Ethereum researcher Malik672 has dropped a research post proposing a new algorithm using random allocation of transaction orders to address the growing problem of centralisation on the Ethereum network. Specifically, he used Byzantine Fault Tolerance (BFT) to remove Maximal Extractable Value (MFV). BFT means the network continues to run despite a dishonest node polluting it with false information. MFV means a dishonest actor uses inside information about the Ethereum network to make money on the market. Toni Wahrstatter, an Ethereum Foundation researcher, found that in the first two weeks of October 2024, 89% of Ethereum blocks were handled by only two block builders, Beaverbuild and Titan Builder. Malik672 found that the figure has since dropped to 80%, but it still shows that Ethereum is far from decentralized. “Ethereum’s shift to PoS and adoption of PBS”, wrote Malik672, “separates block proposing (validators) from building (specialized builders) to address MEV disparities”. “Yet, centralization persists: as of February 2025, approximately 80% of Ethereum blocks are proposed by just two entities—large builder-relay coalitions like Flashbots and their peers—concentrating power and profits. This undermines Ethereum’s decentralized ethos”. For example, let’s assume a trader wants to make a huge transaction on a decentralized exchange (DEX). A trader with inside access, conducts an MEV trade by performing a sandwich attack. This means the trader buys the token before the large transaction and then sells after the price increases. The problem, referring back to Malik672’s research, is that a centralized blockchain allows bad actors to use inside information to have an unfair advantage on the market. The regular trader pays more, while the corrupt insider pays less. Malik672’s system would decentralize block building, using a random generator and distributing the building process to thousands of clients worldwide. The current system uses Prosper Builder Separation (PBS) to limit MEV. However, the new random distribution system could have added benefits, such as optimising mempool handling and faster transactions. “This proposal”, wrote Malik672, “introduces a decentralized random block proposal system. All Ethereum clients—not a handful of builders—construct blocks using a cryptographically random algorithm”. “Validators execute these blocks, achieving consensus via N ≥ 3T + 1 BFT. This eliminates block-level MEV, fully democratizes block proposing across Ethereum’s node network, and aligns with its trustless roots, while remaining compatible with Danksharding’s blob requirements”. However, this approach could have major drawbacks, requiring extensive testing before it is rolled out on the network. One possible drawback could be large overhead problems from distributing the block building process to thousands of clients. This may increase the amount of computation needed, thus making the network even slower. Another drawback could be the increased risk of Sybil Attacks, which means that one bad actor could generate multiple fake clients to gain more inside information and MEV. Further, Decision-Making Complexity may increase because so many block builders may disagree more on transactions, causing more problems for the Ethereum network. “PBS optimizes for rollup scalability”, wrote Malik672, “but this system prioritizes fairness and full democratization—crucial as centralization creeps in. It’s a trade-off: trustlessness over L2 precision”. “For an Ethereum valuing equity over efficiency, this wins; for one chasing scalability, PBS holds. Redesigning for rollup-specific optimization could bridge the gap, making it a versatile contender”.
