Many people think that they can just wave a magic wand and suddenly their crypto currency project is decentralized. In reality, nothing can be further from the truth because decentralization is almost impossible. Bitcoin might still be the only crypto project to claim this feat, and even then, the proper way to describe it is "Sufficiently Decentralized", as the future is always uncertain. This article will attempt to explain why Bitcoin is sufficiently decentralized and why other crypto projects fall short.
What many people take for granted is the fact that Bitcoin is as much a financial innovation as it is a technological one. Saifedean Ammous' book "The Bitcoin Standard" did an excellent job describing the importance of Bitcoin's monetary properties and why it is the most likely candidate to become the future alternative to central banking. However, the only way it can achieve this is by being ownerless, something its creator, Satoshi Nakamoto, must have realized early on and he was happy to walk away from the project very early. To this day, this remains the biggest factor that led Bitcoin on its path to "Sufficient Decentralization". Even though some other projects over the years try to claim something similar, most of them still have leaders (or small groups of leaders under the umbrella of a foundation) that have a disproportionate say regarding the future of the network and the financial asset that powers it.
While there have been several amazing articles written about the parts that secure Bitcoin's network (two of my favorites are: Bitcoin, Not Blockchain & Bitcoin is Not Backed by Nothing by Parker Lewis) the term "decentralized" tends to show up pretty often without a clear description about what it means to the author. While the term is certainly subjective, we will do our best to explain what makes Bitcoin, sufficiently decentralized.
In Bitcoin's Proof of Work (PoW) model, decentralization comes in three parts where each one is equally important to the ecosystem. In no particular order, they are: core code development, mining and validating Nodes. Let us now examine each of these elements independently, while keeping in mind that even if one of the three parts of the triangle below happens to be considered centralized, it's debatable how much damage would be caused due to Bitcoin's dynamic of checks and balances between the three parts. In order to actually cause damage to Bitcoin's decentralization dynamic, a single entity would need to obtain supermajority control of all three parts of the triangle for an extended amount of time.
Decentralization Of Core Code
Let's begin by looking at the core code as this is the spark that started the crypto movement. (If you have not yet had a chance, Satoshi Nakamoto's White Paper titled Bitcoin: A Peer-to-Peer Electronic Cash System is a must read even for non-technical people). In the beginning all the code was written by Satoshi himself, but over time more developers joined the party. Unfortunately, very little analysis has been done on evaluating all the developer contributions to the Bitcoin code. The data is on Github, and it would be great to see analysis even more detailed than this Bitcoin Codebase Deep Dive published in mid 2019. The report showed that Bitcoin had 49 active contributors as defined by having committed code over the prior three months. They also did a thorough breakdown of all the code and the contributions of the top 20 contributors. Prior to this overview, the image below did an excellent job showing the contributions of the top developers up to 2016, which is prehistoric times in cryptoland.
It can be very challenging to explain how and why the Bitcoin Code is being written in a decentralized way. It is possible, if not likely, that there are one or a few higher skilled cryptographers/engineers dedicated to other crypto projects/networks, as well as possibly a larger number of developers programming on the Ethereum network. However, combination of quality and quantity of Bitcoin developers is unmatched by any other project and/or network. The code review that takes place is incredibly thorough, but we have to be realistic and admit that critical potential issues can still sneak in as we saw with the recent inflation bug discovery by a Bitcoin Cash developer in September 2018. However, it demonstrated that developers from other chains still pay attention to what is happening with Bitcoin under the hood. Throughout Bitcoin's history, it has had this strange dynamic where every time a critical potential bug is fixed, it gives people more confidence that the code is even more fully secure; this is usually reflected in the price slightly rising. In altcoins, this is not the case as with every critical vulnerability, they seem to lose users to competitors, with Bitcoin being the biggest beneficiary.
For a specific example, take a look at this contributor comparison in developer efforts between Bitcoin and Bitcoin Unlimited (BU) during the block-size debate that lead to big blockers forking off the network. We of course have to point out that one of the names on the Bitcoin Cash side is John Carvalho (aka BitcoinErrorLog) who is not a fan of the project to say the least, and is on that list for grammar corrections. In addition, several of those BU devs have moved on to Bitcoin Satoshi Vision (BSV) and with another hard fork looming this November, so more segregation between developers is likely. Needless to say, by the end of 2020, it is possible that all the Bitcoin forks will be down to less than a handful of competent developers.
The Bitcoin Code is also very dynamic as it's just the base for all kinds of layers. Is a developer working only on Lightning considered a Bitcoin Core Developer? A great write up by Lucas Nuzzi titled A Look at Innovation in the Bitcoin's Technology Stack tries to summarize all the development that is currently taking place on top of the Bitcoin base layer. Here is the image he put together and I can't wait to see an update next year.
Bitcoin Core Development is extremely decentralized following the best practices of open source development. While an argument can be made that decentralization efforts suffer due to a large number of engineers being paid by Blockstream and Chaincode Labs, those that are experts in open source development do not see this as an issue based on how these companies operate. In addition, several top developers this year alone have been independently sponsored by companies in the crypto industry and this trend is more likely to continue. Bitcoin Core Development continues to set the gold standard for sufficiently decentralized projects where it's all about the scientific contributions and not the status of your name. As stated perfectly by Lucas Nuzzi, "...innovation in Bitcoin requires creativity, patience, and perhaps most importantly, ego-minimization" and those who can't handle the ego-minimization requirement, tend to rage quit and start their own Blockchains.
All this effort to keep Bitcoin Core code sufficiently decentralized is not only important to attract the best talent to advance Bitcoin, but also to prevent malicious spying technology, back doors or anything else that could hinder the efforts of making Bitcoin an unstoppable financial force for decades to come. This was a big deal in 2015 when Mike Hearn attempted to push out a code upgrade that contained an "unmentioned addition which periodically downloads lists of Tor IP addresses for blacklisting." This is something altcoins can't compete with. How do you get the right eyes to voluntarily look at the code? Just prior to publication of this article, we had another example of malicious code being added to an altcoin that went unnoticed for years.
These developers need to listen not only to Node operators and Miners, but also to the overall user base before making any critical changes to the code. These changes MUST be made with consideration on whether it has the potential to increase or decrease future decentralization. For example, had the Core Developers pushed for increasing the block-size during the 2017 block-size debates, it would have probably split the developer community in half, given the dominant miners a slight advantage on finding future blocks (as depicted in Table 2, Page 4 of this BitFury Block Size Increase Research Report), and over the long run, made it more expensive to fire up a Bitcoin Full Node due to the cumulative size of the Bitcoin Blockchain increasing at a faster rate (more on that later).
Thanks to all the current efforts up to this point, it should be fair to say that Bitcoin's Core Development remains sufficiently decentralized.
Decentralization Of Mining
Nouriel Roubini loves his 2015 talking point, stating that Bitcoin is centralized because all the mining is done in China. Debating Nouriel on this point is very frustrating because like many people, he does not realize that virtually all Bitcoin Mining is done in pools. While many pool operators are China-based, the miners inside these pools are global and it is fairly trivial to switch between pools. Because this is not a research paper on the Bitcoin Mining ecosystem, you are encouraged to read the following for a better picture on the geographical distributions and costs to mine a bitcoin by BitOoda.
We have all seen our share of abuse by pool operators misusing the hashrate of their miners in more centralized Blockchains (See: Court Case), but this has never been an issue for Bitcoin due to the competitive nature of pool operators to attract mining operations. The Bitcoin community is also taking critical steps to give the miners inside a pool significantly more voice as to what transaction gets mined once a block is found. It is described here in Bitcoin's Decentralization with Stratum V2 by Braiins but the short version of it is simple; here is what the Bitcoin Mining distribution would look like if every pool adopted it:
Until Stratum V2 is the standard within mining pools, a 51% attack is still of some concern. For those that may not know, if a miner or a mining pool acquires 51% of the hashrate, they have the ability to double spend coins. We have seen this happen many times on significantly more centralized Blockchains with the latest incident being Ethereum Classic. But just because a 51% attack is profitable on altcoins, and theoretically possible to do in Bitcoin, people severely underestimate the Game Theory aspect of how economically suicidal it would be for a miner or a pool operator to even hint at the attempt to cheat. In fact, the last time a mining pool briefly acquired such power, it was down to only 3% of the network within six months and closed operations about a year after that.
Achieving this 51% dominance in Bitcoin is extremely challenging because of the enormous costs involved. The massive amount of energy (hashrate) Bitcoin consumes globally is very critical to its security. Since mining is one third of the Decentralization Triangle, it's important to point out that both, the electricity input costs as well the composition of the mining structure are equally critical to gauge the censorship resistance of the digital asset. Ethereum for example does have a very high energy input and its mining distribution is pretty spread out, unlike most of the other altcoins, and yet the leaders of the coin continue to push for the removal of the one part of the triangle that actually has the highest claim to keep it decentralized with a move to Proof of Stake (more on that later and lots of PoS references at the end). Due to this continued initiative, it is very likely that there is little Research and Development (R&D) investment entering the Ethereum mining ecosystem. Other than Ethereum, the security of your cryptocurrency starts to drop off a cliff with Litecoin 25 times less secure, Bitcoin Cash (aka BCash) 50 times less secure, Monero 100 times less secure, and so on. Just like with Ethereum Classic, as long as you can move your value to a more secure Blockchain like Bitcoin, there will always be risk of a 51% attack. And since these attacks usually target exchanges, we should be expecting them to delist these vulnerable coins, otherwise, they need to do a better job making sure the double spend risk is low.
Now that you got a glimpse of how unsecure all these other Proof of Work coins are based on energy input and R&D to build the most efficient ASICS, the image below shows some of them from their composition side to see if any can claim their mining is sufficiently decentralized. Litecoin and Ethereum seem to be in best shape, but as mentioned previously Ethereum developers are on a mission to get rid of mining. The status of ZCash, Dash, Bitcoin SV and the soon to hard-fork Bitcoin Cash speak for themselves, as it should not take much to coordinate a 51% attack of the network. The best way to keep mining decentralized is to give the industry incentive to innovate. Something that Monero developers refuse to do by hard forking every six months. It's almost like setting off an EMP every time someone tries to innovate a car to make sure we all continue to use horses, but I digress. The best incentive to innovate is to provide confidence that more value will be secured by the network in the future, and hence there is interest in pouring money into mining R&D. If you are holding crypto that isn't Bitcoin, take a long hard look at how their Mining is evolving, is it towards Security & Decentralization or away from it?
So overall, the current Bitcoin Mining structure is not ideal, but things are definitely expected to improve. In my non-professional opinion, I still believe mining is sufficiently decentralized or at worst, significantly more decentralized with an enormous defense shield against attacks than any other Blockchain. If your Blockchain does not have mining, then you will never be able to reach Bitcoin's overall decentralization levels, but more on that later.
Decentralization Of Nodes
The final piece of the Decentralization puzzle is Nodes. Nodes are unique because unlike the prior two parts discussed above, which require a major intellectual barrier to entry (Coding) or a huge financial barrier to entry (Mining), running a full Node requires moderate technical expertise, and a very low financial investment. Nodes are incredibly important because they enforce the rules that coders release in order to keep the miners in check, so they don't cheat. The most important function they serve is probably making sure that there will never be more than 21 Million bitcoins. Nodes are incredibly powerful as was proven during the 2017 UASF, as the 1 mb block limit was enforced by Bitcoin power users going against the wishes of most miners and 'economic Nodes'.
Speaking of economic Nodes, it is probably worth knowing the differences between full Nodes, lightweight Nodes and pruned Nodes (also read about it here). Economic Nodes are nothing special, it's just a term used to refer to large crypto companies that often think they are more important because their Node speaks for thousands (if not 10's of thousands) of people. In reality however, your Node protecting and validating your transactions of a few hundred dollars is just as important to the decentralized nature of Bitcoin as Coinbase's Node which is responsible for over a Billion dollars' worth of value. (Yes, Coinbase is probably running more than one Node, but so can you). Many will also go as far as to claim Nodes are the most important part of Bitcoin's Decentralization, but we will just acknowledge that the more validating Nodes exist around the world, the more sufficiently decentralized our one-third of the triangle is.
The picture above is usually used to represent full Nodes around the world but according to Luke Dashjr, it only shows you the amount of "listening Nodes" while the actual number of full Nodes in the world is much higher. Luke keeps track of these full Nodes on his own site where at the time of writing he has about 45,000 Nodes with 5,000 of them running the latest core client. While there are good arguments to be made that updating your Node to the latest version right away could pose some risks in case that version turns out to have a bug that was missed, running a very outdated version probably has bigger risks. The larger concern however, is that looking at the same picture from 18 months ago, it clearly shows how the number of Nodes has dropped by over 50% while the number of latest up to date Nodes (Green) has dropped by 80%.
This is a pretty scary trend and perhaps it could be a call to readers to consider taking this opportunity to help protect Bitcoin's ecosystem by running your own full Node (directions here). However, it is strongly recommended to stay up to date on the latest updates and understand why you should run a specific version of the software, or upgrade to the latest soft fork. While the pictures above are presenting evidence of significantly lower Node counts, the reality of Node decentralization might not be that dire, as many people have started running Nodes through Onion Routing like Tor. To make this process simple for users, startups like Start 9 Labs have created an affordable solution to those who want to run a full Node while staying anonymous on the network, to where these Node information services don't even know how many Nodes there actually are.
Luke Dashjr is also concerned about the cumulative size of the Bitcoin Blockchain growing too fast. The 2017 SegWit soft fork, which Luke helped implement, was a block size increase in addition to the malleability fix needed for the Lightning Network scaling solution. Bitcoin's Blockchain is now at almost 300 Gigs which is a 50% increase since the start of 2019. While these Node counts are certainly trending in the wrong directions, they remain sufficiently decentralized and head and shoulders above all other Blockchains. There are no stats on how many Full Archive Ethereum Nodes there are, but at a size of over 4 Terabytes in April 2020, before the DeFi transactions wave, it's hard to imagine that in a few years ETH will have more than a dozen Nodes. Some of the academic research that was done in 2017 to prove Ethereum is more decentralized than Bitcoin is already starting to look embarrassing (Full Paper). As for Bitcoin's Hard Forks, there is so little economic use of those chains that they are barely noticeable when compared to the number of Bitcoin Nodes. In addition, currencies like BCH and BSV, will have an impossible task to maintain large Node counts, given their obsession for bigger blocks. Here is a video of me at Scaling Bitcoin (Stanford 2017) asking how I can run a Node on an old laptop and the answer is that you will not be able to.
What about the majority of the bitcoins being held by a few people?
Many people love to demonize Bitcoin because a few people happen to have invested wisely in the early days and now hold a significant amount of wealth in bitcoins. Besides the fact that these statistics are very biased, as many large accounts are custodial holding bitcoins on behalf of many clients, the beauty of the system is that having a large amount of bitcoins does not give you any advantages on how the system functions. This is not true for many other Blockchains. Those with a Proof of Stake system of consensus have made a decision that the early adopters and those who are already wealthy in that coin, should have all the advantages in decision making and gain future wealth by staking those coins, earning even more coins. Please see the reference section for lots of resources and research dedicated to explaining the insecurity of Proof of Stake. Contrary to PoS (or DPoS) promoters, their systems can't possibly be as decentralized as Bitcoin's because of the lack of mining, centralization of coins held, and most likely a single dictator in charge of the code. All of which leads to only a handful of Nodes being relevant with the most notable Node belonging to the lead developer. Hence any Blockchain that hands over decentralization power to the most wealthy is by default an inferior and more centralized system.
Taking a look at some popular cryptos through the lens of decentralization:
- Ethereum (ETH): While the Mining structure is still pretty decentralized, the completely centralized development structure has been promising to remove the one part of the project that actually has decentralization. Once mining is eliminated (with a hard fork that will most likely create an Ethereum Work ETW chain) the system will depend on a socialist style redistribution of wealth to prevent large concentrations of ETH in few hands having too much power. This of course will not apply to the early holders of Ethereum just like any other oligarchy (We have already seen this with the rollback in The DAO hack). As for the Nodes, please try and set up an Archival Full Node and if you can pull that off, try to maintain it for a year validating your Ethereum Translations... good luck!
- Ripple (XRP): This is easy; since there is no mining, it relies on two factors, Development & Nodes. Since development is 100% controlled by Ripple Labs, the only Node that matters is the Ripple Labs Node and it does not matter how many other Nodes exist, they are irrelevant to decentralizing the network. XRP is probably the most centralized project still pretending to be decentralized.
- Bitcoin Cash aka "BCash" (BCH): While BCash has an identical structure to Bitcoin, each of the three elements of decentralization are significantly less decentralized than in Bitcoin. In addition to the same entity (small group of people) having significant control of all three parts that make the network decentralized, they also own the most popular BCH media source, an exchange that prioritizes BCH trading, and a large concentration of coins (though technically this does not make them more or less decentralized). We have seen several examples over the years (like adding checkpoints) to prove the project is not sufficiently decentralized.
- Bitcoin SV (BSV): Same as Bitcoin Cash, but even more centralized across all three parts, as it trends towards more centralization every month with their giant 'empty' blocks.
- Litecoin (LTC): Mining is still somewhat decentralized, but if the price can't keep up with the supply reduction every four years, it can get centralized very quickly. It is very unlikely there are a lot of people running full Nodes, and while the coding may seem decentralized due to Bitcoin developers using some of these projects as Testnets, there are probably very few (if any) pure Litcoin devs. People can argue that Litecoin might still be sufficiently decentralized, but it is way more centralized than Bitcoin.
- Monero (XMR): Similar to Litecoin, but it should be pretty clear that development is way more centralized due to an ability to push out a hard-fork at will. This also means that very few are running full Nodes. Sorry, did not look into research on mining distribution, if the stats are even available. Yes, the transactions are more private and that part will be hard to alter even if the project becomes fully centralized, but it's hard to argue that there is all that much decentralization. The token is almost out of the Top 20 by market cap, hence interest is dropping fast. If you are a supporter of Monero, just ask yourself: Who is developing the code? Who is running the Nodes? And how decentralized is mining not to mention the hashrate strength?
- Other PoW / Hybrid Coins (Ethereum Classic ETC, Dash, ZCash, etc): The Development is completely Centralized, I doubt users are running Nodes, mining is very weak and hence easy to 51% attack as we saw with ETC recently. In the case of Dash, there is additional centralization and control due to Masternodes (A scheme to literally enrich early adopters and give them more control over the protocol).
- Ethereum Competitors (Cardano ACA, EOS, TRON, Tezos, Waves, Neo, OMG etc) - They are all pretty much the same. By not having mining and using PoS, it is impossible for them to be more decentralized than Bitcoin, as a third of Max Decentralization is missing. They have a centralized development team that calls all the shots with the ability to hard-fork the protocol to their desired rules at any time. This is enforced by those with early access to tokens, which are almost always the creators of the protocol. (Bitcoin is truly special because no matter how many bitcoins Satoshi mined, him holding them gives him 0 power over Bitcoin's decentralization). Because of everything that was just stated, the amount of Nodes people have around the world is irrelevant, it could be 100 times more Nodes than the Bitcoin network, but the power will be fully maintained by the Nodes of protocol developers and early token holders. The protocol leaders will try and convince the masses that these groups are independent, but up to now they have always been one and the same. Perhaps Ethereum has the best argument that their tokens are globally spread out as they make the push for PoS, but that's it, that's the list. As centralized as Ethereum is, it is way more decentralized than any of these projects. Since most legitimate projects (and I use the phrase very loosely) building on top of Ethereum don't actually need a decentralized base layer, it is possible for Ethereum Competitors to take market share from Ethereum, but it won't be because they are more 'decentralized'.
- ICO's (Crypto.com CRO, Binance BNB, Chainlink LINK, etc): These are just security scams pretending to be Utility. The more honest ones will be the first to admit there is no decentralization (ex: Binance even though they are convincing you they are building a decentralized platform for the future), while the scammier ones will pretend they are decentralized (ex: Chainlink, Maker Dai)
REFERENCES AND READING LIST
Bitcoin, Money & Economics
Bitcoin: A Peer-to-Peer Electronic Cash System - Satoshi Nakamoto, Oct 31, 2008
The Bitcoin Standard: The Decentralized Alternative to Central Banking - Saifedean Ammous, Apr 24, 2018
Shelling Out: The Origins of Money - Nick Szabo, 2002 (Research Paper)
Core Code
Bitcoin Codebase Deep Dive - CodeAnalysis, Aug 5, 2019
A Look at Innovation in Bitcoin's Technology Stack - Lucas Nuzzi, Dec 3, 2019
Mining
Bitcoin's Decentralization with Strutum V2 - Braiins, Jun 23, 2020
Bitcoin Mining Hashrate and Power Analysis - BitOoda Research, Jul 15, 2020
Digesting 'Quantification of Energy and Carbon Costs of Mining Cryptocurrencies' - Nic Carter, Nov 13, 2018
Bitcoin Block Details by CoinDance
Nodes
Money is a Social Construct and that's Why You Should Run a Bitcoin Full Node - Brand7, Nov 2, 2017
Difference between Hard Fork, Soft Fork and Chain Split - John Light, Sep 25, 2017
Running your Node by Bitcoin.org
Node Count by Luke Dashjr
Listening Node Count by Bitnodes
Node Count by CoinDance
Unchained Capital (Must Reads)
Why It's Hard to 'Get' Bitcoin: The Blockchain Spectrum - Dhruv Bansal, Dec 1, 2017
Bitcoin Can't Be Copied - Parker Lewis, Aug 2, 2019
Bitcoin Does Not Waste Energy - Parker Lewis, Aug 16, 2019
Bitcoin, Not Blockchain - Parker Lewis, Sep 6, 2019
Bitcoin is Not Backed by Nothing - Parker Lewis, Sep 27, 2019
Bitcoin is Not a Pyramid Scheme - Parker Lewis, Oct 18, 2019
Bitcoin Obsoletes All Other Money - Parker Lewis, Jan 24, 2020
21 Million is Non-Negotiable - Phil Geiger, April 9, 2020
Bitcoin is Common Sense - Parker Lewis, May 1, 2020
Bitcoin is One for All - Parker Lewis, Aug 27, 2020
Jimmy Song's (Must Reads)
Why Bitcoin is Different - Apr 2, 2018
Why Bitcoin Works - Apr 17, 2018
Why Hard Forks are Altcoins - May 1, 2018
Why Blockchain is Hard - May 14, 2018
Why Blockchain is Not the Answer - May 7, 2019
Why Cryptocurrency Diversification Makes Little Sense - May 20, 2019
On Altcoin Valuation - Nov 14, 2019
Why Exchanges Delist Coins - Dec 20, 2019
Nic Carter (Must Read)
Is Bitcoin Antifragile - Jul 23, 2017
Unpacking Bitcoin's Assurances - Jan 13, 2019
It's the Settlement Assurances, Stupid - Jul 22, 2019
A Most Peaceful Revolution - Sep 7, 2019
Proof of Work (PoW) vs Proof of Stake (PoS)
Distributed Consensus from Proof of Stake is Impossible - Andrew Poelstra, May 28, 2014 (Research Paper)
On Stake and Consensus - Andrew Poelstra, Mar 3, 2015 (Research Paper)
Proof of Stake versus Proof of Work - Bitfury Group, Sep 13, 2015 (Research Paper)
What's Wrong with Proof of Stake - Bob McElrath, Jun 14, 2016
Vulnerability: Proof of Work vs Proof of Stake - Robert Greenfield IV, Aug 24, 2017
Understanding Proof of Stake through its Flaws [Part II - Nothing at Steak] [Part II - Long Range Attacks] - Abhishek Sharma, Jan 15, 2018
Proof-of-Stake & the Wrong Engineering Mindset - Hugo Nguyen, Mar 18, 2018
Proof-of-Stake, Private Keys Attacks and Unforgeable Costliness the Unsung Hero - Hugo Nguyen, Apr 3, 2018
PoW is Efficient - Dan Held, Sep 14, 2018
Work is Timeless, Stake is Not - Hugo Nguyen, Oct 12, 2018
"Fake Stake" Attacks on Chain-based Proof-of-Stake Cryptocurrencies - Decentralized Systems Lab, Jan 22, 2019
KRNC: New Foundations for Permissionless Byzantine Consensus and Global Monetary Stability - Clinton Ehrlich & Anna Guzova, Nov 14, 2019 (Research Paper)
Blocksize Too Big, Too Small, or Just Right
Blocksize Increase Research - BitFury Group, Sep 6, 2015 (Research Paper)
To Reduce or Not to Reduce the Block Size - Matt, Feb 12, 2019
The Pros, Cons, and Taboo Nature of Enforcing Smaller Bitcoin Blocks - David Canellis, Feb 13, 2019
Other
The #Bitcoin's Balance of Power - Brand7, Feb 26, 2017
Quantifying Decentralization - Balaji S. Srinivasan, Jul 27, 2017
The Ethereum-Blockchain Size has Exceeded 1TB and Yes, it's an Issue - Stop And Decrypt, May 23, 2018
Sharding centralizes Ethereum by Selling You Scaling-In Disguised as Scaling-Out - Stop And Decrypt, Jun 7, 2018
Ether and Bitcoin are Not the Same - Leah Wald & Steven McClurg, Jun 11, 2020