2023-02-19 08:34:16 UTC


If 2021 was the year of the crypto investor, 2022 was the year of the crypto developer.
The Ethereum Merge was a huge milestone for crypto: a $200 billion network porting over to an entirely new, scalable transaction ledger with a brand new security model, in real time, with zero hiccups. It was a spectacular technical feat, even if it ended up being boring to watch.
That it went off without any major hitches bordered on miraculous, and I will be the first to admit that it was completed much sooner and with less drama than I thought it would be. Vitalik and the Ethereum Core team deserve major kudos. They aren't the only ones who have been shipping, though.
The alternative Layer-1 ecosystems have been evolving at warp speed. We highlighted Solana's community traction a couple of chapters ago. Optimistic and ZK-rollup ecosystems have been relentless in onboarding new applications. Cosmos has proven to be a real competitor as it siphoned off more "appchains" to its ecosystem and proposed an overhaul to the economics of its ATOM token. Meta spinout, Aptos, launched its mainnet.
If security and scalability is what's required to accelerate adoption and make room for new killer applications, then the L1 developers killed it this year.
That's not to say L1 development has been all sunshine and rainbows. Upgrades take time and patience, and core developers are scrambling to solve MEV issues, reduce platform downtimes, and ensure their blockchains do not become censorable or damaged by centralization risks.
But the future looks bright.
*Note: this stuff is highly technical. I'm working on directional correctness and concision, and allowing for some imprecision for the sake of human readability.

The New Ethereum Roadmap

The Merge was a major technical update six years in the making. Now that it's been successfully deployed, laypeople might not appreciate just how much is still left to build on the Ethereum core team's aggressive roadmap. Many of these subsequent releases will come faster (and can be tackled more iteratively) than the Merge. In a single tweet, Vitalik laid out what's next:

Let's walk through this all in plain English:
  • The Merge: Part one of The Merge forked out Proof-of-Work consensus and switched to Proof-of- Stake (the big milestone this fall – Why Proof-of-Stake). Now developers are working on ensuring transaction validation is well distributed, allowing stakers to withdraw from the staking contracts (est. March 2023), and tackling more wonky items like improving signature aggregation.
  • The Surge: The goal is to reach 100,000 transactions/second by introducing a new Ethereum transaction type called "blobs," where rollups will have a specific allocation of blockspace to post their data. The initial form of blobs will be delivered in EIP-4844 or "Proto-Danksharding." In its full form, "Danksharding,"fees on both L1 and L2 should be reduced while tackling Ethereum's version of Data Availability Sampling.
  • The Scourge: This is a new phase added recently (Vitalik's tweet), largely in response to community concerns over risks that maximum extractable value ("MEV") was creating potential transaction censorship (more on this in two sections). It also includes full Danksharding of blob transactions. That's a real sentence I just wrote.
  • The Verge: A "fully SNARKed Ethereum" makes verifying blocks easier and introduces something called Verkle trees, cryptographic proofs that are smaller than their Merkle tree cousins, paving the way for stateless clients (better mobile support!) on Ethereum.
  • The Purge: Lots of little code clean-ups that simplify the Ethereum protocol, eliminate technical debt and limit costs of participating in the network. This will include state expiry in EIP-4444. Again, things that reduce costs and improve performance.
  • The Splurge: What Vitalik calls "fix everything else." These are EVM improvements, account abstractions, and the creation of a "quantum-safe" Ethereum.
If you want to nerd out and get deeper into the weeds on what all of this means, Vitalik did a two hour episode with the Bankless gentlemen on why each bucket is important to the Ethereum roadmap, and an encore episode on what to watch for in 2023. Otherwise, you can take my word for it that the next couple of years are about cleaning up tech debt, nailing the scaling and security of rollups, and ensuring the EVM stays censorship resistant.

Merge Economics

The Merge marked a fundamental shift in the economic model of Ethereum. The switch to Proof-of- Stake yielded a protocol whose environmental footprint was reduced in size by ~99%, making Ethereum more compelling as an investment to environmental, social, and corporate governance (ESG)-minded institutions. It also reduced new token issuance by 90% and eliminated nearly $500 million of monthly sell pressure from miners. Finally, it created a net deflationary asset with real yield as a result of the fee burn mechanism Ethereum implemented in August 2021 in EIP-1559.

Since EIP-1559 went live last summer, the Ethereum protocol has "burned" about 85% of all transaction fees, with the remaining 15% triaged to miners as "tips." Ether's supply will become net deflationary (tantamount to programmatic stock buybacks) if burnt transaction fees exceed the network's staking issuance rate. We think the network could have steady-state deflation of 1–2% per year depending on the demand for blockspace. No other project in crypto has achieved these supply dynamics.
Depending on the number of active stakers on the network as well as the level of network activity, yields could range from 5-7% in 2023, setting a sort of "risk-free rate" for Ethereum's financial system. Some DeFi protocols are already starting to build out this yield curve for their investors to monitor. Headline yields today are a bit higher, though we expect those will decline and normalize over time, as staking becomes easier and derivatives like Lido's staked ETH (stETH) become ubiquitous.


(Required Reading: State of Ethereum Q3 2022, The Investor's Guide to the Merge: Understandingand Playing the Opportunity, What Will Ethereum Miners do After the Merge?, How Traders are Positioning for the Merge, Staking Done Well)

MEV & Censorship

Maximum Extractable Value (formerly Miner Extractable Value), or "MEV", is one of the most fascinating technical challenges in all of crypto, and it's attracted some of the industry's top technical and financial minds to solve its riddles.
In short, MEV is a byproduct of the power dynamics that exist between a blockchain's security providers (miners / validators) and the network's users. Since the security providers (or "sequencers" in the case of a rollup) are responsible for determining the transaction order and inclusion of transactions per block, they can impose a tax on users.
This tax, MEV, exists on every blockchain, so it can be thought of as a feature, not a bug, in some ways. MEV can arguably make protocols more liquid, efficient, and antifragile. A healthy MEV sub-economy would redistribute most MEV rewards to staking tokenholders, and ensure that the transaction processing "supply chain" separates block proposers (anyone) and block builders (specialists) in a global, decentralized block building competition.

A block proposer is the same thing as a transaction validator. It's simply a node/person/entity that creates a block of transactions and propagates it to the network for inclusion in a blockchain. Pre-Merge, these proposers were the miners. Post-Merge, proposers are ETH stakers, validators, that are programmatically selected by the protocol to propose the next block on the chain. This sounds like a lot of work, but historically, a lot of proposers have simply created new blocks based on gas fee and sent them to other nodes in the network.
Under proposer-builder separation, or "PBS," block builders are disaggregated from proposers. The builders create "exec block bodies" that order the list of block transactions and submit them to proposers along with a fee. The proposers (validators) are incentivized to simply accept the exec block body with the highest bid. This helps to specialize block building (where it pays to specialize in efficient block bundling), without jeopardizing network decentralization.
Obviously, there are some negative consequences of this. Block builders may try to extract value from the network through asymmetric knowledge of users' transaction instructions (e.g., front-running a trade) and fleece other network stakeholders with an aggressive network "tax" rather than redistribute some of the rewards across the transaction processing supply chain. And some transactions could even be censored outright if block builders collude or blacklist certain addresses.
One of the leading MEV R&D shops, Flashbots, proposed a framework to categorize types of MEV based on their value accrual and possible externalities. If we can design systems capable of controlling for certain types of "bad" MEV, we can (hopefully) create more equitable protocols in the future that socialize MEV's gains.

In fact, censorship is the primary post-Merge concern with Ethereum right now.
When the U.S. Treasury Department sanctioned Tornado Cash, many post-Merge validators (who had become more recognizable and sophisticated in a MEV-rich post-Merge environment) began to censor certain addresses and transaction types. Large custodians, exchanges, cloud staking services (like Alchemy), and even decentralized RPC relayers (like Pokt) became OFAC-compliant, basically overnight. OFAC-compliant blocks hit nearly 80% of network validation in mid-November, though that is (thankfully) back down to 65%.

(Source: MEV Watch)
With a new block produced every 12 seconds, this means that "censored" transactions will be posted in under a minute, on average. Having to wait an extra ~36 seconds isn't the apocalypse, but it does strike many in the Ethereum community as a bad precedent.
If more types of transactions are deemed sanctionable and subsequently censored by producers, there will be more competition to cram "non-compliant" transactions into blocks produced by other validators. Worse, there is some degree of concern that financial authorities like the U.S. Treasury will soon see that they had a direct (and credible) route to enforce compliance requirements over entire blockchains by forcing large U.S. custodians to make a choice between transaction censorship or non-participation in staking economies (not really viable).*
Ethereum devs need time to work their magic on "The Scourge" part of the roadmap and solve the base layer censorship problem before regulators somewhere do something ill-considered and technically impossible like demand entire chains are OFAC-compliant.
In the meantime, major custodians and exchanges will make their own determinations on behalf of customers as to whether they will censor transactions and only leverage software and network infrastructure that complies with OFAC sanctions. But we should keep an even closer eye on what wallets (e.g., MetaMask) and decentralized node operators choose to offer their users by default. When it comes to public policy, it's helpful to have that OFAC-compliance ratio around 70% for now.
(*Yes, I'm aware of slashing, but I assume Coinbase isn't ever going to get slashed, so this will be handled "out of court" as they say.)
(Required Reading: The Book of MEV: Current Landscape, Future Solutions, and Key Considerations, Ethereum's Compliance Concerns are Greatly Exaggerated)

What Vitalik's Excited About

What's Big V excited about this year?
  1.  Privacy-preserving and real-world asset backed stablecoins (this happens to map to my thoughts from last chapter, but I promise I didn't peek when I was writing mine.)
  2. Within DeFi, Vitalik likes prediction markets. I think prediction markets are dumpster fires and no one has yet figured out how to make them interesting. This could be a contrarian bet that pays off big in the future if I'm wrong, but even V acknowledges that he doesn't expect them to make "multibillion-dollar splashes," so maybe the interest is academic.
  3.  Identity modules around things like authentication (sign-in with Ethereum), names (ENS), attestations (decentralized social), proof-of-humanity (Worldcoin? jk), etc.
  4. DAOs, which he breaks down into communities that are decentralized either for robustness, efficiency, or interoperability.
  5. Hybrid applications that use blockchain and non-blockchain systems like voting, "auditable centralized services," etc.
If you're a builder, I'd read through this carefully, and pick an underdeveloped area that Vitalik likes. It seems that every single application envisioned in the Ethereum whitepaper is now enormous, and you can draw a straight line from this 2017 post on market makers to what happened with Uniswap's invention in the last bear market, and what is now the backbone of DeFi and a $5 billion protocol whose volumes rival Coinbase.
(Required Reading: EVM-Chains in the Bear Market)

Exploding Bridges

I wrote in last year's report that blockchain interoperability and bridging protocols would be one of the top three most important emerging areas of development to watch.
"All of these new blockchains (plus Ethereum's Layer 2 rollups) will need to talk to each other, so the most acute pain point in crypto today may be the lack of bridges. If the future is multi-chain, then those who build better cross-chain connectors and help move assets fluidly across parachains, zones, and rollups will inherit the (virtual) earth."
I didn't say it would be easy! In fact, that prediction aged (catastrophically) well.
Axie's Ronin "sidechain" was hacked for $600 million. Wormhole's ETH-SOL bridge lost $320 million. Nomad lost $200 million. Today's bridges have proven to be, shall we say, unstable.
Rollups are a bit different. They're basically just blockchains with built-in bridges for transferring value and settling transactions across EVM chains. Rollups, as their name implies, are blockchains that process their own transactions, but "roll up" to Ethereum for settlement and leverage Ethereum's robust security.


As an individual user, you don't really care if the button you push on a front-end service is routing your crypto transaction over a bridge to an EVM compatible "alt-L1"  like Avalanche or an "appchain" built on Cosmos, or whether it's processed on a rollup chain like Arbitrum.
What you do care about is that your assets are secure when in transit, and that they are there on-chain when you want them back.
High net worth investors, corporates, and governments certainly care about security. If a rollup suffers an outage or vulnerability you can always claim funds on its associated parent L1. If a Layer-1 blockchain has an outage or significant performance issue, you're out of luck until things are back up and running. When everything is working, there's no difference between rollups, appchains, and alt-L1s.
But the security issues can get ya. Bad.
It's all pretty similar to the global banking system, actually: you don't really care how your neighborhood bank routes your payment through the banking network. You just don't want your money lost in transit, and you want your savings to be secure and insured. Individual bank accounts are FDIC insured up to $250,000, but bigger depositors face more risk.
Vitalik expounded on his thoughts around why the future of *cross-chain* bridges isn't that bright. There are "fundamental limits to the security of bridges that hop across multiple 'zones of sovereignty'" and the modular blockchains thesis where "you can't just pick and choose a separate data layer and security layer. Your data layer must be your security layer."
If that's the case, it reinforces my hunch that the L1 "blockchain wars" will look similar to the browser wars and mobile OS wars. That is, the EVM and one or two others might win at scale, but we won't see dozens of L1 blockchains proliferate.
Let's drill down the different types of rollups and the "modular blockchain" thesis, then talk about today's two largest non-EVM ecosystems, Solana and Cosmos.

Rollups and Modularity

Rollups improve blockchain scalability by executing transactions on separate blockchains, then posting compressed transaction data to an underlying L1. Traditional rollups that rely on Ethereum L1 for settlement, consensus, and data availability were first to market with Arbitrum and Optimism Layer-2s launching in 2021.
The new kid on the block, Celestia, ushered in the "modular" thesis which became all the rage this year. While traditional rollups execute transactions and rely on Ethereum to settle, verify, and store their data, modularity allows developers to mix and match how their protocols handle each step. We'll touch on modularity a bit more below, but first let's dig into the rollups to watch this coming year.

Smart Contract Rollups: A smart contract on L1 verifies batches of compressed transactions using different flavors of "proofs." They are "enshrined" to the L1 as they rely on the smart contract for final settlement (security). The two main types we have today are:
  • Optimistic Rollups: Optimistic rollups rely on "fraud proofs." They are "optimistic" as batches are assumed to be valid until a fraud proof submitted by "verifiers" indicates otherwise (within a set challenge period). Currently, Arbitrum and Optimism are the two largest optimistic rollups, and  they account for ~80% of the $4.3 billion in TVL on L2 blockchains today (not counting Polygon's PoS chain)
  • Zero-Knowledge Rollups (ZK Rollups): While similar to optimistic rollups, ZK Rollups rely on "validity proofs" to verify transactions. Each batch posted to L1 includes a validity proof that the smart contract then verifies, and hence, a long arbitration period isn't required. In contrast to their optimistic counterparts, it's much more technically complex to achieve EVM-compatibility for the zk bunch. Polygon, zkSync, Scroll, StarkWare, Loopring (Taiko), and ConsenSys have all announced plans to launch zkEVM rollups "soon."
  • Many developers are more excited about the long-term potential of zkEVM implementations (and I predicted in last year's report that they would soon dominate), but it's really going to come down to usability. Optimistic rollups may have the edge for the foreseeable future due to the first-mover advantage. ZK Rollups may proliferate if and when their scalability benefits become more obvious as they *theoretically* promise cheaper transaction fees for end users.
"Sovereign" Rollups: In the same way that you can deploy a virtual server with one click from a cloud provider, developers can spin up sovereign blockchains with negligible upfront costs and eschew the burden of bootstrapping a distributed validator set. Sovereign rollups execute and validate transactions and post data to a consensus and data availability layer like Celestia, Polygon Avail, or Ethereum (when data availability sampling is achieved).
The key difference from smart contract rollups is that they are "sovereign" in their decision making over the validity of transactions and the rollup's state. They can be optimistic, or zk-based, or whatever their communities' desire. Rollup nodes execute and verify transactions according to their own rules and proofs and do not rely on the opinion of an underlying smart contract layer for final settlement. They merely use the base layer to store data and refer to the order of past blocks to compute the next. The use of a separate settlement layer is optional.
The universe of possible modular configurations can make your head spin. Consider the FuelVM, a modular execution layer with parallel transaction processing.

(Source: Fuel)
Fuel Labs is building an optimistic rollup that settles to L1 Ethereum, but the FuelVM itself is configurable with different settlement, consensus, and data availability solutions. Paired with Celestia, it's a Sovereign Rollup. When using an independent settlement layer like Cevmos, it becomes a "Settlement Rollup."
Easy, right?
I have rewritten this for concision and accuracy multiple times, and it's impossible.
(It's directionally correct, though. I will leave it to the Messari analysts to parse the nuances – and offer more depth – in their excellent ongoing research in the new year. In the meantime, keep in mind that crypto scalability hinges on the "modular" and "monolithic" designs.)
Value accrual on rollups and modular networks is questionable because it is unclear how much economic value will actually flow to the consensus and data availability layers versus transaction settlement and execution.
Some DeFi applications in particular will have to monitor whether liquidity ultimately gets fragmented between rollups thanks to insecure cross-chain infrastructure. Expect some novel tools and protocols in this space (we still need bridges!) as we continue to march towards a multi-rollup world with a decreased dependence on Ethereum L1 and a heightened search for cheaper transactions and data availability.
Rollups are a boon for application development (better interoperability and user tradeoffs, better throughput and lower fees), but the jury is still out on whether they will compete effectively with alternative Layer-1s like Solana and Avalanche. Traditional rollups get security benefits (and trustless bridge benefits) from their ties to Ethereum, but transaction costs are still an order of magnitude higher than many alt-L1s.
That leaves an opening for a couple of big, if beleaguered, competitors.
(Required Reading: The ZK Everything Report, How zkEVMs Can Keep Ethereum Competitive withAlternative Layer-1s, The Value of Optimism)

Too Close to the Solana

Many in the crypto community have taken to referring to blockchains with a single settlement layer as "monolithic" versus their "modular" counterparts.
Wrapping transaction execution, settlement, network consensus, and data availability on one main chain may not be the ideal way to scale a blockchain ecosystem, and more communities are gravitating towards a more segmented approach with various layers optimized for different levels of performance and decentralization. You can think of this similarly to microservice architecture in traditional app development.
What would it take for an alt-L1 like Solana to win in the endgame versus modular blockchains?
  1. Computational advances that make running high-powered nodes cheaper over time
  2. Delays or securities issues with rollups
  3. User preferences and experience
  4.  Product-market fit of killer apps on a specific chain
  5. How much centralization and MEV tolerance users are willing to accept.
There are no explicit downsides to monolithic chains; it's just that the engineering challenges are different. With monolithic blockchains, the innovations have to happen in hardware and transaction scaling, whereas with modular designs, the bigger challenges are arguably around blockchain communications, bridges, and solving other interoperability edge cases.
Solana has had a tough go in the wake of the FTX bankruptcy. After a red hot 2021, Solana has come back down to earth. Last year, I wrote:

"I will acknowledge the recency bias, but only if you also acknowledge the fact that Solana is really good at the things Ethereum doesn't even try to be good at. Solana is not trying to out-EVM and out-modularize Ethereum. It's trying to fit everything it can into its base chain."
FTX, Alameda, and some of their closest investors had close ties to Solana (SOL and SRM were major holdings on FTX's balance sheet according to bankruptcy documents) and were active, enthusiastic funders of its early ecosystem. But the Solana team and its early community were eager to embrace a mentality of "eating glass" during the last bear market and may be able to rebound again once this hurricane passes.
Solana has continued to improve its speed and node decentralization, if not its server/cloud level decentralization. Though the immolation of FTX/Alameda presented a setback to some of Solana's core supporters, another enterprise-grade rug pull may have more lasting impact.
At the end of September, the Solana network had a concentration of validators and staked SOL with Hetzner Online, a German web-hosting company. Hetzner hosted over 40% of the network's validators and over 20% of staked SOL, with a particular concentration in one German data center. In early November, Hetzner removed all Solana-related activities in their German site and stated publicly that they would forbid future crypto-related activities on their servers.
Though the incident did not interrupt the Solana network, it brought the issue of centralization back into focus.
Solana's validator count has tripled in the past two years, and roughly 70% of the network's tokenholders stake their SOL. Validators are distributed across 35 geographic locations around the globe and more than 100 unique data centers. And the Solana Foundation released a network health report in August 2022 intended to track centralization risks and showcase Solana's strength.
Still, data center concentration remains an area of concern for Solana and other L1s. Whether it's Hetzner or AWS, using large cloud providers to run validators gives those providers disproportionate power over the network.

Fortunately, hundreds of Solana validators were able to migrate from Germany to other countries such as France and Finland (and away from Hetzner to hosting providers such as OVH SAS) after the abrupt policy change in November.


On the plus side, Solana seems to be the ecosystem that has been pushing hardest on mobile accessibility. The Solana Mobile Stack represents a big bet on mobile infrastructure in order to power a generation of new crypto applications. Its Web3-native Software Development Kit (SDK) is built on top of Google's Android OS with a new smartphone (the "Saga") to match. The SDK allows any Android phone to run Solana's software, while the Saga offers Solana's software natively at a price point of $1,000.
Crypto isn't very mobile friendly, and the Saga is a bet and hedge that Solana developers will be able to iterate faster and more reliably on top of crypto-native designed hardware than the competitive devices from fee-extractive monopolies such as Apple. Saga also plans to leverage Helium for mobile coverage, which could boost the adoption of a major adjacent decentralized wireless protocol if it gains adoption.
Rooting for Solana means rooting for concurrent hardware, a mobile platform, and network advances. I'm not sure any other L1s (beyond Ethereum) can make that claim as credibly as Solana. They'll rebound in 2023.
(Required Reading: The Solana Ecosystem Overview, How Solana Wins the Endgame vs. ModularChains, Solana DeFi: Ramifications Post-FTX Collapse)

Cosmos & Appchains

Though we didn't call them "modular" blockchains last year, I wrote:
"Cosmos was the first community to work on a modular network of blockchains, and Ethe- reum's rollup-centric scaling plans seals the deal: the "one-chain-to-rule-them-all" thesis is dead, and Cosmos' Inter-Blockchain Communication protocol (IBC) does something Polkadot and Ethereum don't, keeps the protocol entirely open and independent of the Cosmos "Hub" and its native token, ATOM."
Now it's time to pay the piper. The Cosmos community is toying with what it now calls the "ATOM 2.0" era, which aims to enshrine the Hub as Cosmos' central data router and shared source of security. It's unclear whether ATOM 2.0 will work (it has received pushback on its ambition as a proposal for its adoption was rejected by the community), but while that developer debate plays out, Cosmos will likely remain the top ecosystem for "sovereign" appchain developers.
The flexibility Cosmos affords developers in creating their own chains (custom inflation schedules, transaction fees, transaction types, security models, coding languages, etc.) and the value accrual mechanisms the IBC affords appchains (MEV, transaction costs, etc.) make the ecosystem compelling for those with the technical chops to vertically integrate their apps. Look no further than dYdX, arguably the largest application in crypto, which migrated from a ZK-rollup to an appchain this year.
While L2s may inherit the $100 billion security budget of Ethereum at virtually no cost (periodic call data and no need to support a dedicated validator set), the tie-up with the EVM affords less flexibility, more interoperability challenges, lower throughput, and higher direct costs. When one of the most successful, high-throughput dApp developers switches backends, others will likely listen. Unless L2 speeds increase significantly or more non-financial transactions are pushed off-chain (DeSoc likes, gasless Snapshot voting, etc.), the apps people use could end up on Cosmos. The new standard operating model for successful decentralized applications could be to leverage an L1 as an initial launchpad: build, scale, and then migrate to a custom appchain.

It's very simple.
Sei and Canto are new L1s built on the Cosmos SDK with a focus on DeFi apps. Sei has a built-in central limit order book (CLOB) with a parallelized order matching engine. Its objective is to become the NASDAQ of crypto, and its shared liquidity and centralized order book model makes it optimized for DeFi use cases. Canto recently launched an EVM-compatible L1 which has pre-built core protocols such as an AMM DEX, lending protocol, and a stablecoin. It aims to make these core protocols free public goods for users and developers. Canto has chosen to take a democratized approach to user acquisition. There are no venture backers, no pre-sale, no foundation — just the chain. The objective  is to minimize rent-seeking activity and give users and developers simple, free infrastructure. While commendable, without revenue potential, products (and people) aren't motivated to compete, and developers may be hesitant to build enhanced features.


Other L1s

I don't want to gloss over some of the other L1 chains I wrote about last year. Instead I'll consolidate them here. Some are dead. Others live on. But I only have so many hours to write about these things, and there are still several chapters ahead.
Cardano had a big year of technical upgrades, including Plutus smart contract capabilities and the Vasil hard fork with upgrades to Plutus and better scaling. Will the slow and steady approach prove to be a prudent path to scaling? Cardano's volumes (transactions, TVL, and developer activity) have paled in comparison to bigger ecosystems so far. But 2023 will be a pivotal year.
Polygon crosses a bunch of different sections in the Rollup and Modularity sections above. I'd encourage you to read more about the project in our latest quarterly report and our ongoing analyses on all things EVM, rollups, and modular blockchains. They were a bright spot in Q3, setting new all-time highs in active addresses and NFT wallets. They've also got one of the best BD teams in crypto, one that notched partnerships with Reddit, Meta, and Starbucks this year.
Polkadot founder Gavin Wood was the co-founder and technical mastermind of Ethereum, so it's no surprise that Polkadot's design as an interoperable "chain of chains" looks like Ethereum post-Merge. Gavin's often a step ahead technically, and the Polkadot developer community consistently ranks  highly relative to peer communities. We'll see if that translates to more killer ecosystem applications in the new year. (Gavin is stepping down as head of the project, but continuing on as its chief architect in 2023.) You can read our full Q3 report on Polkadot here.
I covered Terra last chapter, so…
MOVE-ing On to New Stuff
The same strategy is evolving with a new batch of L1s seeking to solve crypto's user experience and scalability challenges. These upstart L1s are more specialized than their predecessors and can offer faster execution, lower transaction costs, and specialized components tailored to application verticals like DeFi.
What we call the "upstart L1 thesis" boils down to whether you believe that it's still possible to innovate around an enhanced user and developer experience at crypto's base layer. This year, we saw two potential candidates emerge:
Both Aptos* and Sui spun out of Facebook's nixed Diem project.
Aptos and Sui inherit years of R&D and partnership discussions from a world-class engineering team. Their new Rust-derived smart contract development language, Move, aims to give developers better control over their data management and more secure execution. Both projects pitched themselves as high-speed, high-scalability chains (they decouple consensus and parallelize transaction processing to finalize transactions in less than a second). These projects have strong teams, backers, and networks.
That said, Aptos had hiccups with its early launch, and Sui is still merely available on testnet. Sui may also have clawback issues from the FTX bankruptcy process. FTX Ventures invested $75 million of the
$350 million that Aptos raised this year, but that deal appears to have closed after the typical 90-day bankruptcy clawback window. The $100 million to Sui is a different story. It looks like a third of Sui's
$300 million in Q3 funding may have been received within the 90-day clawback date. We'll see if the FTX investments come into play for either network.
Success for any upstart L1s hinges on business development and, more importantly, developer courtships: a tall order in bull markets that's only made that much tougher in bears.

I'm skeptical about how much value will ultimately accrue to these upstarts in an icy winter. While the TAM of L1s are significant, I made the mistake last year of betting against Ethereum's dominance. I won't make the same mistake again now that the champ has passed its toughest trial yet. I'd expect Ethereum to stay over 70% dominant until we're out of crypto winter.
At the end of 2020, I thought Ethereum's lead was unassailable. At the end of last year, I wasn't so sure, as I was bearish on The Merge's timely completion. Now that I'm bullish Ethereum dominance again, I'm not sure the community should be rejoicing. I consistently get Ethereum directionally wrong in both directions.

Other, Other L1s

I should reiterate something I wrote last year in the preface to the L1 Chapter. Once again, I want you to stop reading, go to a mirror and repeat after me:
"It isn't TBI's job to hype my project onto the top L1 list. The above projects were selectively reviewed for timeliness and thanks to prior Messari analyst research. If we didn't get written up this year, there's always next year, or Messari's quarterly reporting. I understand that it's simply not possible to cover every single project that has exploded in this broad, complex market, even in subsections of a 150 page report. The fact that our protocol wasn't covered was not bias or incompetence on the part of the author. If anything, it's a wake-up call that we have to hustle to be so good we can't be ignored."
Last year, I also reflected on how incredibly cool it was that we could now essentially produce "earnings reports" for top crypto communities without the need for any central, corporate investor relations team, and that we could do so over any arbitrary time period and update it in real time. We put that 1000x improvement in investor information symmetry to the test this year and now offer quarterly reporting for all of these L1s:

These metrics are still evolving, their usefulness as valuation reference points is still in question, and I'm sure more than one person will copy this chart, crop out this subsequent paragraph, and dunk tweet on us for saying "ETH expensive, TRX is cheap" because it doesn't fit their priors.
But the metrics themselves are what they are. It's on us to make sense of them and come up with better ones, as needed, as we enter a more fundamentals-driven era for crypto.
If your community wants to start working with us on quarterly reporting, let's do it! Here's a taste of what we do for over a dozen L1s and 40+ major token projects:
EVM: Ethereum Q3| Polygon Q3| BNB Chain Q3| Avalanche Q3| TRON Q3

Solana Q3| TRON Q3| Polkadot Q3| Tezos Q3| Hedera Q3| NEAR Q3
And we push this data and these reports to Bloomberg, S&P, Refinitiv, and other top financial data aggregators - for no charge - in order to help crypto cross the chasm.
To preempt your mean tweets, I'll add that we're adding quarterly coverage for Cardano, ICP, and (I hope) Cosmos in Q1 as well as many of the other top 100 assets by market cap. Stay tuned for all of these project-specific updates in January.