This proposal introduces the MEVless protocol, a novel approach to prevent Maximum Extractable Value (MEV) attacks by implementing blind transaction sequencing. The core principle of MEVless is to prevent validators from seeing transaction content during the ordering phase, thereby eliminating the fundamental prerequisite for MEV attacks. The protocol separates block production into two phases: sequencing blocks that order transactions based on transaction hashes without revealing transaction content, and execution blocks that execute transactions in the predetermined order. This design eliminates the ability of miners and validators to perform MEV attacks such as sandwich attacks and front-running by removing their access to transaction content during the critical ordering phase.
Motivation
MEV attacks, particularly sandwich attacks and front-running, represent a significant threat to blockchain ecosystems, especially in DeFi applications. The fundamental prerequisite for MEV attacks is that validators can see transaction content before ordering, which allows them to:
Sandwich Attacks: Insert buy and sell transactions around user transactions to manipulate prices and extract value
Front-running: Execute similar transactions before users to capture arbitrage opportunities
Back-running: Execute transactions after users to benefit from price movements
Unlike legitimate arbitrage activities that provide liquidity, these attacks are purely extractive and harmful to the ecosystem. They reduce user confidence, increase transaction costs, and can lead to significant financial losses for users.
Current solutions have limitations:
Encrypted Mempools: High computational overhead for encryption/decryption
Private Mempools: Centralized and not fully decentralized, use trustful instead of trustless.
PBS (Proposer-Builder Separation): Still allows builders to see transaction content and becoming more and more centralized.
The MEVless protocol addresses these issues by fundamentally changing how transactions are ordered, making it impossible for miners to perform MEV attacks while maintaining decentralization and efficiency. By eliminating MEV attacks at the protocol level, DeFi developers can focus on building innovative financial products and user experiences without constantly worrying about MEV protection mechanisms, leading to faster development cycles and more robust applications.
Specification
Protocol Overview
The MEVless protocol operates on a two-phase block system:
Sequencing Blocks: Order transactions based on transaction hashes and prepayment amounts without revealing transaction content. Sequencing blocks interval can be shorter than execution blocks.
Execution Blocks: Execute transactions in the predetermined order from sequencing blocks
Block Types
Sequencing Blocks
Purpose: Order transactions and collect prepayments
Frequency: Every odd-numbered block height
Operations:
Receive transaction hashes from users
Collect prepayments (gas fees + optional tips)
Order transactions by prepayment amount (highest first)
Publish ordered transaction sequence as commitment
Store transaction hashes in block
Execution Blocks
Purpose: Execute transactions in predetermined order
Network executes transactions in predetermined order, deduct the remaining fee and 50% of the actual difference between gasUsed and gaslimit for tax ( prevent the gasLimit much more than gasUsed so that validators can entire the whole block to get MEV)
Gas Economics
Prepayment Structure
Total Prepayment = Base Gas Fee + Optional Tip
Base Gas Fee: Fixed cost for hash storage and sequencing computation
Optional Tip: Variable amount for priority in transaction ordering
Gas Cost Calculation
Sequencing Phase: Minimal gas cost for hash operations
Execution Phase: Standard gas cost for transaction execution
Rationale
Why Blind Sequencing?
Blind sequencing eliminates the root cause of MEV attacks by preventing miners from seeing transaction content during the critical ordering phase. This approach is more efficient than encryption-based solutions and more decentralized than private mempool approaches.
Why Two-Phase Blocks?
Separating sequencing and execution allows for:
Efficient Ordering: Hash-based ordering is computationally lightweight
Network Efficiency: Small hash data reduces bandwidth requirements
Commitment Mechanism: Public ordering commitment prevents manipulation
Why Random Block Production is Critical?
Random block production is essential to prevent speculative MEV attacks. When block producers cannot predict when they will be selected to produce blocks, they cannot reliably perform speculative MEV attacks because:
Unpredictable Selection: Attackers cannot know if they will be the next block producer
Economic Risk: Speculative transactions require prepayments that are lost if the attacker is not selected
Decentralization Amplification: More validators mean lower individual selection probability, increasing attack costs exponentially
Backwards Compatibility
This protocol will add some RPC methods to support the two-phase block system:
eth_sendTxHash: Users submit transaction hashes instead of full transactions
eth_getTxSequence: Returns the ordered transaction sequence that is committed
Test Cases
Basic Transaction Flow
User submits transaction hash with prepayment
Network orders transaction by prepayment amount
User submits transaction content
Network executes transaction in predetermined order
MEV Attack Prevention
Attacker submits transaction hash for sandwich attack
Network orders transaction without revealing content
Attacker cannot see victim transaction content during ordering
Attack fails due to lack of transaction visibility
Security Considerations
The MEVless protocol requires a consensus mechanism with unpredictable block production to prevent speculative MEV attacks. Without unpredictable block production, block producers can perform cost-free speculative MEV attacks by submitting transaction hashes and then deciding whether to submit their transaction content based on other users’ transaction content, thus performing cost-free speculative MEV attacks. Since the prepayment ultimately goes to the block producer themselves, predictable block production allows them to recover their prepayment costs through block fees, making MEV attacks economically viable. Random block production forces attackers to pay prepayments for speculative transactions without knowing if they will be selected to produce the next block, making the prepayment cost real rather than recoverable through block fees, thus making such attacks economically unviable. The anti-MEV effectiveness increases exponentially with network decentralization - more validators mean lower individual selection probability, making attacks economically unviable.