Idea Propagation in Crypto: A Case Study on MEV
-------------
Thesis -> Antithesis -> Synthesis
Auctions -> FIFO -> CR Auction -> Proposer commitments
-------------
Problem: How to order transactions in a block?
--- Afaik Flashboys 2.0: first detailed description of problem scope by
@phildaian (phd work with
@AriJuels):
arxiv.org/abs/1904.05234
-------------
Thesis:
Idea: Auction the block.
Protocol: Validators sell the block to the highest bidder
Manifestation: MEV-geth, MEV-boost
@phildaian key proponent
-------------
Antithesis:
Idea: Order transactions based on whichever transaction is received by a majority of validators earlier (FIFO = first in first out).
Protocol: Validators each submit their local ordering of the blocks, and the final ordering aggregates all the local ordering.
Manifestation: Theory - Aequitas (
eprint.iacr.org/2020/269), Practical Protocol: Themis (
eprint.iacr.org/2021/1465)
@_mahimna was key author of these two papers. He was a student of
@AriJuels as well :)
@soubhik_deb and I collaborated with
@_mahimna and
@AriJuels on Themis.
---------------
I was on the FIFO side and we used to have some heated discussions across the two groups (the auction gropu was the dominant one) :)
Here is a fun story. At the stanford conference Sep 2022, there was a panel with
@drakefjustin @VitalikButerin @fradamt @phildaian and
@zmanian
Someone asked "why does fifo make sense" and
@VitalikButerin answered, tongue in cheek, "thats like asking a panel of vegans why factory farming makes sense" conveying the Auction-centric nature of this group :)
youtu.be/Z9VCdiSPJEQ?si=uTQu…
---------------
Synthesis:
- Idea: Censorship Resistant "Pre-Block", Auction to find the best ordering.
- Protocol: Include all transactions sent to a majority of users (using Themis-like protocol). Order them via auctions.
- Benefits: Prevents some types of toxic MEV (excluding liquidations via bribing etc), while still capturing CEX-DEX arbitrage MEV etc.
Most vocal proponent:
@maxresnick :)
He even made a T-shirt called "End the proposer monopoly"
Manifestation: Multiple-proposer based ordering is being proposed in
@ethereum ,
@solana, and now pretty much everywhere else.
---------------
It is exciting to see these ideas converge!
---------------
My vote for the best protocol for high-throughput censorship resistance is a paper by my former student Bowen Xue (who is a contributor to
@eigencloud) created called BigDipper. It is a clearly highly biased vote.
arxiv.org/abs/2307.10185
---------------
If anyone is interested in experimenting with consensus protocols rapidly on
@eigencloud, please DM me.
---------------
Super Synthesis:
---------------
Beyond CR Auction, there are many more complex methods for proposer commitments.
---------------
Beginning from our work on MEV-Boost (
youtube.com/watch?v=ywJNXIUS…)
to our work on multi-commitment systems (
youtube.com/watch?v=zTPKnKsf…)
Now there is a thread of work led by
@barnabemonnot at EF, called PEPC (protocol-enforced proposer commitments).
---------------
Many of these commitments can be built out of protocol on
@eigencloud. Since EigenLayer uniquely supports native restaking, it allows all validators to equally receive rewards on such MEV enhancements, satisfying
@ethereum non-centralization considerations. A new
@eigencloud upgrade (
x.com/eigenlayer/status/1831…) made native restaking even more affordable!
---------------
With new technologies coming up like
@Commit_Boost for adding arbitrary new sidecars, there is a lot of scope for experimenting and building new MEV management techniques both for the L1 as well as for L2s. If you are interested in building MEV management on
@ethereum and want to explore
@eigencloud please DM me.