Superchain Circuitry

Takens Theorem
Etherscan Blog
Published in
7 min readSep 6, 2023

--

Coinbase’s OnChain Summer is now complete. During August 2023, OnChain Summer celebrated the activation of Coinbase’s Base, a second layer (L2) on Ethereum (and built from Optimism’s OP Stack). Daily events across August included many NFT mints and other on-chain Base activities. Thousands of wallets participated. OnChain Summer was so active that you can see on-chain Base usage spike and show an approximate daily cycle, probably around when a mint or other activity became available.

Data from adapting this Dune dashboard by @tk-research

Second layers like Base offer users cheaper and faster transactions. The OnChain Summer event was meant to illustrate to users the fun and ease of participating. As unfortunate evidence of these benefits, when Base was opened to transacting, there was some event log spam on its ledger. Some transactions logged 10,000 events and slowed some explorer interfaces when viewing these transactions (see here for an example on BaseScan; caveat: it may freeze some browsers!).

Example: BaseScan view of event log spamming on transaction

Initial hiccups aside, OnChain Summer seemed largely successful. It culminated in the Base Wars NFT project. Base Wars is part of the Finiliar family of NFT projects, in which “animal software” of cute little avatars respond to on-chain data. Thousands of distinct wallets minted a Base Wars NFT.

Base Wars for OnChain Summer

OnChain Summer provokes a broader question about the role of an L2 in Ethereum or any protocol: What are second layers for? One simple view is that an L2 is a kind of simulacrum of the hallowed mainnet. This simulacrum’s “reality” is assured by its relationship to the real thing (mainnet) but it permits the same use cases that are facilitated by cost and speed improvements.

This answer to the question oversimplifies the concept of an L2 and the many relationships it may have with its L1. In a blog post last year, I summarized how users may engage L2 for a variety of reasons. This may include cheaper trading, developing faster and less expensive data-heavy on-chain applications, more affordable NFT experimentation and more.¹

This suggests an L2 has a more complex relationship to its L1.² One way to think about it is that L1 and L2 are part of a broader “circuit” in which users find distinct purposes for each. The L1 may be seen as ultimate “finality,” as its ledger is most likely to have the best security guarantees. An L2 greases the wheels on transactions, and so users may visit L2 for distinct reasons, such as gas-heavy NFT participation. In this simplest case, the L2 is participating with the L1 for a specific functionality. It is not simply copying a user’s overall engagement.

What are second layers for? Data source.

Example on Base

To quantify these ideas we can use the API for Etherscan and BaseScan. I grabbed transactions from Base blocks 3344550–3344559 yielding 122 unique transactions. I then sampled the Ethereum mainnet (L1) and Base (L2) ledgers for 87 from wallets in those transactions. All together, this process returned about 34,000 on mainnet and 110,000 on Base for all 87 addresses (some are high-transacting, such as bridge addresses). I categorized each of these transactions as ERC-721 (“NFT” activity), ERC-20 (“DeFi” activity) or other (e.g., mainly other contract calls, regular transfers, etc.).³

We can use a ternary plot to visualize the concentration of this activity, and how it may differ across Ethereum mainnet L1 and Base L2. A ternary plot is basically a triangle with three categories at each point. Plotting inside that triangle represents how prominent the different categories are. When the point is right at the triangle’s center, it means the three categories are balanced — in our case, it might mean equal DeFi, NFT, and Other transaction counts. But if a wallet has its data point at the very top (say, near Other), it means they do not engage in any ERC-20 or ERC-721 transactions at all and only Other. If a point is on one of the triangle’s edges, it means there’s some balance between the two categories that connect that edge (such as Other and DeFi) and no transactions of the other (third) point.

This is illustrated below for four addresses that were included in those Base blocks. The green dot shows the distribution of transactions over three types (NFT, DeFi, Other) for mainnet, and the blue for Base. The top row shows that some addresses have similar activity, because the two dots are close together in this space. Those on the bottom show a deviation of Base activity from the way they transact on Ethereum. (For example, the bottom left ternary plot shows a Base transaction distribution of Other, NFT, and DeFi as about 75%, 20% and 5%, respectively.)

Green = Ethereum L1 activity; blue = Base L2 activity

We can also look to the preference of addresses in L1 vs. L2 usage by taking the ratio of transactions between L1 and L2. This is shown below. We see a wide range — some addresses focus on one or the other, while many are at a balance of L1 and L2 usage. The concentration of activity differs widely across addresses.⁴

Focal Base addresses by relative L2/L1 usage (negative y-axis, more relative L2 use)

Superchain Circuitry and Explorers

OnChain Summer and Base offer a recent opportunity to see this wide range of distinct chain usage. Relationships among chains (L1, L2, etc.) and their users will probably be complicated. A fun and philosophical way of thinking about “layers” is that they are part and parcel of one larger circuit — a superchain, as Optimism’s OP Stack calls it. Layers are secured by the mainnet, but the flow of information inevitably includes bridging and transacting in a wide variety of ways. These chains should not be considered “apart” from one other. Instead, they are functionally coupled such that they should be considered one superchain or “meta-chain.”

Optimism OP Stack

This promotes new potential ideas for the role of blockchain explorers too. What I’ve shown above is just the simplest such analysis — showing distinct L1 or L2 activity. But in the future, we may be able to analyze and categorize users and applications in more complex ways. These future analyses may show more complex relationships that the “superchain circuitry” enables, perhaps especially when EIP-4844 arrives with its blobs making L2 usage even cheaper.

In a similar way, designers of explorers like Etherscan could think of themselves as mapping one big superchain circuit. New tools might facilitate discovery of how users are engaging L1/L2, and how users could engage them in new ways. Here is some mild speculation.

Teleportation links. Labeling contracts as bridging contracts should invite all related links on a bridge transaction to the associated chain on the other side. This would permit quick juxtaposition of an address on L1/L2.

Relative activity badges. The simplest possible way to mark a wallet on L1 or L2 is to specify by chain metrics whether it has relatively higher transactions or transaction volume on L1/L2.

Hypothetical relative activity bands (right) show chain usage; click, go to explorer for that address

Value lock-up engagement / time. A similar analysis may be how long bridged value has lasted on the L2. Next to a bridging transaction, a ratio of L2 transaction volume by value bridged could show how much usage is associated with that bridged value.

Usage fingerprinting / cross-chain entropy. A small marker consisting of a ternary plot or miniature bar plot could indicate the pattern of activity of a wallet on a given chain vs. other chains; it could, for example, be represented by a single metric such as “usage entropy” measuring whether the distribution of wallet activity on one chain deviates from mainnet or others.

Hypothetical “badge” showing engagement fingerprint

Conclusion

Many explorers like Etherscan, Nansen and others have features like this, but they are almost always focused on one chain per interface or explorer. In the coming years, these could be adapted to describe this “superchain circuit,” to connect information across explorers and facilitate discovery. The result might very well be a UX feel that sees mainnet and its second layers as part of one big superchain. Users see one explorer interface, with icons or other markers that specify in which part of that circuit a transaction has settled.

Endnotes

  1. This is what Optimism means by its “superchain” — a development stack that can interconnect and advance new use cases and applications across chains by leveraging their individual strengths and lowering the barriers for them to mutually interact.
  2. In fact the superchain concept noted above and the many more chains that may interact could challenge the “L1/L2” dichotomy altogether. Concept worthy of mention, but outside scope of the post.
  3. Obviously this leaves out ERC-1155. I do this for simplicity, though I may include it in a follow up analysis later.
  4. This analysis can’t tell if some of these origin wallets on mainnet are just hot wallets and there may be other activity on another main address. In many chain analyses, such confounds are inevitable so it’s an important caveat.

About Takens

I was not paid for writing this post. I wrote it for fun. I do creative projects and other work in the crypto space. I am on Twitter, and welcome a hello on my main timeline!

--

--

Takens Theorem
Etherscan Blog

Dynamic distributed data displays. Intermittent. Friendly.