The debate over MEV definition and classification: Who has the final say?

Tina, founder of Flashbots, said on Twitter that the word MEV has been abused, and 99% of discussions about MEV use this term casually, either to describe the problems that the product can solve, or to criticize the Ethereum phenomenon they don't like, ignoring its formal definition. So should the definition of MEV cover the potential extractable value, or the actual extracted value? Are all extraction behaviors counted, or are only harmful extractions considered MEV?

Views from all sides: The definition of MEV

Back to last week, Fastlane CEO Thognad expressed his views on the definition of MEV on Twitter. He believes that the MEV problem cannot be solved at all, because as research deepens, new ways of extracting value will always be discovered and incorporated into the definition of MEV, making this definition continue to expand. This view has sparked discussion in the community.

In response, Nathan Worsley, consultant to Flashbots, said that the definition of MEV should be limited to the value that block proposers can observe and exploit before the block finally becomes part of the consensus. If the definition is too broad, it will lead to management complexity. However, Thognad feels that this limitation only focuses on the role of block proposers in MEV and ignores other possible ways of extracting value.

Then, Ethereum researcher mteam joined the discussion, arguing that MEV is any value that can be extracted from a privileged position within the system. Thognad further refined this definition: Any player who has a privileged position within the system and actively exploits system vulnerabilities or design flaws through this position to obtain additional value beyond what passive players can obtain should be considered MEV. And proposed a baseline concept of "zero MEV", that is, if a block builder only selects transactions from the public transaction pool according to transaction fees and time order to build blocks, without performing any other operations, this should be regarded as zero MEV. mteam responded that because the blockchain is distributed, there are differences in network speed and connection quality, and different nodes may see different transaction orders and contents, so it is difficult to set a unified standard. He believes that the only useful definition is "harmful MEV".

Tina retweeted the question mentioned by mteam in the discussion: Should MEV be regarded as the value that can be theoretically extracted from the blockchain system, and REV (Realized Extractable Value) as the actual extracted value distinguished from the theoretical value. Tina pointed out that the discussion on MEV on Twitter usually ignores the formal definition of MEV, and uses the term more arbitrarily based on personal opinions and convenience, which has caused confusion in the market's definition and understanding of MEV.

Tina further mentioned that a GitHub issue discussion on MEV classification (closed in December 2021) may provide a systematic solution to clarify the definition and classification of MEV. And consider whether this discussion should be reopened to better clarify the definition and classification of MEV, thereby promoting a deeper understanding of MEV.

Discussion on MEV classification

Although the above GitHub discussion failed to reach a final unified conclusion, it proposed multiple directions to improve the definition and classification of MEV. It can be divided into the following key points:

1. How to clearly distinguish between potential extractable value and actual extracted value:

• Extractable Value: refers to all possible value opportunities that can be identified by analyzing the transactions of the previous block in the current state of the blockchain. This includes the maximum value that can be extracted theoretically through transaction reordering, insertion, etc.

• Extracted Value: refers to the value actually extracted in a specific block, that is, the value obtained by miners or other participants through actual operations (such as transaction reordering). This is a practical and quantifiable value.

2. How to distinguish good from bad MEV:

• For example, arbitrage operations may be considered "good MEV" by helping the market reach equilibrium, while manipulating the order of transactions to cause market slippage or unfair trading may be considered "bad MEV". Such distinctions help identify the social impact of MEV and propose corresponding countermeasures for different types of MEV.

3. Redefine terms:

• Some participants proposed introducing new terms, such as realized extracted value (REV), to clearly distinguish between the value that has been actually extracted and the theoretically extractable value. Dark MEV, to describe those manipulation behaviors that are difficult to detect, such as manipulation through covert market intervention.