Opinion: Pros and Cons of "Frontrunning" on Ethereum
Editor's Note: This article comes fromEthereum enthusiasts (ID: ethfans)Editor's Note: This article comes from
Ethereum enthusiasts (ID: ethfans)
Ethereum enthusiasts (ID: ethfans)
, Author: Ed Felten, translation & proofreading: Min Min & A Jian, reprinted by Odaily with authorization.
Vitalik Buterin responded to my article with an argument that I don't think is true, which I will explain later. I later had a long discussion with Vitalik on Telegram and finally understood his point of view. Long story short, I still don't think he's right.
Vitalik believes that there must be a certain miner who is better at frontrunning transactions than other miners, and seeks economic benefits from it, thereby increasing the proportion of his computing power and eliminating other miners who are not good at frontrunning transactions. In this way, the trend of centralization of computing power will be formed, which will destroy the stability of the network.
secondary title
When frontrunning deals become a service
image description
-Sorry, I announce that you can now get high-end customization services from my side-
I do not share this view. why? Let's say Fred is the best frontrunner in the entire world. He has a deep understanding of the way Ethereum blocks are packaged, and his ability to profit from front-running transactions is unmatched. Also assume that Alice is a miner who controls 2% of the computing power. Naturally, Fred will sell the frontrunning deal to Alice as a service. If Alice purchases Fred's service, Alice's income will increase by V every time she digs a block, and then she pays Fred no higher than V. In this way, both parties will benefit, and Alice and Fred certainly will not refuse. Alice is nothing special here. The problem is that Fred will sell this service to every miner.
(Alternatively, suppose Fred mines himself, but has no cost advantage over other miners. Fred could offer front-running transactions for his mining operations at a discounted price or even for free, but doing so would do Fred no good: a $X discount means This means that Fred will earn $X more for his mining business, but he will earn $X less for his front-running business. Therefore, if Fred also buys front-running trading services for his mining business like Alice did, then he is affected by The incentives are the same.)
The end result is that Fred will sell his front-running service to all interested miners. Other miners will also buy.
As a result, front-running trading as a service (FRaaS) has become an inevitable result: in this active market, front-running trading experts Freds are vying to sell block packaging services to Alices. This does not drive mining centralization at all. FRaaS allows Alices to hand over to Freds the problem of how to use front-running transactions to increase revenue, thereby lowering the entry threshold for mining. Therefore, this is in favor of promoting the decentralization of mining.
FRaaS and mining pools
Ultimately, the entire market will consist of a group of mining pools. Fred will run a mining pool that can pay miners a higher rate of return than other mining pools because Fred's mined blocks will always generate higher returns than other mining pools. In the FRaaS market, Fred's competitors will do the same. Most miners will choose the honest pool that pays the highest yield.
Importantly, miners are free to choose which mining pool to work with, avoiding misbehaving pool operators, and can switch to other mining pools every time a block is mined. Therefore, the mining pool operator's power to determine the order of transactions can only be used on one block at a time. Of course, Ethereum mining has given the miners who packaged blocks the power to determine the order of transactions. Therefore, the "monopoly power to order transactions within a single block" that needs to be earned on the open market is exactly our status quo, not something that violates the agreement.
secondary title
How to Design a System for FRaaS
If the FRaaS industry will be formed in the future, how should we design a more robust system? I have two principles.
1. Maintain competition in the FRaaS industry
The first principle is that if the FRaaS industry will be formed in the future, we hope that there will be internal competition in it, so that the right to sort transactions will not be monopolized by a certain party. Generally speaking, monopolies usually have more power than market leaders in competitive markets, and they often use their power to capture value, resulting in less efficient markets and loss of end users. It can be seen from this that we should not artificially give a certain party the monopoly right to pack blocks. Otherwise, this leads to centralization, compromised community principles, and abuses of power that would not be possible in a competitive marketplace. Even if a monopoly is only temporary, it will only lead to unnecessary centralization of power.
The second principle is that we have to remember that front running is harmful and should be minimized. We should take frontrunning resistance as one of the criteria for evaluating systems and protocols.
For example, when evaluating any new protocol, we should ask ourselves this question: Will this protocol increase or decrease the front-running rights of miners or other participants?
In the future, I will detail how this principle plays out in our Arbitrum protocol design, and why we believe that a Layer 2 solution that actively decentralizes transaction ordering is superior to Ethereum.
