Centralizing Ethereum mining could make the network healthier – yes.

Flashbots has successfully on-boarded some 80% of hashrate to itself, with another percentage using variants of Mev-Geth software as well.

The remaining percentage, often termed rogue miners, operate their own sequencing projects or incentivize priority gas auctions (PGA) by not sequencing transactions. PGAs can be thought of as a negative externality pushed onto network users by driving up the average gas fee. MEV, on the other hand, is showing early signs of alleviating this pressure.

So, what if Flashbot miners could force the rogue miners to use Flashbots? Such a move would threaten the status quo, but could prove to be beneficial for the chain. Two methods are available: Orphaning blocks of non-Flashbot miners and/or time-bandit attacks against non-Flashbot miners.

  • Orphaned blocks are blocks that are no longer part of the heaviest chain. These blocks receive no reward (outside of variable Uncle rewards) for the miner’s work or transaction fees as they are excluded from the chain. Flashbot miners could collude to exclude non-Flashbot miners by orphaning blocks until they implement Mev-Geth.
  • Time-bandit attacks are what is considered a blockchain re-organization. Here, colluding miners can build an alternate chain outside of the purview of other miners on the network, similar to selfish mining. They can sequence the chain’s history differently – such as removing non-flashbot miner transactions – to increase extractable value on their chain.

As the seminal MEV paper Flashboys 2.0 puts it:

“[T]ime-bandit attacks can be subsidized by a malicious miner’s ability to rewrite profitable trades retroactively, stealing profits from arbitrageurs and users while still claiming gas fees on failed transactions that attempt execution.”

Past research has concluded that all proof-of-work (PoW) blockchains are susceptible to time bandit attacks. The math involved is simple: If the coinbase rewards over the earlier period are less than the transaction fees over the same period, then an attack to extract possible MEV is economically rational.

Of course, there are a few asterisks involved as noted by “A model for Bitcoin’s security and the declining block subsidy.” Time-bandit miners must account for the depreciation of other assets on hand caused by an attack on the network. The chain’s asset, hardware on hand and assets extracted may decrease in value, among other things.

Which brings us back to block censorship with one question: Would a miner led initiative to censor miners who create negative externalities for the chain be accepted by the larger community?

Historically, any Ethereum miner collusion has been viewed with scrutiny. For example, miners planned a “show of force” on April 1, 2021 to show displeasure with July’s EIP 1559. Negative press and souring relationships with core developers dissuaded miners against it, but not before pushing developers into moving to proof-of-stake (PoS) early.

However, miners colluding against other miners in favor of other end users is characteristically different, even if it does constitute an attack on the Ethereum network. Lowering the cost to interact with the network could be viewed favorably and, to some degree, is validated by Ethereum’s well known philosophical persuasion toward breaking changes.