Vitalik proposes anti-correlation incentives to promote decentralization in staking protocols by penalizing wrongdoers and applies them in various scenarios supported by empirical data.
Vitalik Buterin recently published a thought-provoking post article which delves into the concept of supporting decentralized staking by using anti-correlation incentives. Authored by Vitalik Buterin, the paper offers preliminary research and encourages independent replication attempts to validate the proposed ideas.
The main focus of the paper is to address the challenge of driving better decentralization within staking protocols. The author suggests that penalizing correlations between actors can be an effective mechanism to promote a more distributed and sustainable ecosystem.
The current approach in Ethereum’s scaling mechanics now includes an element of anti-correlation incentives. However, the paper argues that relying solely on edge-case incentives that can only arise in very exceptional attack situations may not be enough to drive decentralization.
To further improve anticorrelation incentives, the paper proposes an extension of this concept to handle more common errors such as missing attestations. Larger bettors, including wealthy individuals and betting groups, are believed to often run multiple validators on the same internet connection or physical computer, resulting in correlated failures. The article acknowledges that expecting these stakeholders to create independent physical settings for each validator would eliminate economies of scale in staking.
To validate the hypothesis, the author combined attestation data from recent epochs with information mapping validator IDs to publicly known clusters. By analyzing cases of co-failures (instances where two validators within the same cluster fail during the same slot), the paper provides empirical evidence of a redundancy of correlated failures within clusters. These data support the idea that validators in the same cluster are more likely to miss attestations at the same time compared to validators in different clusters.
Based on this analysis, the paper proposes a penalty mechanism based on the current number of missed slots compared to the average of the last 32 slots. This mechanism ensures that penalties for missed attestations are proportional to the number of validators that have failed in a given slot compared to recent slots. The paper highlights the robustness of this mechanism, as it is not easily manipulated and provides no incentives for participants to intentionally fail.
The research presented in this paper contributes to the ongoing discourse on decentralized staking and provides insight into the potential benefits of anti-correlation incentives. By driving decentralization and mitigating correlated errors, staking protocols can become more robust and resistant to attacks.
It is important to note that the research presented in the article is preliminary and the author encourages independent replication attempts to support the findings. The code used for the analysis is available on GitHub for reference.
In conclusion, supporting decentralized staking through anti-correlation incentives offers a promising path to improve protocol decentralization. By penalizing correlations between misbehaving actors, staking protocols can foster a healthier and more sustainable ecosystem. Further research and experimentation in this area will contribute to the evolution of decentralized blockchain networks such as Ethereum.
Image source: Shutterstock