The market doesn’t care about your narrative. It cares about execution. Last week, Vitalik Buterin’s address was front-run by Jaredfromsubway.eth—a degenerate MEV bot that has been harvesting sandwich attacks for years. The community’s response? A flurry of proposals to encrypt the mempool. EIP-8184, LUCID, EIP-8105, FOCIL. The goal: hide trades from bots before they can attack. Noble. But the market doesn’t reward nobility. It rewards efficiency. And the encrypted mempool, as currently designed, is anything but efficient.
We didn’t always see the blind spot. For years, the crypto industry accepted MEV as an unavoidable tax—a cost of doing business in a decentralized order flow. Private relays emerged as a band-aid: send your transaction to a trusted intermediary, bypass the public mempool, pay a premium for privacy. It worked, but it created a new centralization vector. The Ethereum protocol development community, led by researchers like the authors of EIP-8184, now proposes a more elegant solution: encrypt the entire public mempool. No more sandwich attacks. No more front-running. Just pure, permissionless, fair ordering.
But the devil is in the cryptographic details. And the details are a mess.
Context: The MEV Arms Race
Ethereum’s current architecture is a paradise for MEV bots. Transactions are broadcast to the public mempool in plaintext. Bots scan them, simulate outcomes, and insert their own orders ahead of yours. The result: you pay more, the bot profits, and the network remains decentralized—but at the cost of user experience. Over the past two years, the total value extracted by MEV has fallen from its 2021 peak, but the frequency of attacks on high-profile addresses (like Vitalik’s) keeps the narrative alive.
The proposed solutions fall into three categories:
- Encrypted Mempools (LUCID via EIP-8184): Transactions are committed to the block in encrypted form, then decrypted after the block is proposed. The key is held by a separate “key publisher.”
- Trust Graphs (EIP-8105): A registry of providers who can be trusted to order transactions fairly, with a social-based reputation system.
- Inclusion Lists (FOCIL/EIP-7805): Multiple validators force a block builder to include specific transactions, reducing the builder’s ability to censor or reorder.
All three are being discussed in the “Encrypt the Mempool” conference calls, the most recent of which was on August 19. The timeline: FOCIL is a priority for the Hegotá upgrade (expected 2027). The encrypted mempool is still in the research phase.
Core: The Cryptographic Elephant in the Room
Here’s the problem: there is no known cryptographic primitive that satisfies all of Ethereum’s requirements for an encrypted mempool. The authors of EIP-8184 admit this openly. They need:
- Small public keys (to fit in block headers)
- Non-interactive decryption (so the block builder can decrypt without talking to the sender)
- No trusted setup (to maintain decentralization)
- Feasible ciphertext size (to keep gas costs low)
- Strong chosen-ciphertext security (to prevent malleability attacks)
- A credible path to quantum security (because Q-Day might come as early as 2029)
No existing scheme checks all these boxes. The current best candidates—like lattice-based encryption—are either too large in key size or require interactive decryption. The result: LUCID’s design is a “commit-reveal” hack. The sender submits a sealed transaction, then later reveals the decryption key. But the key revelation is not enforced by the protocol; it relies on a third-party key publisher. If the publisher goes offline or colludes with the builder, the transaction is lost.
This is a blind spot. The crypto community has been so focused on the elegance of the commit-reveal pattern that they’ve ignored the trust assumption underneath. The key publisher becomes a new centralized point of failure. And because the protocol cannot punish a failed key reveal (it can’t distinguish between a malicious publisher and a network failure), the economic penalty is a blunt instrument: a reservation fee equal to one-eighth of the block gas limit, which is forfeited if the key doesn’t arrive. That’s a heavy cost for honest users if the network glitches.
We didn’t fully appreciate the economic implications until I ran the numbers. A reservation fee of 1/8 of the block gas limit for a single transaction means that a user executing a 100,000 gas swap would need to reserve ~2.5 million gas of space. That’s a 25x overhead. Even if the fee is refunded upon successful decryption, the opportunity cost of locking that block space is real. In times of high demand, this could make the encrypted mempool more expensive than using a private relay.
Contrarian: The Real Value Is Raising Attack Costs, Not Eliminating MEV
The market doesn’t care about technical purity. It cares about what works. The contrarian view is that the encrypted mempool, even in its current flawed form, achieves something important: it makes MEV attacks economically unattractive for bots.
Consider the bot’s perspective. Today, sandwiching a transaction costs only the gas price and a bit of computational power. With LUCID, the bot would need to reserve block space just to see the encrypted transaction, then risk losing the reservation fee if the key is delayed. The 1/8 gas limit also means that bots cannot easily batch multiple attacks in a single block. The cost of an attack rises from a few dollars to potentially hundreds of dollars per attempt.
This is a subtle but powerful shift. The goal is not to eliminate MEV—that’s impossible in a permissionless system. The goal is to raise the cost of extraction until it becomes unprofitable for all but the most valuable transactions. In that sense, the encrypted mempool is a form of economic defense, not a cryptographic silver bullet.
EIP-8105’s trust graph takes a different approach. Instead of hiding the transaction, it builds a network of reputation-based providers. Providers can selectively trust each other, creating a decentralized web of order flow. The downside: no on-chain slashing, no trustless enforcement. It’s a social solution, which is fragile in a game-theoretic environment. But it has the advantage of being deployable today, without waiting for new cryptographic breakthroughs.
Takeaway: The Road Ahead
Ethereum’s fight against MEV is entering a new phase. The encrypted mempool is a long-term bet, tied to the Hegotá upgrade and the eventual availability of post-quantum cryptography. The immediate timeline: FOCIL will arrive first, likely in 2027, giving validators more power to enforce inclusion. The encrypted mempool will follow, if at all, no earlier than 2029.
In the meantime, the market will continue to use private relays. But the landscape is shifting. The rise of “key publisher” services could create a new intermediary layer, one that regulators will eye with interest. And the quantum threat—now projected to hit as early as 2029—means that any long-term solution must be quantum-safe, adding another layer of complexity.
The blind spot is not just cryptographic. It’s economic. The community assumes that users will flock to the encrypted mempool because it’s permissionless and fair. But if the cost is higher than using a private relay, users will choose the cheaper option. The market doesn’t care about your narrative. It follows the path of least resistance and lowest cost.
We didn’t design for that. And that’s the real challenge.