On September 15th, the Ethereum Merge went live with a quarter million people observing the transition. Shortly after the success of the Merge, a separate set of infrastructure came online, enabling the MEV relayer networks to start sharing blocks with Ethereum validators. This new block marketplace is now baked into the Ethereum protocol making it more important than ever to provide additional transparency into the Ethereum block supply chain.
When an Ethereum validator creates a block, it can decide to construct its own block using transactions pulled from the public mempool OR it can query a marketplace of private relayers who bid on their block to be included in the next slot. The relayers act as trusted* aggregators of proposed blocks, verifying the block’s validity and sharing the most profitable block header with the validator. Relayers receive these blocks from Builders who find opportunities for making profit by reordering the transactions (ex. arbitrage) and are willing to pay validators for the right to create a block. A validator can opt-in to using relayed blocks to increase their profit by running mev-boost and adding one or more of the following relayers:
Flashbots - https://boost-relay.flashbots.net
Bloxroute - https://bloxroute.ethical.blxrbdn.com
Blocknative - https://builder-relay-mainnet.blocknative.com
Eden - https://relay.edennetwork.io
Manifold - https://mainnet-relay.securerpc.com
*This role will become more trustless in the future with new models of Proposer Builder Separation.
To help provide more visibility into the relayer marketplace, I’ve created a set of datastreams and dashboards. You can track these dashboards over time or plug directly into the data to answer MEV relayer related quetions.
The first page of the dashboard focuses on the overall relayer market share and profitability for validators.
What is the market share breakdown by relayer?
Which relayers provide the highest ETH rate per block?
How many of each relayer’s blocks are ignored by the proposers?
Are there any competing blocks from relayers for the same slot?
What is the ETH payout difference for competing blocks?
How many unique builders and proposers are connected to each relayer?*
*Note that the data here looks at the blocks which are actually surfaced from the relayers. There may be more builders that never produce winning blocks or proposers who have not yet had a chance to include a block.
The second page digs into the transaction type differences by relayers.
Which relayers have more ERC20-type transactions in their blocks?
Which relayers have more ERC721-type (NFT) transactions in their blocks?
It’s only been one week so these findings are likely to change. We see a clear dominance in blocks produced by Flashbots, followed in distant second by Bloxroute’s Max Profit (aka sandwiching🥪) relay. The non-Flashbot relays make up around 20% of the MEV blocks, providing some censorship resistance to the network, even if all validators ran mev-boost.
We see a small handful of blocks that were suggested by relayers but ignored by the validator. The validator opted to build their own blocks instead of choosing those from the relayer. At this time, there haven’t been any competing blocks produced by two or more validators for the same slot. Once there are, we can compare the bids to see which relayer can produce the highest fees.
Flashbots is again the clear winner in both number of builders and number of proposers (validators). Once a relayer has large proposer market share, builders will gravitate toward this relayer since they’re more likely to get their block included. If you run a validator, consider adding some of the under-represented relayers to your mev-boost config to help balance the market.
Finally we crack open the MEV blocks by relayer and see if we can spot any differences in the underlying transactions. We see some differences when looking at ERC20 and ERC721 (NFT) transaction inclusion by relayer type, but the sample sizes are still small enough that the differences could be attributed to random chance. It’ll be interesting to see if relayers specialize on any particular type of transaction (NFT mint front-running or ERC20 arbitrage).
Check back in on the dashboards over time to see how the data evolves!
To collect the data, I’ve put together a simple monitor that queries the /data API of each relayer and pushes the resulting blocks data to Google BigQuery. The monitor is open source and I’ve provided instructions to run your own. You can also query the datastream directly and build your own views and dashboards. This can be joined with the public Ethereum BigQuery dataset for deeper analytics.
If you have an idea for a new dashboard or question about using the data, reach out on Twitter!
collect://