My name is Chase Coleman and I am one half of what we call DART, which stands for Data Analytics and Research Team, at Across. My role is to uncover and analyze cross-chain data and look into how it relates to our protocol and, more generally, the economics of Across.

My background in crypto leading up to my current role at Across took a pretty interesting route. I did my PhD in Economics at NYU. While I was finishing my degree, I was put in touch with Hart, UMA/Across Founder, who introduced me to the core concepts behind smart contract technology and crypto. Admittedly, I had previously thought Bitcoin was a little pointless in much of the developed world (I don’t think this now), which people can laugh at me for, but I was really interested and impressed when he started talking about Ethereum and the composability behind smart contracts.

At some point between 2018 and 2021 we all can admit that we thought Bitcoin was a little pointless, right?

I ended up doing some consulting work with Risk Labs early on. After that stint, I ran my own company for a few years, but in 2021, I saw that Risk Labs had an opening for someone to do data and economics-focused work. I had really enjoyed my time with Risk Labs and I thought that I would be a good fit, so, I applied, and here we are!

I spend most of everyday looking at cross-chain data that pertains to internal research projects designed to help our team decide what we should be working on next. Our team emphasizes data-driven decision making, and because everything is transparent and accessible in crypto, you can get a lot of information about what is happening both internally and externally.

Pieces of this process have always felt like a less-malicious version of corporate espionage. Traditional companies like Wal-Mart or Costco would pay unbelievable amounts of money to be able to observe transactional level data about their competitors and, because of the transparency of the block chain, we can get that information “for free” (obviously, there’s lots of work that goes into keeping our data pipelines operational). So, the fact that we can do this and make decisions based on a holistic market outlook is super cool and very valuable [to consumers and the Across staff].

Wondering what the “Wal-Mart” of crypto is when writing this, but then remembered that Wal-Mart is apparently going to be the Wal-Mart of crypto soon.

We’ve chatted quite a bit about what we’re working on during community calls and shared pieces of it on Twitter. We are focused on compiling accurate cross-chain data that goes further than TVL and one-sided volumes. We look at lots of things, such as: which tokens are trending on bridges, how much aggregator volume accounts for total bridge market share, and research other ways that we can improve the efficiency of Across for bridgoors.

There isn’t standardization in the cross-chain bridging space yet, so there’s still a lot of code guesswork that needs to be done before data can be confidently retrieved and analyzed with accuracy.

The thing that gets me most pumped, and that might be a little bit surprising, is our median transfer speed over time. I remember when we ran early versions of the Risk Labs relayer bot and we used to be stoked when our median transfer time was 5–6 minutes. We would literally be like: “Hell yeah! We’re the fastest bridge in the world!” If you look at today’s speeds, the (decentralized) Across relayers are now writing their own custom bot implementations and the median transfer speeds on most routes are sub-10 seconds. If you deposit something on L1, before the next L1 block is mined, your funds are already on an L2. I think it was 10.3 seconds or something like that a few weeks ago and it’s just insane! It’s a bit funny to remember celebrating sub 5 minutes when we now are disappointed if the median time bumps above 30 seconds.

Welcome to my world! This is what I look at all day long and I think why our team (especially DART) is so passionate about getting the word out about the design of Across.

Leagues below the competition on average transfer times — which is actually leagues ahead, don’t let the chart fool you!

It depends on what you mean by instantaneous, but 10 seconds is getting pretty close when you consider that L1 has 12 second block times. If you only looked at chains that have less than 12 second block times, we’re seeing 25–40% of transfers happen in less than 3 seconds.

I was experimenting the other day (another great part of my job) and noticed that some transactions were getting filled before my computer even saw the deposit event. It could just be that I was particularly “far” away (geographically) from those blocks but, to me, this feels like relayers are already optimizing on things like “what geographic region should I run my bot from” so it’ll be fun to see how they keep optimizing.

While I do think there’s room for us to keep getting faster, we are in this heyday where relayers are deeply competitive but the only dimension on which they can compete is speed. I think as the Across, and more generally the bridging space, evolves we will see additional dimensions of competition and added complexity. It’s possible both of these things result in some degree of a slow down.

At any point in time, we want to be the cheapest, fastest and most secure bridging option. Security doesn’t show up in the data (until funds are hacked…), so we use data to improve our data-driven decision making on pricing and speed after ensuring that we have prioritized security.

Cost curve is most important, but it’s been really interesting to see how much further Across can optimize to compete on speeds.

In order to evaluate how we compare in terms of pricing and speed, we need to be able to observe when transactions start, how much was sent, where it arrived, and how much arrived. This is one of the reasons that DART’s job can be a little tricky — Matching a transfer from an origin chain to a destination chain is a bridge-by-bridge process and there is no real standardization here. It’s a bit like working on a puzzle in the sense that we approach it one piece at a time and fill in the gaps when we can. That said, while it has been a tough nut to crack, I think we’re at a point where the data is quite accurate and provides strategic insights to the Across community and can help us make good data-driven decisions.

It really comes down to matching transactions. DeFiLlama has a great compilation of what they call “Deposits” and “Withdrawals.” So, for any bridge they track, you can see the total “Deposits,” which is sending funds from origin chains, and total “Withdrawals,” which represents you pulling your funds out on a destination chain. They have this data for almost every bridge, but the thing that they don’t do (because it’s a massive pain in the ass) is match deposits with their corresponding withdrawal. This is the biggest difference between the DeFiLlama data and the data that DART has built. I think for most things, the DeFiLlama is great, but if you want to focus on some of the things that we are interested in, you need the matching.

You can check out the different bridge data on DeFiLlama here: https://defillama.com/bridges

One thing that Dylan (the other half of DART) tackled early on was being creative in how he used our data with Dune data and building datasets where we can observe the first 5 transactions that users submitted once their bridge had finalized. This data allowed us to examine what people were doing once they had their funds on the destination chain and there were some pretty neat results.

For Across users, some of the most frequent actions included a swap or depositing funds into a lending protocol of some sort. This helped influence the type of action we plan to support with our composable bridging (coming soon!).

One of the more amusing things that we learned from this exercise was that for some bridges, the most likely thing their users did once they received their funds was use that bridge again. You can imagine that this was a pretty amusing thing to find in the data and the DART team laughs about this frequently.

We’ve had lots of discussions about this internally but there’s a fundamental question of “what defines a user on the blockchain”. Lots of people, including myself, have multiple addresses and I use them all with some frequency (everyone needs a degen wallet). So, it’s this kind of odd case where you have some number of active addresses, but, if you asked me how many unique users we had, I think that’s a hard question to answer accurately.

If I had to give an answer, my logic would look something like:

In the month of October we had roughly 30,000 addresses interact with Across. If you suppose that there are 2 wallets per person on average (I probably have more active wallets than the typical person) then we land at about 15,000 active users.

If you expand that to be the bridging space as a whole, I find it harder to answer. You see days where there are massive spikes in users — sometimes up to 250,000 per day — and it’s just hard for me to believe that there are even 50,000 individual people doing a bridge transaction per day. I think we are still pretty early and my suspicion (based on nothing but my raw instinct) is that we’re probably seeing closer to 5,000–10,000 individuals bridging per day.

I think so. I know a lot of people still use the Uniswap frontend but I deeply believe in the mission of aggregators both on the DEX side and on the bridge side. We are a long ways from where we want to be but I think the UX is starting to improve with the great bridge aggregators we have already and with various wallets already adding their own bridge tabs (often powered by the great bridge aggregators!). As “cross-chain living” becomes more native, you’ll start to see more and more bridge volume.

We hope to see Across integrated on more in-wallet bridging dapps!

I feel like I have some insensitivity to this, and although I don’t exactly know why I have this insensitivity… Maybe I can attribute it to having been in the crypto space for a few years now?

If I put things in perspective and really imagine things going to zero, what’s the worst case? I’m confident that the world will continue to provide interesting things to work on and there are lots of opportunities. That said, I deeply believe in our mission and so, obviously, I’d love for Across to be successful, but, if it all fell apart, my life wouldn’t fall apart. That context often helps keep the stress levels down.

We’re at about 14 years for Bitcoin, but only about 8 for Ethereum. Right on schedule?

The way that Ethereum is built, any changes that they make to improve mainnet, improves L2s even more so. Obviously you could make fundamental changes to how Ethereum works that would make it scale better, but I do think that as it is designed now, especially with things like 12-second blocks or the gas limits that they recently posted about, further supports the consensus that the only way to scale Ethereum is going to be through L2s.

Ethereum’s recent upgrade saw network costs take a downward path.

If you think about what we want from crypto, exponential growth really seems like the end goal. If you cut costs in half that’s a linear improvement and linear improvements cannot keep up with exponential growth.

I feel like I’ve seen some of this, but these people tend to be folks who took big gambles on certain tokens where it’s paid off, and I don’t feel a particularly deep connection to that part of the crypto community.

Eventually I’d like every person in the world to be able to use crypto on a near daily basis to replace the entire financial industry as we know it . And with that large number of adoption, we’ll continue to see even more subsets of crypto communities, like the Ferrari or Lambo-chasing community that I think is a bit coupled with this. I think these folks are also just very early to the space and all the wealth is cause for attention for people outside of crypto, but if you asked me what these people were building, I wouldn’t know the answer. The people I respect the most in the space are the ones that are building really cool things and I think these are ones that typically make a lasting impression.

I do have lots of books and “my favorite book” depends on what I’m working on since “my favorite book” is the one that’s most helpful for me at any particular time. There are a few books that deserve a shoutout since they’ve been repeatedly helpful: Recursive Macroeconomic Theory was a staple during my PhD and is a nearly encyclopedic reference for modern macroeconomics, Statistical Rethinking by Richard Mcelwreath, Economic Dynamics by John Stachurski (John is one of the clearest thinkers that I’ve ever met and this clarity comes through in his writing), Tim Roughgarden has a number of books I’ve found useful, and currently I’m working through a book called Introduction to Modern Cryptography.

In terms of books that are “more interesting for a general audience”, my brother gifted me ‘Against the Gods’. The book is a history of risk and probability and the path to how people started thinking seriously about these things. My brother also gifted me ‘From Foot Soldier to Finance Minister: Takahashi Korekiyo, Japan’s Keynes’ which is a fascinating biography on Takahashi Korekiyo who was the Finance Minister of Japan leading up to WW2. His fiscal policies helped bring Japan out of the Great Depression but, in doing so, he had upset various individuals inside the military and ultimately was assassinated.

Chase’s love for reading is contagious! He also started a team-wide internal cryptography learning initiative.

I really do like reading textbooks (economics+math+programming preferred) and I actually do a bit of that for fun. Other than reading, I like to golf — I’m not particularly good at it, but I enjoy the act of being outside and “walking with a purpose”. I also like spending time with my family. I have a two year old son and I love doing things with him!

Can you guess which one is Chase? This picture was taken at the Across off-site in Portugal earlier this year!

I do! It’s called the Pennybacker Bridge. It’s in Austin and, while it’s not that impressive of a bridge, I have lots of good memories of being on the lake around the bridge and that’s why I like it.

Bridge in terms of impressiveness, would be the Verrazano-Narrows bridge. Just a New York City shoutout!

This bridge was, at one point, the largest in the world! It also bends during storms to maintain structural integrity.

Across Protocol is an intents-based interoperability protocol, capable of filling and settling cross-chain intents. It is made up of the Across Bridge, a powerfully efficient cross-chain transfer tool for end users, Across+, a chain abstraction tool that utilizes cross-chain bridge hooks to fulfill user intents and Across Settlement, a settlement layer for all cross-chain intent order flow. As the multichain economy continues to evolve, intents-based settlement is the key to solving interoperability and Across is at the core of its execution.

Website | Forum | Discord | Twitter | YouTube | Docs | Github