Insightful Mining Equilibria

Mengqian Zhang, Yuhao Li, Jichen Li, Chaozhe Kong, Xiaotie Deng

Introduction

Since the introduction of Bitcoin, the assumption of strict incentive compatibility has been challenged. Selfish mining, first described by Eyal and Sirer in 2013, allows a large mining pool to strategically withhold blocks, forcing honest miners to waste computational power on stale branches and allowing the attacker to earn disproportionate rewards. While much research has focused on optimizing selfish mining, less attention has been paid to how victims can strategically defend themselves.

This paper introduces "insightful mining," a strategy where a mining pool uses an undercover "spy" to monitor a selfish pool's hidden blocks. By transforming private information into asymmetric information, the insightful pool can consistently outperform selfish miners. This research provides a robust framework for analyzing the Nash equilibria of blockchain mining as a multi-agent game.

Methodology and Strategic Framework

The core of the insightful mining strategy is the "undercover miner." An insightful pool plants a spy within a suspected selfish pool. Because the selfish pool manager must assign new mining tasks based on its private, hidden blocks, the spy can deduce the number of hidden blocks by observing the "previous hash" in assigned tasks.

Research roadmap for Insightful Mining Strategy and Game.

The authors model this system using several advanced techniques:

  1. Two-Dimensional Markov Reward Processes: The state of the blockchain is modeled based on the lead of the selfish pool over the honest pool ($x$) and the lead of the insightful pool over the selfish pool ($y$).
  2. Game Theory: An $n$-player normal-form game is formulated where miners choose between "Refined Honest Mining" and "Insightful Mining."
  3. Markov Decision Processes (MDP): Used to determine the optimal actions (wait, match, or override) for the insightful strategy.
  4. Simulations: Discrete-time random walk simulations of 2 billion steps validate the theoretical proofs.
Flowchart of the mining competition involving honest, selfish, and insightful pools.

Key Insights and Theoretical Results

The fundamental "aha moment" is that selfish mining's advantage relies on secrecy. By using a spy, the insightful pool can strategically decide when to mine honestly to negate the selfish pool's lead.

The Dominating Theorem

A key theoretical contribution is the Dominating Theorem. The authors prove that if a selfish pool and an insightful pool possess identical hashing power, the insightful pool will always earn a strictly higher expected revenue. Simulations further show that an insightful pool can outperform a selfish pool even with less hashing power.

Equilibrium Stability

The research characterizes the Nash equilibrium for a network of $n$ strategic pools:

  • Honest Mining Equilibrium: Stable only if no single pool controls more than 1/3 of the total hashing power ($m_1 \le 1/3$).
  • Single Insightful Pool Equilibrium: Occurs if $m_1 > 1/3$ and the second-largest pool is relatively small.
  • Dual Insightful Pool Equilibrium: Occurs when both the first and second largest pools are sufficiently large. Interestingly, the model predicts that no more than two insightful pools will exist in a stable equilibrium, regardless of power distribution.

Discussion and Limitations

While insightful mining offers a powerful defense, it relies on certain assumptions. The model assumes the spy's hashing power is negligible and that the selfish pool cannot easily detect or eject them. In practice, selfish pools might implement counter-espionage measures. Furthermore, the current model focuses on block rewards, whereas a future dominated by transaction fees would introduce significantly more complexity.

Future work may explore strategies where all pools mine behind the latest hidden block—even without knowing its full contents—potentially rendering selfish mining completely ineffective.

Summary updated

BibTeX
@inproceedings{zhang2022insightful,
  author = {Mengqian Zhang and Yuhao Li and Jichen Li and Chaozhe Kong and Xiaotie Deng},
  title = {{Insightful Mining Equilibria}},
  booktitle = {WINE},
  year = {2022},
  doi = {10.1007/978-3-031-22832-2\_2},
}
Figure