Ittay Eyal and Emin Gün Sirer Cornell University showed that selfish mining—withholding blocks to gain an advantage—can become profitable for rational miners under typical proof-of-work rules. Subsequent work by Yonatan Sompolinsky and Aviv Zohar Hebrew University proposed the GHOST family of chain-selection rules to reduce such incentives, and Ethereum implemented related mechanisms through uncle inclusion following recommendations from Vitalik Buterin Ethereum Foundation. These studies together indicate which protocol changes most effectively curb block-withholding strategies.
Protocol-level defenses and why they work
Changing the chain selection and reward structure directly targets the chief cause of selfish mining: the outsized marginal value of private blocks relative to published blocks. The GHOST rule selects the heaviest subtree rather than the longest linear chain, which reduces the payoff for hiding blocks because withheld branches do not gain the same advantage when weight from multiple children is counted. Similarly, uncle inclusion and partial rewards for stale blocks reduce the penalty for honest miners whose blocks arrive late, narrowing the profit gap that selfish strategies exploit. Eyal and Sirer Cornell University also discussed tie-breaking rules; adopting uniform tie-breaking so miners randomly choose between competing tips removes predictable network-propagation advantages used by selfish pools.
Consequences, trade-offs, and social relevance
These protocol changes reduce incentives for centralization by lowering the return on collusion and block withholding, which has cultural and territorial implications where mining clusters around low-cost energy sources. Reduced selfish-mining profitability helps maintain a more distributed mining base, decreasing leverage that large pools can exert on local grid use and regional policy. There are trade-offs: GHOST-style rules increase protocol complexity and may alter confirmation finality characteristics, and uncle rewards change economic dynamics of fees versus subsidies. Implementation choices reflect priorities between simplicity, security under high throughput, and equitable reward distribution.
Empirical and analytical work by Eyal and Sirer Cornell University and by Sompolinsky and Zohar Hebrew University provides the strongest evidence that chain-selection and reward redistribution are the most effective levers. Protocol designers should weigh those benefits against engineering complexity and possible new attack surfaces, and communities should consider governance mechanisms that align incentives as well as technical fixes.