Ethereum Research Paper Challenges Centralization With Decentralized Block Proposal System 

A new research paper published on Ethereum’s community research forum proposes a decentralized system to replace Ethereum’s centralized block-building model, eliminating maximal extractable value (MEV) risks while democratizing participation.

Can BFT and Randomization Fix Ethereum’s MEV and Builder Dominance?

Ethereum’s current Proposer-Builder Separation (PBS) framework, designed to mitigate MEV concentration, has led to centralization, with two entities controlling 80% of block production as of February 2025. The research suggests replacing PBS with a Byzantine Fault Tolerance (BFT)-backed system where all clients propose blocks via a shared random algorithm, aiming to restore decentralization.

Under PBS, specialized “builders” craft blocks to optimize MEV extraction, but this has concentrated power among a few players like Flashbots. The proposed system mandates that every Ethereum client—such as Geth or Nethermind—generates blocks using a Verifiable Delay Function (VDF) and RANDAO for randomness, ensuring uniform transaction selection. Validators then prune invalid transactions and reach consensus via a BFT model requiring a 3T+1 threshold.

By randomizing block construction, the system “eliminates block-level MEV,” preventing transaction reordering for profit. It also distributes block-building across thousands of clients, contrasting sharply with PBS’s reliance on centralized builders. The author “malik672” claims this could reduce Ethereum’s 12-second block intervals to 6-8 seconds through parallel processing.

Compared to PBS, the system prioritizes decentralization over optimization, potentially sacrificing some rollup efficiency. While PBS excels in scalability and MEV redistribution, the proposal offers stronger resilience via client redundancy and simpler validator roles. Trade-offs include challenges in synchronizing mempools and managing blob inclusion for Danksharding.

The paper concludes that Ethereum faces a choice between PBS’s scalability and this system’s egalitarian ideals. Future work includes testing blob-heavy mempool synchronization and refining rollup-specific optimizations.