Selfish Mining

💡 The Plain-English Definition

Selfish mining is a theoretical attack strategy where a mining pool withholds newly found blocks — mining privately on a longer chain — then releases them at a strategic moment to invalidate competitors’ work and claim a disproportionate share of block rewards.

🤔 But Why Though?

Standard mining is honest: when you find a valid block, you immediately broadcast it. This lets the whole network build on your block, and you earn your reward. Selfish mining deviates: when you find a block, you keep it secret and immediately start mining the next one on your private chain. If you find a second block before the honest network finds their next block, you now have a two-block lead. When you release both blocks simultaneously, the honest network’s current block (if they found one) becomes an orphan block (a valid block rejected because a longer chain exists), and your two-block chain is accepted. You’ve earned two block rewards while the honest miners earned nothing for their last block.

Researchers Ittay Eyal and Emin Gün Sirer demonstrated in 2013 that a pool with greater than approximately 33% of total hash rate (total computing power devoted to mining) could profit from selfish mining even accounting for the risk of their private chain being overtaken before release. Below 33%, the strategy is unprofitable. Above it, selfish miners can earn more than their fair share while simultaneously reducing the effective revenue of honest miners — creating a centralisation pressure where rational miners join the selfish pool to avoid being victims. Why hasn’t this happened at scale? Several reasons: maintaining a secret chain requires keeping it completely private within a pool, which is difficult and risks defection. The strategy is detectable with network analysis. And reputational damage from being identified as a selfish miner would likely cause a mass exodus of participants. At current mining pool sizes, several pools operate near or above the 33% threshold intermittently — making this a real concern worth monitoring, even if it hasn’t been actively exploited.

🌍 The Real-World Analogy

Think of selfish mining like a student in an exam who secretly looks up answers but waits to write them until they see a classmate finishing — then rapidly fills in their answers before time is called, invalidating the classmate’s efforts by submitting a “better” paper at the last moment. The honest student did real work but gets credit for none of it. The cheating student’s advantage grows if they can do this repeatedly and quickly.