Briefing
The core problem addressed is the divergence between the theoretical promise of Proposer-Builder Separation (PBS) and its real-world outcome, specifically the increasing centralization of block building on Ethereum. This research introduces an empirical game-theoretic analysis using the meta-game framework to model builder strategic bidding under asymmetric conditions. The foundational breakthrough is the formal proof that advantages in network latency and, critically, exclusive access to private orderflow act as self-reinforcing competitive edges, driving the MEV-Boost auction toward an oligopolistic equilibrium. The most important implication is that the current implementation of PBS, absent a protocol-enforced, trustless mechanism for fair orderflow distribution, fundamentally undermines the decentralization of block production, posing a long-term systemic risk to censorship resistance.
Context
Prior to this research, the prevailing theoretical solution to validator centralization from Maximal Extractable Value (MEV) was Proposer-Builder Separation (PBS), implemented via MEV-Boost. The established theory posited that by decoupling the block proposal and construction roles, a competitive, efficient, and decentralized market for block building would emerge, where builders would bid away all MEV to proposers. This model assumed a near-ideal market with symmetric access to network resources and orderflow. The unsolved foundational problem was understanding why, despite this mechanism, empirical data consistently showed a rapid consolidation of block-building power among a handful of entities, contradicting the expected competitive equilibrium.
Analysis
The paper’s core mechanism is an empirical game-theoretic model that simulates builder behavior in a dynamic, multi-agent auction environment, specifically accounting for real-world asymmetries. The model fundamentally differs from previous approaches by allowing agents to explore a wider range of “meta-strategies,” which are more complex and adaptive than simple heuristics. The analysis demonstrates that a builder’s ability to secure private orderflow → transactions sent directly to them to bypass the public mempool → provides a non-biddable informational advantage.
This advantage, coupled with superior latency, allows a dominant builder to consistently construct more profitable blocks and outbid competitors. The new primitive is the identification of a centralization-reinforcing feedback loop → winning blocks attracts more private orderflow, which increases future win rates, leading to an economy of scale that marginalizes smaller, less-resourced builders, thus formalizing the oligopoly’s emergence.
Parameters
- Dominant Builder Share → 90% (Two builders controlled this share of MEV-Boost blocks as of April 2025, demonstrating market concentration )
- Top Builders Block Share → 97.51% (The top 10 builders produce nearly all blocks built via MEV-Boost, indicating an extreme oligopoly )
- MEV-Boost Block Share → 90% (The proportion of Ethereum blocks currently built using the MEV-Boost mechanism )
- Public MEV Value → 31% (The estimated percentage of total block value derived from public transactions, the remainder is from private orderflow )
Outlook
This research opens new avenues for mechanism design focused on trustless orderflow distribution. The next critical step is the formal development and deployment of cryptographic solutions, such as encrypted mempools or decentralized private orderflow routing protocols, that can credibly enforce symmetry among all block builders. In the next three to five years, this theory will drive the architectural shift toward a more robust form of in-protocol PBS that actively mitigates private orderflow advantages. This will unlock the potential for truly decentralized block production, moving the system away from reliance on a small set of privileged entities and enhancing the network’s resilience against censorship.
