, whether a single party or a cabal of conspirators, from changing the rules of the system without the consensus of a vast majority of the network as a whole, which includes both producers (miners) and consumers (users and businesses).

In their minds, this is essential to the security of the Bitcoin system, which is driven not by the design of Satoshi, but by the network consensus (which again includes all network participants regardless of their status as miners). In a sense, BTC supporters are proponents of the “death of the author” concept, at least as relates to Bitcoin.

All of the above is to say, BTC supporters believe that small blocks are necessary to keep the system “decentralized,” and that “decentralization” is essential to the security of the system.

What about Lightning Network though?

For those who believe as the BTC supporters do, the Lightning Network and similar “second layer scaling systems” are a necessary and natural byproduct of what they believe are inherent limitations to Bitcoin’s design.

For those like myself who believe the BTC supporters make fundamental errors in their understanding of how Bitcoin works, Lightning seems problematic at best. But more on that later.

From a more technical perspective, the Lightning Network is intended to enable participants to transact many times in a form of payment channel where BTC coins are locked in an on-chain contract (establishing the payment channel) with a built-in incentive structure to ensure that the channel remains “trustless” such that if the other channel participant tries to cheat you, you can proactively prevent such cheating, and usually punish the cheater.

Once coins are locked in a channel, the coins within the channel can be exchanged back and forth an unbounded number of times without ever posting a transaction on-chain.

Additionally, if your channel partner has channels open with other people, that partner can “relay” transactions between yourself and those other people.

This allows you to create a mesh network of channel participants who can exchange directly with those who they are connected with, or indirectly with anyone who can be routed to through existing channel participants.

In this way, the Lightning Network is said to enable second layer scaling on top of the BTC network layer. Said differently, you can theoretically transact as much as you like on the Lightning Network, and all that is recorded in the small block, limited BTC chain is your channel opening transaction.

Now, if you ever wish to close a channel (thus unlocking your funds on the base layer), this requires another on-chain transaction similar to the channel opening transaction. Once closed, a channel loses all incentive structures, and is subject to the consensus established on-chain regarding the state of that channel.

Even if it’s possible to create seemingly valid transactions within the payment channel structure, the channel has been settled on-chain, where consensus of the actual coin ownership is recorded on the public ledger.

Denying the ownership records established on chain is, therefore, either an exercise in “running your own node” where you disagree with the rules established in the chain you are rejecting, or rejecting the “not your keys, not your coins” mantra that is also common in the BTC community. Neither of these options is consistent with a smooth operation of the Lightning Network.

Now kindly tell us how most, if not all, Bitcoin SV supporters think about the idea of Lightning Network.

Matthew Zietzke: The Lightning Network is a boondoggle project that has demonstrated, repeatedly, that it is an impressively bad security risk for its participants, a supremely poor user experience, and a never-ending repetition of the idea that it will be ready in “18 months”.

Speaking for myself, I see no way that the developers of the Lightning Network can possibly solve the complex combination of problems surrounding the routing of payments across its mesh network given fee, channel state, and node uptime variability, and restricted by the requirement for coins to never be double-spent (you may only execute a transaction through one route, atomically, which contrasts with data routing over the internet where information may be sent many times to overcome a failure of transmission or routing at any point along the route).

In fact, solving these problems would appear to me to solve the problem of increased validation costs with increased block size on the base layer. For those interested in diving more deeply into why this is, these are all NP-hard problems. Solving one of them, it is said, proves that P = NP, meaning that there is a scalable solution to all problems currently classified as NP and NP-complete problems. This includes the asymmetric cryptographic functions that make Bitcoin and the Lightning Network possible in the first place (and the internet for that matter), thus bringing into question the very foundations of the Bitcoin system. However, the specifics of this are well beyond the scope of this interview.

To put it more plainly, if the Lightning Network’s routing problem can be solved, then scaling on-chain in an affordable way can be solved, too, making the Lightning Network unnecessary in the first place even assuming the BTC supporters are correct about the security problems resulting from increased costs of validation. And, if they can solve that problem, the internet is likely to be broken, too, and Bitcoin along with it, in every form it currently takes.

I, for one, don’t see an upside for the Lightning Network anywhere in that realm of possibilities.

To understand your thread better, feel free to let our readers know what Miner-Extractable Value (MEV) is.

Matthew Zietzke: Miner-Extractable Value is, generally, income that miners may take (and therefore, may also distribute,) as a result of how and what they choose to mine transactions. Specifically, in cryptocurrencies with second layer architectures such as Ethereum’s Dapps and DeFi systems, miners (or stakers, in Proof of Stake systems,) may choose to include transactions in their chain in a different order than the transactions were actually sent by network participants. They do this by including or not including given transactions in their block(s). This is called transaction reordering.

Miners may be able to take advantage of transaction reordering to extract value (income) out of the immutably posted actions created by users of the system. The specifics of how this is done vary by the systems in use, and it appears strategies will evolve over time in an attacker-defender balance structure. For a time, those seeking to unfairly extract value will have more success, then defenders will implement methods to prevent such gaming to occur, which leads attackers to develop new strategies to game the system.

Currently, DeFi systems with the ability to create oracle driven smart contracts to place a spot purchase or sale order are especially at risk, because a miner may see the spot order, then post a spot order of their own and include it before including the user’s order. Then, when the user’s order is fulfilled, it is at a different price, which was affected by the miner’s fulfilled order.

In fact, if the same miner is lucky enough to publish a block that includes that user’s order, this miner may sell the coins they bought at the price the user sought to purchase at, and to sell them at a price that is advantageous for the miner. Near-instant profits for the miner, with minimal risk, all enabled by the fact that the miner chooses transaction ordering.

Miner-Extractable Value is a general class of problems that are generally a result of small block ecosystems, as in these systems, the first-seen rule cannot be applied. It is the limitation of block space that leads miners to optimize transaction inclusion in a block by fees or some other metric, rather than by the order in which they are seen.

Miner-Extractable Value also comes into play when interacting with other second layer systems such as the Lightning Network, where the “trustlessness” of payment channels depends on the incentive structures involved in its creation.

If a miner can choose which transactions to include in its chain irrespective of the double-spend rule, then it can act as an external incentive structure that breaks the rules of the subordinate system.

Said differently, miners have incentives, too, and those incentives aren’t controlled by the Lightning Network’s internal incentive design. Any consensus problems that exist on the main chain are inherently a problem on a subordinate system such as the Lightning Network.

In this case, a majority miner may choose to partner with Lightning Network channel partners to close those channels with a state that is advantageous to the Lightning Network participant, for a modest fee given to that miner. To the miner, it’s a win-win: they have moved transaction volume back onto their chain, while getting paid to settle the channel in a manner advantageous to the Lightning Network participant.

This would be called a breach in Lightning Network vernacular, and it is typically thwarted by sending a “punishment transaction” which moves all funds in the channel into the victim’s wallet. This, however, requires the victim to be able to get their transaction on chain before the time-lock on the channel closure transaction is passed.

This is not possible if the majority miner refuses to include such a transaction in their chain. As I understand it, at best, if the breach transaction is discovered and is replayable on the victim’s preferred chain, the victim may broadcast their punishment transaction on their chain. However, this is not necessarily the case, and is likely to be thwarted by any miner seeking to implement such a strategy.

You see a fatal flaw in Lightning Network:

Majority miners can literally incentivize LN users to unilaterally close channels (and to do so at the expense of the other channel partner, if there's a previous channel state that allows this,) so that they can operate directly on the majority miner's chain…

— Matthew Zietzke (@MZietzke) April 19, 2021

Our readers—coming from within the Bitcoin SV space—might ask themselves: closing channels? Isn’t Bitcoin about not needing to have payment channels at all that could be closed?

Matthew Zietzke: Payment channels aren’t in and of themselves problematic. In fact, the concept of a payment channel has been around for a very long time, and is even being implemented in BSV. The difference in BSV is that you don’t require a transaction to open a payment channel on chain, and settling the payment channel is a single transaction that occurs on chain, and may even be indistinguishable from a normal transaction.

Instead of seeking to create a second layer network with payment channels, in BSV, we use payment channels to transact directly between two parties, with channel closure occurring at the end of their interaction. A proposed use case for this design is to enable streaming data where users pay per packet, and events such as a failure to pay ends the stream, with the channel closed and payment settled on chain.

In BSV, we don’t rely on kludgy solutions to shore up problems created by shortsighted, “expert” developers.

You tweeted that the main problem with Lightning Network is rejecting Nakamoto Consensus. Could you walk us through this step by step? What is the Nakamoto Consensus and how does Lightning Network reject that eventually?

Matthew Zietzke: I would say that this is less an issue involving the Lightning Network, rather, it is a general rejection of Nakamoto Consensus by BTC supporters.

Nakamoto Consensus, perhaps oversimplified, is the process of enforcing rules in Bitcoin through the repeated, purposeful choice (often automated) by miners to build on one block as opposed to another.

This process is repeated every time a new block is created and propagated, and the primary factor involved in determining how many blocks a miner creates is the amount of hash power they are applying to finding a nonce (a slight variation on each block,) and resulting hash of the block header that satisfies the difficulty threshold at the time of mining. In short, the miner with more hash power finds more blocks, and the miner with less hash power finds less blocks. If you (and miners who build on your blocks) have greater than 50% of hash power, over time you will be expected to build the longest chain.

Through this repeated choice, miners come to a consensus on the longest chain, and all other chain tips are rejected as part of that longest chain’s history (though not necessarily discarded entirely).

This process is incentivized, as mining a block directs both the coinbase subsidy, which creates new coins, and more importantly, the transaction fees into your control. If your block is rejected by the rest of the miners, then the record of this income is rejected, too. It is as if you never received this income.

Through this process, miners enforce the rules of the Bitcoin system.

As mentioned above, BTC supporters believe that miners (earlier referred to as producers of the chain) are not the only ones involved in consensus over the rules in the system.

They believe that by “running their own node” they can choose whether or not to follow a chain tip produced by miners. If they choose not to follow it, then they reject that chain tip as valid history in the chain, and they instead only follow miners who produce blocks that conform to their preferred rule set.

It is here that BTC supporters are rejecting Nakamoto Consensus, as they believe that many users who “run their own node” may choose to reject blocks produced even by a majority of hash power of miners, thereby keeping “their chain” intact and moving forward with new blocks by incentivizing miners to build on “their chain” with a combination of their transaction fees, the selling of the non-conforming miner’s coins on exchanges, and the purchasing of the conforming miner’s coins on exchanges.

This is a belief in a coordinated boycott of the majority miner chain as a means of “vetoing” miner’s production decisions.

In short, BTC supporters don’t believe miners are the final arbiters of consensus in Bitcoin, rather, users are.

Is your criticism of BTC and especially Lightning Network acknowledged in the BTC sphere?

Matthew Zietzke: Parts of my criticism are certainly acknowledged in the BTC sphere, though they are rarely accepted as valid. Others are not acknowledged at all.

In particular, I mentioned earlier that BTC supporters believe that barriers to entry are the primary factor in determining how “centralized” or “decentralized” the system is. This is a naïve understanding of the economics of barriers to entry and how they relate to things like market structures.

In fact, long term market structures are more heavily swayed not by the barriers to entry, but by the differentiation of products in a given market segment. Said differently, if there are more niches that can be filled in a market, there will tend to be more firms (companies) producing in it over the long run, while less niches result in less firms in that market.

This is visible when looking at any other established market. Global automobile manufacturing is more “decentralized” than SHA256 hashing is, yet it would be absurd to suggest that the barriers to entry into the global automobile market are lower than those required to begin mining Bitcoin on any of the three primary forks sharing the Genesis block history.

Compare this with cola flavored soft drinks, or cream cheese, or x86-64 chips and you’ll see a vast difference regardless of the barriers to entry involved in each market.

As applied to Bitcoin, this understanding leads to the conclusion that BTC’s block size limit actually reduces the number of niches in mining to one, that of hashing efficiency, whereas in BSV, miners must compete over many different factors of block production, including transaction sourcing, block and transaction propagation and validation, as well as any number of ancillary services that follow from production of the blockchain.

In fact, BTC’s limited block sizes lead directly to market consolidation into a monopoly or duopoly, both of which mean that one miner has greater than 50% hash power, whereas in BSV, it is likely that mining will be vertically disintegrated into many firms providing subsets of the block production process.

It is this criticism of mine which leads inevitably into a consensus crisis in BTC, where a single miner has complete control over consensus, and as a result, even the Lightning Network will be crippled, despite beliefs that it can allow users to transact even if an “evil miner” tries to take over their chain.

Why this is so readily dismissed is bewildering to me.

I found this entertaining tweet in your thread:

(Yeah, that's right. I understand LN better than most LN supporters do.)

— Matthew Zietzke (@MZietzke) April 19, 2021

How come that Lightning Network is better understood by you, someone that is full on Bitcoin SV, than by those who actually depend on Lightning Network to function?

Matthew Zietzke: On one hand, I’m psychologically incentivized to know the flaws in the Lightning Network, while they are not. For many, they have no interest in learning what they must learn in order to understand the system, so they stop when they believe they have a “good enough” knowledge on the subject, usually provided by influencers and socially reinforced narratives that “lazer eyes” provide them with unceasing frequency.

On the other hand, why would they listen to someone who is having fun staying poor? Getting rich quick is attractive, and it’s very easy to believe what someone is saying when there are plenty of examples of people winning the lottery by hodling and otherwise following the social conditioning which punishes people for even listening to an alternative voice.

For my own part, I have been involved with Bitcoin for a long time now. The reason why I started with Bitcoin was to use it as a practical example to help me learn monetary theory in economics. I’m literally that nerdy. I want to understand these systems, and place them in a broader context of knowledge that helps me to understand the world. It’s fun for me.

You might say I’m having fun whether I’m poor or not.

If Lightning Network is technically flawed and probably has serious legal implications too, yet Bitcoin SV does offer full scale and low cost transactions already, why don’t BTC supporters simply use Bitcoin SV?

Matthew Zietzke: To put it bluntly, they don’t know any better.

I want to be clear; I’m not insulting them here. Most of these people are just ordinary folks who are trying to find some vehicle with which to increase their wealth, allowing them to escape debt and live a more comfortable life for themselves and with their families.

Most of these people have full time jobs, think they’ve learned something about economics (perhaps because they supported Ron Paul in 2008), and on a superficial level, BTC, ETH, etc. seem to line up with everything they want them to be. Wishful thinking is a powerful incentive.

The problem is, they have entered the mad house, and they’ve gotten wrapped up in the insanity. Tribalism is a powerful drug, and it is being augmented with incredible earnings in paper wealth.

They literally don’t know what Bitcoin is, or why it is so much better than the things they are supporting. And, it sure does look like a poor financial move to learn about these things when they get to tell those of us who have learned to “have fun staying poor.”

Bubble markets, especially those driven by low information speculation like this one, are hard to wake up from until it’s too late, and oftentimes, those who did wake up are hurt just as badly as those who didn’t when all is said and done.

But hurt they will, and they have actively participated in creating the pain that they and many others will experience. There are real consequences to the perpetuation of this bubble, and I am truly sorry for what those consequences will be. I hope people have bought in only with what they can afford to lose.

In your Twitter thread, you kind of concluded with, “What a shame, that these people hate Bitcoin so much, but think they’re supporting it.”

Thanks for your time and efforts, Matthew!