Author: Shijun

Preface

Every so often, the blockchain community reignites a major discussion about "decentralization": some are still defending the ideal, some have already given up and say this has always been a false proposition, while others have long turned to a realistic route focused on performance, profitability, and regulatory friendliness. Yet, we forget why we embarked on this path in the first place.

But I have always felt that we have not truly understood one question: when we say "decentralization," what exactly are we avoiding? What kind of forces are resisting its move towards centralization, and what kind of logic allows it to continuously self-correct?

Over the past few years, I have observed a particularly ironic trend—chains that are labeled as decentralized are gradually giving rise to new centralized agents, new power structures, where former challengers become new vested interest groups, and then transform into the very dragons they once opposed.

They are not companies, not banks, not governments—they could be miners, MEV seekers, block builders, or even the chains themselves.

Choosing MEV (Maximum Extractable Value), in my view, is because it is the most authentic and naked mirror of this ecosystem.

It pulls blockchain from the pure world of mathematics and cryptography back into the real dimensions of game theory, institutional design, and even power politics.

No matter how decentralized a chain is, as long as there exists differentiated power in transaction ordering and block packaging, MEV will not disappear; it may even intensify, becoming more covert and systematic.

Thus, I decided to write this article, not just to "popularize" MEV, nor merely to summarize existing solutions, but to attempt to delve deeper from an evolutionary perspective—using a critical lens to sort out the structural issues and evolutionary mechanisms behind MEV, clarifying how it reappears in every chain and every round of technological upgrades, taking on different roles and institutional forms, and continues to occupy the center stage in a new guise.

We all say "code is law," but MEV shows us another side of the reality: "power is law." Whoever controls the ordering power controls the distribution of on-chain information flow, the decision-making power of transaction execution, and even the redistribution of on-chain wealth.

In this article, I want to start from the following questions:

• Why is MEV not an incidental "arbitrage behavior," but a structural phenomenon?
• How does it undergo role migration with the evolution of protocol design, consensus mechanisms, and on-chain economic structures?
• What exactly do Ethereum's PBS and mev-boost "solve," and what do they leave behind?
• How does the design of "high-performance chains" like Solana change the form and participant structure of MEV?
• Is it really possible for us to "eliminate MEV"? Or can we only coexist with it and try to tame it?

Only by deeply understanding MEV can we truly grasp the deep institutional logic of blockchain.

This is not only about balancing fairness and efficiency but also about whether our industry is heading towards a "neoliberal high-speed financial laboratory" or still retaining some kind of "open, neutral, and anti-censorship" idealistic spark.

II. Ethereum: From Dark Forest to Layered Game, How Has MEV Evolved?

In fact, the term MEV can be misleading because people tend to think it is miners extracting this value. In reality, the MEV on Ethereum is primarily captured by DeFi traders through various structural arbitrage trading strategies, while miners only profit indirectly from the transaction fees of these traders.

It is difficult to pinpoint exactly when the term "MEV" was invented, but it is certain that it was never some kind of sudden mechanism, but rather a "pain point that had to be named."

As early as 2020, Paradigm wrote that now-classic article "Escaping the Dark Forest," comparing the entire on-chain trading environment to a dark forest filled with undercurrents. The suffocating feeling of "you just finished writing an arbitrage transaction ready to send out, but before it spreads in the mempool, it gets front-run by another bot at a high price" is a fear familiar to anyone who has navigated the depths of DeFi.

If a hacker is foolish, they will directly execute profit-seeking methods and get front-run by hunters.

If a hacker is smart, they might adopt a contract-over-contract (i.e., internal transaction) approach to hide their ultimate profit-seeking transaction logic. Unfortunately, the outcome is not like that of the author of "Escaping the Dark Forest," who ultimately succeeded, but rather they still get front-run.

This also means that hunters not only analyze the parent transactions on-chain but also analyze every child transaction, simulating profit scenarios. They even further detect the deployment logic of gateway contracts and replicate it, and this is done automatically within seconds.

Before this, people might have vaguely sensed that on-chain transactions were not equal, but had not systematically summarized it, and the emergence of the term "MEV" was the first time this structural inequality was clearly named.

The now-defunct model "Time Bandit Attack" existed before Ethereum transitioned to POS; if reorganizing a single block yielded extremely high profits, miners could form alliances to create a higher blockchain to replace the already mined ledger. For example, the author of "Escaping the Dark Forest" discovered a $9.6 million vulnerability on-chain, ultimately contacting privacy transactions to mine blocks. Even with white hats assisting throughout, their main concern was not just the risk of transaction leakage being front-run, but also the type of malicious miners discovering and forcibly constructing large-scale reorganizations (do not think reorganizations are rare; the author once ran a BSC node and found that even in this super node model, there would be daily reorganizations five times when the block height was insufficient).

“Rescue of 25,700 ETH”

This is also why MEV creates a tense relationship with decentralization.

Originally, we thought chains were open, transparent, and decentralized, where anyone could participate in transactions, only looking at the rules and not the identity. But the reality is, you can have an address, but you may not have the ordering power; you can send transactions, but you may not be packaged before others; you can be an ordinary user, but some "Searcher" can always profit from you. This "structurally deficient fairness" is essentially a "power misalignment": the power of ordering is not in the hands of users, yet it affects their costs and destinies.

And this misalignment is the starting point of the story of MEV's evolution on Ethereum.

2.1 First Stage: Chaotic MEV Free Market (2018–2021)

In the early Ethereum chain, MEV was a battlefield without unified rules.

Miners were responsible for packaging transactions, while users could only send transactions hoping to be timely included on-chain. Theoretically, everyone could compete for fair inclusion based on gas fees, but in reality, it quickly became apparent that if I could monitor the mempool, I could immediately copy and slightly increase the gas of someone else's arbitrage transaction and package it first, thus reaping the arbitrage profits for free.

This is known as "front-running," also seen as the prelude to sandwich attacks.

At this time, the most active role was the Searcher—they were like a pack of wolves, constantly watching every transaction in the mempool, simulating results, automatically judging arbitrage opportunities, and sending their constructed front-running transaction packages at a very high frequency, striving to be prioritized by miners.

Miners were not passive either. Initially, they only passively collected higher gas fees, but gradually some miners directly bypassed Searchers, deploying MEV bots themselves to participate in arbitrage, cutting others out.

Thus, we entered the so-called "miners self-extracting MEV" phase. Some miners even sold packaging rights—openly stating, "I can put your transaction first, as long as you share the profits with me."

The most prominent characteristics of this period were:

• The ordering power was not layered; miners were both consensus and ordering roles;
• MEV was chaotic; anyone could front-run, forming a free market of competition;
• Users had no voice; the very openness of their transaction information was a form of deprivation.

Paradigm's article "Dark Forest" sparked heated discussions not only because its writing resembled science fiction but also because it captured an essence: blockchain transactions before being mined are actually in a state of "information openness but unconfirmable suspension," and the essence of MEV is to exploit this "deterministic lag" to gain structural profits.

But continuing down this path was not a solution. Miners front-running, Searchers harming each other, and users bleeding out—the overall user experience on the chain was deteriorating.

Thus, we entered the next stage.

2.2 Second Stage: Flashbots and the Attempt at "Extractable Fairness" (2021–2022)

Flashbots proposed a seemingly illogical but very Ethereum-style solution: since front-running is inevitable, why not turn front-running into a standardized service?

Thus, the Flashbots Auction system was born—Searchers no longer monitored the mempool to speculate alone but submitted their constructed "transaction bundles" to Flashbots, which organized multiple bundles into blocks and auctioned them to miners (or later to validators), who only needed to choose which block was the most profitable.

This mechanism solved at least two issues:

1. Responsibilities were layered: Searchers were responsible for proposing profit schemes, Builders (Flashbots) were responsible for organizing transactions, and miners were responsible for block production. Responsibilities were clear, reducing the incentive for miners to participate in speculative arbitrage.

2. User privacy was protected: Many transactions no longer entered the public mempool but went through private channels (Flashbots Protect), avoiding public dissemination and front-running.

At this point, MEV entered the "mechanized mining" stage. Flashbots became the core infrastructure, and the vast majority of organized Searchers and Builders participated in block construction through it.

However, this did not mean the problems were resolved.

Because Flashbots itself is a centralized system, and only transactions following its route could potentially avoid being front-run, this left "fairness" still unresolved, merely hiding it in a different way.

The real turning point came from "The Ethereum Merge."

The Ethereum Merge refers to the upgrade of its consensus mechanism from POW to POS, where the final merging scheme was based on the lightest reuse of the pre-merge Ethereum infrastructure while separately isolating the consensus module for block decision-making.

For POS, blocks are generated every 12 seconds, rather than the previous fluctuating values. The block mining reward has decreased by about 90%, from 2 ETH to 0.22 ETH.

This is very important for MEV for the following two reasons:

1. The block interval of Ethereum has become stable. It is no longer the previously random and discrete situation of 3-30 seconds, which has both advantages and disadvantages for MEV. Although Searchers no longer need to rush to see slightly profitable transactions and can continuously accumulate a better total sequence of transactions to submit to validators before block production, it has also intensified competition among Searchers.

2. The reduction in miner incentives has encouraged validators to be more willing to accept MEV transaction auctions, leading to MEV reaching a 90% market share in just 2-3 months.

The actual impact is also significant.

In the year before the merge, the average profit calculated from MEV-Explore was 22MU/M (from September 2021 to September 2022, before the merge, the values included Arbitrage and liquidation models).

In the year after the merge, the average profit calculated from Eigenphi was 8.3MU/M (from December 2022 to the end of September 2023, the values included Arbitrage and Sandwich models).

The conclusion regarding the change in final returns is that after excluding hacker events that should not be attributed to MEV from the above data, the overall return rate has significantly decreased by 62%. Note that since the statistics from MEV-Explore do not actually cover sandwich attacks data and include liquidation profits, the decline in pure Arbitrage comparisons may be even greater.

So why is this the case? It has to do with the mechanisms after the merge.

2.3 Third Stage: Post-Merge, New Complexities from Block Separation (2022–2024)

Flashbots proposed a seemingly illogical but very Ethereum-style solution: since front-running is inevitable, why not turn front-running into a standardized service?

In 2022, Ethereum completed its historic merge, officially transitioning from PoW to PoS. This was a revolutionary event at the consensus level, and for MEV, it was a "structural earthquake."

With the introduction of PoS, block production became predictable (every 12 seconds), the miner role disappeared, replaced by Proposers and Validators, and the powers of ordering and block production were further separated.

This is known as the PBS (Proposer-Builder Separation) mechanism.

https://twitter.com/vitalikbuterin/status/1588669782471368704

Flashbots intervened again, launching MEV-Boost—a modular extension that allows validators to "outsource" block construction rights to third-party Builders while receiving "build bids" from different Builders through relays, choosing the optimal plan for block production.

The entire process is as follows:

1. Builders create a block by receiving transactions from users, Searchers, or other (private or public) order flows.
2. Builders submit the block to the relay (which has multiple Builders).
3. The relay verifies the block's validity and calculates the amount it should pay to the block producer.
4. The relay sends the transaction sequence package and profit price (also the auction bid) to the current slot's block producer.
5. The block producer evaluates all the bids received and selects the sequence package that offers the highest profit.
6. The block producer sends the signed header back to the relay (completing this round of the auction).
7. After the block is published, rewards are distributed to Builders and Proposers through the transactions within the block and the block reward.

Market share shows rapid growth.

https://mevboost.pics/

From then on, the lifecycle of MEV transformed into a "supply chain model." On-chain games entered an era of "layered specialization":

• Searchers need to optimize algorithms and order flows;
• Builders need to optimize resources, stability, and simulation speed;
• Relays act as "data messengers," needing to be neutral and efficient;
• Validators only look at bids, trying to interfere as little as possible to avoid regulatory pressure.

This model is, in some sense, successful—by the end of 2023, Flashbots held over 90% of the MEV-Boost market share, forming a de facto "off-chain ordering hub."

However, it also brought two new problems:

The first is the centralization issue. The number of Builders and Relays in Flashbots is very limited; currently, over 90% of block construction comes from four Builders, which has led the originally decentralized structure back towards an oligopoly.

“Distribution Trend of Builders Involved in Block Production ”

The second is the lack of incentive issue. Relays essentially do not charge fees but bear the costs of computation, storage, and network. Several Relay operators, such as Blocknative, have announced their exit, and the exit of long-tail participants further exacerbates the risk of centralization.

We can actually understand why this system has evolved in this way.

This is not only due to the complexity of building off-chain systems but also because of the competitive nature of information silos. In the face of the ubiquitous resistance of MEV, some have gradually emerged, abandoning the AMM mechanism in favor of auction bidding protocols, such as UniswapX (which essentially sacrifices the real-time nature of trades for better exchange prices), allowing ordinary users not to search for the best trading paths on the client side but letting professional players combat MEV. Besides the specialized path simulations and off-chain trading willingness matching, UniswapX also encourages quote providers to build private order flows to create low-competition situations for MEV, avoiding excessive competition for profitable orders that would benefit ordinary users, ultimately reducing the experience gap between DeFi and CeFi.

In this context, private order flows will inevitably not be completely private. When Searchers discover profit delivery traps, Builders can also discover and even front-run transactions, constructing block sequences. Of course, for those who always front-run transactions, they will gradually lose the strategy discovery capabilities from Searchers (Builders are not unique; overly opaque Builders will be blacklisted by Searchers), so unless profits are extremely high, they will still engage in profit-sharing to maintain a delicate balance.

In terms of system structure, they can replace each other, ultimately all revolving around order flow. Searchers will hope to gradually expand their profit margins, which requires their private order volume to be large enough (ultimately, the constructed block profits must be high), thus gradually taking on the role of Builders.

In terms of regulation, there is relatively more breathing room.

As the cryptocurrency industry matures, regulation is inevitable. All entities registered in the United States and their operations of Ethereum POS validators must comply with OFAC requirements. However, the system mechanisms of blockchain ensure that it will not only exist in the United States; as long as there are other relays that comply with local policies, it can ensure that at some opportunity, it can be propagated on-chain.

Even if over 90% of Validators review transactions routed through MEV, those anti-censorship transactions can still be included on-chain within an hour, so as long as it is not 100%, it is equivalent to 0%.

“Block Distribution Map Complying with PFAC Requirements ”

2.4 Fourth Stage: Decentralized Builders on Trusted Hardware (2024-~)

If the initial MEV-Boost revolution led by Flashbots was an "open dismantling" of centralized miner power, then the emergence of BuilderNet attempts to decentralize this dismantling itself—providing a deeper structural response to the question of "who will be the builder."

Since the launch of the BuilderNet project by Flashbots at the end of 2024, a certain directional branch of Ethereum MEV has entered a new stage based on TEE (Trusted Execution Environment) and in the form of a decentralized builder network.

Mechanically, this is not just another product update; it may signal the redistribution path of block ordering power in the coming years.

BuilderNet is essentially a decentralized Ethereum block construction network that operates on TEE and shares MEV with the community.

【BuilderNet Architecture Diagram】

It involves roles: Builder node (TEE), Node operator (managing block builders), Flashbots infrastructure, MEV-Boost relay (final blocks to them), Order flow source (wallets, users, Searcher bots, etc.)

Its core is the multi-operator concept.

Multiple parties can operate the same block generator.

Each operator runs an instance of an open-source generator in a Trusted Execution Environment (TEE), and order flow providers (such as applications, wallets, users, and Searchers) can verify and send encrypted order flows to it.

Each instance shares the order flows it receives with other instances in the network and submits blocks to the MEV-Boost relay as usual.

When a BuilderNet instance wins a block, it will calculate refunds based on the value added to the block by order flow providers and distribute them to the order flow providers.

This forms an economic closed loop, striving for order flow through user feedback.

Detailed Explanation of the Operation Mechanism under TEE

BuilderNet serves as a systematic response to this issue, advocating a clear and direct approach: sharing order flows among multiple builder nodes running in TEE, using open-source, verifiable sorting code to construct blocks, and then connecting to the final block production path through MEV-Boost.

In brief, its working logic is as follows:

• Each builder node operates in trusted hardware such as Intel SGX or Azure TDX, and external parties can verify through remote attestation mechanisms that the sorting algorithms it runs have not been tampered with.
• Order flow providers (wallets, DApps, bots, etc.) can send transaction flows to these TEE builders through encrypted channels (HTTPS/TLS), ensuring that their strategies are not leaked.
• All TEE nodes share order flows within the BuilderNet network, but the information is only readable within the TEE, preventing the builder nodes themselves from peeking, thus ensuring the protection of order flow privacy.
• After a winning block is generated, the system allocates refund incentives to the order flow sources based on the construction value, allowing users or wallets to receive subsidies.

In other words, BuilderNet allows builders to be "depersonalized"—the nodes are not companies or exchanges, but verifiable instances of sorting algorithms.

BuilderNet is conceptually very similar to the previous Flashbots MEV-Share; it can be said that BuilderNet is an "evolution of MEV-Share based on trusted hardware."

Through TEE and a multi-node collaboration mechanism, BuilderNet has made a more thorough structural decentralization attempt regarding the original sorting rights issue. It does not trust any single organization or server but instead allows code and hardware to assume the role of trust.

From an adoption perspective, BuilderNet currently has a limited market share on the Ethereum mainnet, according to Dune data, having constructed only a small number of blocks. However, this is not surprising:

• The construction process is slower; the TEE execution environment has performance bottlenecks compared to native kernel operations.
• Multi-node sharing of order flows means increased communication load, significantly challenging real-time performance.
• Users and wallets have not yet widely accessed its order flow channels (although the refund addresses have been open-sourced and are operational).

Importantly, its system logic loop has been established, and every component (node registration, remote attestation, order flow encryption, block auction, refund calculation) has been open-sourced and deployable.

A prototype of a "decentralized builder network" has already materialized.

Summary

Is MEV revenue really declining?

According to EigenPhi data, in the year following the Ethereum merge, the average MEV profit was about 8.3 MU/M, while in the year before the merge, it was 22 MU/M, a drop of as much as 62%.

Does this indicate that MEV is "receding"?

I believe not entirely.

It is more like a "weaker party in the game": in the early days, due to chaotic mechanisms, low arbitrage thresholds, and frequent front-running, Searcher profits were unrealistically high.

After the merge, with transparent mechanisms, layered sorting, and intense competition, profits have been diluted across multiple roles in the entire chain, also due to intense competition upstream, with profits migrating to the backend of the chain.

In other words: MEV has not disappeared; it has simply "lengthened" and shifted the "revenue chain." In this process, users have actually gained a portion of fairness.

As it stands, the issue of extreme concentration in the builder market has become more severe.

In the past 30 days, the top five builders constructed over 80% of Ethereum blocks, one root cause being the monopoly of order flows: large builders control high-quality order flow sources, enabling them to consistently provide higher-value block packages through exclusive agreements or preprocessing capabilities.

This is a dangerous centralization feedback loop— the more blocks are built, the more order flows tend to concentrate, leading to even more construction.

Because miners ultimately produce blocks, they are most capable of extracting MEV themselves, while Searchers, after bidding, face execution risks, limitations in expressing Ethereum's native transaction types, and concerns about miners front-running their strategies.

Due to this mutual suspicion and concern, miners and Searchers integrate with each other; miners can increase their share of MEV by reaching customized agreements with Searchers and outsourcing the search for MEV opportunities to them, while Searchers can better control transaction order by establishing relationships with trusted miners.

This will concentrate power in the least ethical participants while reducing network execution and user protection.

Looking back at the evolution of MEV in Ethereum, it is essentially a modernization process from "natural front-running" to "institutional layering."

The earliest MEV was a primitive jungle, relying on seizing, speed, and burning money;

Later, it became a contractual society, establishing order through Flashbots, but also leading to centralization;

After the merge, it entered a layered society where different roles perform their duties, but inequality still exists, just becoming more logical.

This resembles an evolution of social governance:

• From chaotic violence (primitive front-running),
• To black-box agreements (early Flashbots),
• Then to constitutional decentralization (PBS architecture),
• But still with power concentration and elite rule (monopoly of a few Builders and Relays).

One of the prototypes of the next stage is BuilderNet, which brings a reconstruction of the "builder" role itself and represents the most structurally just design of a sorting system.

It is neither absolute idealism nor complete cynicism, but a form of "realistic governance order," resembling "blockchain-style democracy": there is freedom, but it must be fought for;

There is competition, but it is still better than chaos.

Another Structural Solution: Solana's Pre-Auction and High-Speed Chain Scenario

Beyond the "open market + modular role game" model like Ethereum, MEV has another model that does not build a multi-role relay network but instead pre-positions auctions and incorporates them into the client protocol stack, attempting to complete transaction sorting and profit distribution at the chain level.

The most representative of this "centralized path" is Solana and its core MEV infrastructure: Jito.

Let’s look at the miraculous speed of its market share development through a timeline, paying attention to the staking rate and related partners.

• Established at the end of 2021
• Launched on the Solana mainnet in June 2022, with 200 validators by September of the same year, covering 15% of the staking volume
• From 2022 to 2023, financing + iteration + collaboration with the Solana Foundation, the Jito client was included in the official recommendations
• In 2023, TGE, staking Jito gained MEV revenue bonuses, forming a staking and restaking model.
• In Q1 2024, due to strong community opposition, the channel for sending transactions from jito-solana to jito-blockengine was closed.
• In Q2 2024, the number of cooperative validators exceeded 500, covering 70% of Solana's MEV, processing 3 billion transactions in 2024.
• By Q1 2025, the staking coverage rate had reached 94.71%. Today, the importance of cross-chain bridges remains self-evident.

Through the staking volume, it can be said that Jito is the leading infrastructure in the current Solana MEV ecosystem, having established a solid support base among Solana validators over the past three years, ensuring that the vast majority of transactions pass through Jito's system.

Its system diversion has significantly reduced Solana's downtime,

It has provided high-profit returns for front-runners,

And it has allowed Solana validators to gain an additional 30% in MEV revenue, and steadily so.

Moreover, it has transformed from an initial dragon-slaying warrior into a dragon itself, oscillating between the warrior and the dragon, sometimes fierce, sometimes benign.

In today's mainstream meme narratives, it has become a two-faced player that caters to both sides.

In the past year, it has initiated a total bundle of 4.3 billion transactions, generating total tips amounting to 5.51 million SOL. At a market price of 140, this has resulted in an additional revenue of 7.7 billion dollars alongside Jito's infrastructure.

3.1 MEV Auctions and Block Production Process on Solana

The architecture of Jito can be divided into three parts:

• Block Engine: Responsible for the auction logic of MEV bundles, allowing front-runners (Searchers) to participate in auctions by submitting bundle packages (combinations of multiple transactions) and attaching tips to validators to achieve sorting priority.
• Jito-Solana Client: This is Jito's custom validator client, which integrates native support for the Block Engine. Validators running this client can prioritize processing auctioned bundles, bypassing the regular transaction queue.
• Jito Relayer: This is the hub for sending and receiving transactions. In the early days, it was reported to have leaked transaction flows to the Block Engine and even to others through a 200ms delay mechanism, essentially a privatized mempool charging mechanism. Today, it is dynamically officially stated that it no longer engages in transaction selling.

This system is not only used for MEV; objectively speaking, it serves all scenarios that require acceleration and batch transaction bundling.

For example, during the lively opening activities on Solana, market makers utilize the bundling and acceleration mechanisms of bundles to open and deploy chips, among other operations.

Major exchanges can also avoid being attacked by bundling large transaction tips for users. However, it is important to note that these measures cannot prevent scenarios where validators engage in malicious behavior (in reality, you cannot determine which validator is acting maliciously).

This stems from Solana's unique chain mechanism.

Solana is designed for high speed and does not have a public mempool in the traditional sense; transactions are directly broadcast to the next round's block leader through peer nodes. Its structural characteristics are as follows:

The most unique aspect is the absence of a memory pool; however, Solana does not have a memory pool, it merely reduces public transmission rather than completely eliminating (which is impossible) transmission. Due to this characteristic, Searchers on Solana have become exclusive to high-end players.

Additionally, there is a prediction mechanism.

Validators will be randomly sampled from 1,300 validators every epoch (approximately 2-3 days), using the VDF algorithm, which also has a staking-weighted effect.

For example, if the total staked amount of SOL is 2 million, and you stake 200,000 SOL, you will have a 10% probability of being selected each time.

If selected, you will be responsible for producing blocks for the next 4 slots (the equivalent concept of blocks in Solana) for about 1.6 seconds.

This speed is very fast, so any effective node can calculate who the next validator is and attempt to connect with them to submit user transactions. Due to network latency, it is also easy for transactions to miss the current leader and be delivered to the next leader.

Why is MEV harder to govern on Solana?

In fact, these mechanisms seem to favor users by making it difficult for them to be front-run, so why are front-runners more rampant on Solana? The key lies in other reasons.

It is difficult to prove that a leader is acting maliciously; both leaders A and B can access all user transactions, so the cost and ambiguity of malicious actions by leader B are lowered.

Imagine, as the second leader, I see a profitable transaction, so I quickly construct a front-running attack to submit to the block engine for auction. Under the 80% Bundle priority mechanism, my attack will naturally take effect first, even though it is packaged by leader A.

So how can you determine that I, leader B, am the attacker?

Moreover, why are Solana validators prone to betrayal?

Because the profits are incredibly high, and the costs are also very high, forcing validators to continuously expand their sources of income.

Validators incur about 300-350 SOL in voting costs per year (estimated at $42,000 based on a market price of $140) and $4,200 in hardware costs (not including dynamic network costs).

The significant node configuration burden on Solana requires node performance to be at least 24 cores, 256GB of memory, and 2*1.9TB NVME.

Among the commonly used custom Latitude models in the market, currently, 14% of validators are using them, which costs $350 per month.

Ultimately, this has resulted in only 458 out of 1,323 Solana validators being profitable. This is also why the "SIMD-0228 proposal" was voted down.

From the results, this proposal would further reduce block production incentives, inevitably forcing smaller validators to exit, leading to an irreversible centralization of the platform. And when MEV revenue increases while the essential work revenue decreases, what do you think will happen?

3.2 Understanding Jito's Contributions and Shortcomings on Solana

This is my viewpoint: a truly good market will continuously have new competitors entering, while an oligopolistic market will lead to the blockage of challengers. What kind of market does the platform layer hope to become?

Jito's oligopoly is also a side effect of Solana's high-speed structure, which indirectly transfers the opportunities for fair on-chain competition to a few participants who possess data and block production rights.

Jito was once the "dragon slayer" solving Solana's downtime issues, but as its client monopoly and data privileges gradually solidified, Jito has also gradually become a representative of "becoming the dragon."

Through a comparison of core process coins, it can be said that Ethereum is an open, multi-role game MEV market; while Solana resembles a closed-loop order system managed by a dominant platform.

What we see today is no longer just a competition of different MEV tools or tactics, but a fundamental divergence in institutional design:

• Ethereum chose an open game market logic, where each role has participation rights and exit mechanisms;
• Solana chose a centralized model of platform internalization and packaging efficiency, prioritizing system stability and throughput.

Solana has successfully achieved "extreme throughput + extreme sorting control," but at the cost of compressing the on-chain gaming space.

However, I believe that performance issues can ultimately be resolved.

Any entity can perform multi-node global optimization; ultimately, the leader will establish the shortest path in whichever continent they are on, allowing their transactions to reach the leader quickly. They can also establish multicast strategies to divert different user demands, and the future competitive outcome will inevitably be the result of refined operations.

But what cannot be resolved is the issue of market competition squeezing out others.

If jito-solana uses its oligopolistic advantage to modify the Bundle priority strategy from 80% to 90%, or even 95%, then ordinary users will have to infinitely raise Priority Fees to compete for the 5% CU space that is left out.

So why is the market competition in ETH more open? And why is Solana's competition more exclusive?

I believe the root cause is the lack of a bidding role for Builders.

ETH can have multiple Builders producing multiple final block sequences, while validators only verify and choose which one.

However, Solana only has multiple block engines (and they are all from the same entity), and the transaction queue provided to validators is actually a single Bundle (5 transactions), which lacks the competitive element of multiple Builders.

Objectively speaking, from ETH's history, there is no absolute fairness; this competition significantly expands the earnings of validators while reducing the earnings of Searchers. When Searcher earnings decrease, the incidence of attacks will also reduce, ultimately achieving a balance.

4. The Shadow of Freedom

So far, we have seen MEV evolve from Ethereum's "decentralized gaming mechanism" to Solana's "platform-built auction distribution logic." Although the technical solutions differ greatly, they both point to the same core issue:

Who controls the sorting rights?

The control of sorting rights is not a "technical detail" or "optimization item," but the deepest institutional cornerstone of blockchain: the guarantee and breaking of decentralized power structures.

4.1 Vitalik's Reflection: Is MEV a Design Failure or Inevitable?

In Vitalik's 2024 article "Futures yet to be decided," he had a philosophically profound reflection:

"MEV is not a bug, but a natural byproduct of a system's power structure. Blockchains either explicitly grant sorting rights or implicitly create sorting advantages… In the absence of coordination mechanisms, sorting advantages often evolve into opaque operations and rent-seeking structures."

In other words, blockchains can never completely eliminate sorting rights; at most, they can decide: who controls it, whether it can be verified, and whether it can be exited.

Today's MEV designs, whether Ethereum's MEV-Boost architecture, Solana's Jito auction, or other chains' private relay mechanisms, essentially answer three questions:

• Who owns the sorting rights?
• Is the sorting behavior transparent and verifiable?
• Do ordinary users have exit or self-selection mechanisms?

4.2 Is MEV a Poison of Decentralization or a Catalyst for Evolution?

Many people view MEV as a tumor in blockchain systems, but others believe it is a natural result of market mechanism evolution. We can look at it from two dimensions:

Positive Perspective: MEV Incentive Mechanism Enhances Efficiency

• Validators have a stronger motivation to maintain node operation because they can earn additional income through sorting fees;
• Front-runners (Searchers) introduce more complex intelligent arbitrage behaviors, enhancing price efficiency in the DeFi market;
• Sorting auctions provide programmability and funding allocation mechanisms, even giving rise to new scenarios like cross-chain arbitrage and NFT liquidation.

This is a mindset of "acknowledging the existence of sorting rights and opting for market-based handling."

Negative Perspective: MEV is an Invisible Corridor of Centralization

• Sorting rights shift to non-consensus entities like relays, builders, and client operators, leading to governance structures being bypassed;
• The information gap among front-runners widens, turning ordinary users into data "harvesters";
• The chain-level security model gradually relies on closed-source components, violating the original intent of transparency and verifiability.

Jito's "client-integrated sorting logic" solution has both the merit of efficiency improvement and the drawback of bypassing governance and community consensus, forming a de facto sorting oligopoly.

Both aspects have their highs and lows; without positive modules, validators themselves may struggle to survive, and without negative modules, profit-seeking entities (attackers) would lack space for generation. Together, they form an ecosystem that realizes continuous evolutionary competition between life and death.

This also creates a more vibrant system.

Vitalik mentioned:

"Trying to eliminate MEV entirely is futile. The best we can do is to make it transparent, fair, and optionally avoidable."

This means that MEV cannot be "eliminated"; it is a side effect of the system's openness and programmability. What we should truly consider is "allowing it to operate in a more transparent, open, and equally participatory manner."

The game of MEV is both a manifestation of economic freedom and an outlet for power centralization.

Just like the early internet claimed to "decentralize connections to the world," it ultimately gave rise to super platforms like Google and Meta. The commodification of sorting rights will inevitably attract resources, capital, and authority to tilt towards specific nodes.

Without intervention from chain-level governance and structural mechanisms, MEV will accelerate the formation of "protocol oligopolies"—client providers, relay networks, and builder alliances will gradually control the "fate" of user transactions.

This is not about who designs better, but rather the natural consequences of institutional prototypes and incentive paths.

4.3 The Next Stage of MEV

The past decade of MEV has been a history of the evolution of sorting rights. From miners' private mempools to Flashbots' builder networks, and then to Jito's on-chain auction systems, MEV has continuously transformed, differentiating new roles, but it has never disappeared; it has only become more refined, institutionalized, and concealed.

The future sorting market will evolve into an interchain MEV market, where the main chain serves merely as a settlement layer, and sorting and priority selection occur on a decentralized, cross-ecosystem sorting protocol. We are entering a "monetization era" of a freely floating market.

Privacy becomes a prerequisite for new institutional design, no longer just "data protection." "Privacy is not about hiding information, but about balancing power." This has disruptive significance for the MEV world.

In the current architecture, the premature exposure of information (such as public mempool) is the main incentive for MEV. Therefore, many solutions aim to prevent front-running through "encrypted mempool." Suave, Fair ordering, FHE-Rollup, and Time-lock Encryption are all exploring this path.

So I want to revisit the initial question.

Why is MEV not an incidental "arbitrage behavior," but rather a structural phenomenon?

Because sorting rights are an inevitable resource allocation mechanism in all open systems. The existence of sorting is not a mistake, but a fundamental metaphor for power structures. Without addressing the governance of sorting rights, a chain cannot truly be "decentralized"; yet each attempt to govern sorting rights creates new institutional centers and interest structures.

How does it undergo role migration with the evolution of protocol design, consensus mechanisms, and on-chain economic structures?

From miners to proposers, from builders to client operators, the structure of MEV participants is constantly shifting.

MEV is an "endogenous paradox" of the decentralized world.

It forces us into a dilemma: we want to eliminate it because it brings injustice and systemic exploitation; but we cannot truly eliminate it because it is an inevitable byproduct of consensus mechanisms and on-chain markets.

We can only continuously compromise, redesign, and redistribute, making it "seem controllable" under new institutional structures.

By contemplating MEV and facing this mirror, we can see what we call the "ideal of decentralization," and how fragile it is in the face of economic incentives and institutional realities, yet how precious it is.

Because the existence of MEV games allows freedom to find space for survival.

Like a shadow, through it, we can see another dimension, our true selves.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。