Inside the Dark Forest: The Maximal Extractable Value Playbook
Chapter 1: What Is MEV?
💡 Definition
Maximal Extractable Value (MEV) refers to the maximum value that can be extracted from transaction inclusion, ordering, and exclusion within a block. MEV is an emergent behavior of permissionless blockchains that support complex, expressive smart contracts — particularly in DeFi environments. It’s the surplus profit available by manipulating block composition beyond normal transaction fees and rewards.
🔍 Key Insight
MEV is not a bug — it’s a structural consequence of transparent mempools, programmable state transitions, and unordered transaction flows. It is value that exists because actors are incentivized to optimize for their own profit.
🔁 How MEV Emerges
Imagine a DEX with large, pending trades in the mempool. If you can see those trades before they’re finalized, you can place your own buy or sell orders ahead (front-running), or in between them (sandwiching), to extract profit. That’s MEV.
💥 Examples of MEV in Action
- DEX Arbitrage: Buying from one DEX and instantly selling on another with a better price.
- Sandwich Attacks: Inserting a buy before and a sell after a large user trade.
- Liquidation Sniping: Racing to liquidate an undercollateralized loan before other actors.
- Backrunning: Executing a trade immediately after another to benefit from price movement.
📊 MEV vs Normal Validator Rewards
Validators (or miners pre-Merge) can extract MEV directly, but most of it is now found and submitted by independent actors known as searchers.
🧪 Why It Matters
MEV is central to:
- The economics of validators
- The behavior of searchers
- Gas price volatility
- Fairness and ethics in DeFi
Understanding MEV is the key to understanding how blockchains really work under pressure — and how users, builders, and protocols can protect themselves.
🧠 Core Takeaways
- MEV is profit from transaction ordering — above normal gas or rewards
- It affects everything from DEX trades to lending markets
- It’s a constant battle between searchers and other searchers, not just users and validators
⛏️ Miner Extractable Value in Proof-of-Work
In Ethereum’s original Proof-of-Work (PoW) system, miners controlled transaction ordering. Since they were responsible for proposing blocks, they had full discretion over the order of transactions. MEV in this context was known as Miner Extractable Value, and it was limited primarily by the miner’s technical sophistication and their access to mempool data.
Miners could:
- Reorder transactions
- Include their own transactions first
- Exclude competing transactions
This power created incentives for miners to extract value by participating in arbitrage, liquidations, or front-running — sometimes directly, sometimes via partnerships with off-chain bots.
🔁 The Merge: Ethereum’s Transition to Proof-of-Stake
In September 2022, Ethereum transitioned from PoW to Proof-of-Stake (PoS), in an upgrade known as The Merge. This replaced miners with validators, who now take turns proposing blocks based on their staked ETH.
The core idea remains: the block proposer has the ability to choose and order transactions. However, in practice, PoS introduced a new ecosystem of actors and a more structured MEV pipeline.
🧱 Validators and the Modern MEV Supply Chain
Under PoS, validators are selected to propose blocks based on the amount of ETH they’ve staked. But not all validators have the infrastructure or knowledge to identify profitable MEV opportunities. This gave rise to new roles:
- Searchers: Independent actors who scan the mempool and construct profitable bundles
- Builders: Entities that collect bundles from searchers and assemble full blocks
- Relays: Middlemen who connect builders with validators, often used to enforce censorship resistance or protocol rules
This system is often referred to as Proposer-Builder Separation (PBS), and it formalizes the MEV flow in PoS:
Searchers → Builders → Relays → Validators🧠 Implications of the Transition
- MEV extraction is now more specialized: Searchers compete to find the most profitable bundles.
- Validators outsource MEV: They earn rewards from builders without doing MEV search themselves.
- Transparency and fairness tradeoffs: Private relays (e.g. Flashbots) help protect users, but also create centralization concerns.
🪙 What Stayed the Same?
- Block proposers (miners or validators) still control transaction ordering
- MEV still exists because of DeFi composability and open mempools
- The battle for profit still favors those with the fastest, most efficient infrastructure
🧠 Core Takeaways
- MEV began as Miner Extractable Value in PoW
- Ethereum’s move to PoS shifted the MEV landscape but didn’t remove the incentives
- Proposer-Builder Separation (PBS) defines today’s MEV supply chain
Chapter 2: From Miners to Validators — The Evolution of MEV
Ethereum didn’t always operate the way it does today. In its early days, the network ran under a Proof-of-Work consensus mechanism. Miners were the heart of the protocol, consuming energy to solve cryptographic puzzles in order to produce the next block. But alongside block production came power — the power to control what transactions were included and in what order. That power birthed the original concept of MEV: Miner Extractable Value.
In the Proof-of-Work era, miners had total discretion. They could choose to prioritize transactions with higher fees, reorder trades to favor themselves, or even censor competing activity. Some ran sophisticated bots to detect arbitrage opportunities between DEXs, inserting their own transactions to capture profit. Others partnered with off-chain searchers who would pay them for guaranteed block inclusion. The mempool was a hunting ground, and the miner was both gatekeeper and participant.
But everything changed in 2022 with Ethereum’s historic upgrade: The Merge. The network transitioned from Proof-of-Work to Proof-of-Stake. Instead of miners, validators would now take turns producing blocks, selected based on the amount of ETH they had staked. The environmental cost dropped overnight. The architecture of MEV, however, only grew more complex.
Validators inherited the same block production powers as miners, but most lacked the infrastructure to identify and extract MEV. This gave rise to a new supply chain of MEV professionals. Independent actors known as searchers began scanning the mempool with real-time bots. When they detected a profitable opportunity, they would package multiple transactions into what’s called a bundle. These bundles would then be submitted to specialized block builders — entities whose sole job is to construct the most valuable blocks possible.
Because validators only care about maximizing rewards, they often outsource the block-building process. Relays emerged to connect builders with validators. These relays serve as trusted intermediaries, forwarding the most profitable blocks to whichever validator is proposing the next one.
The end result was a modular, layered MEV pipeline: searchers at the bottom, builders in the middle, validators at the top. This structure is now known as Proposer-Builder Separation, or PBS. It was never part of Ethereum’s original plan — it emerged as a market solution to the inefficiencies of PoS MEV extraction.
The PBS model has benefits. It allows specialization, reduces validator complexity, and creates open markets for block production. But it also creates new risks: centralization among builders, opaque relay behavior, and increasing barriers to entry for searchers without the fastest infrastructure.
Still, the fundamental dynamic hasn’t changed. Whoever controls block content controls MEV. In PoW it was the miner. In PoS, it’s the validator — aided by builders, fueled by searchers. The power to reorder is the power to extract. And that power still defines the economics of Ethereum today.
Chapter 3: Who Controls Transaction Ordering in PoS?
At the heart of Ethereum’s block production lies a fundamental power: the ability to decide what goes into a block, and in what order. This seemingly simple responsibility — selecting and sequencing transactions — drives one of the most competitive and lucrative layers of the blockchain economy. Whoever controls ordering controls opportunity. And in the post-Merge world, that role belongs not to miners, but to validators.
Every 12 seconds, a validator is selected at random from a pool of stakers to propose the next block. That validator, in theory, can fill the block however they wish. They can select transactions from the public mempool, prioritize their own, or work with third parties who offer prebuilt blocks in exchange for a share of the profits. This process isn’t just mechanical — it’s strategic. The contents and ordering of a block can mean thousands, sometimes millions, of dollars in arbitrage, liquidation fees, or slippage exploitation.
But the validator rarely acts alone. Over time, Ethereum has evolved a complex hierarchy around block building. Validators began outsourcing this task to more sophisticated entities — block builders — who specialized in constructing high-value blocks using bundled transactions submitted by searchers. These searchers run bots that scan the mempool for opportunities, then assemble their bundles into profitable sequences. The builder selects the best combinations and packages them into a block. That block is then handed off to the validator, often via a relay like Flashbots, which acts as the gatekeeper between builders and proposers.
This architecture — known as proposer-builder separation (PBS) — wasn’t part of Ethereum’s original design. It emerged organically, then was embraced by the community to help decentralize MEV extraction. Rather than letting validators race to become the best searchers, PBS allows specialization. Validators validate. Builders build. Searchers search.
The result is a high-frequency, highly competitive supply chain where power is distributed — but not equally. Builders and relays now hold enormous influence. If a relay censors a block, it disappears. If a builder prefers one searcher over another, it shapes the future of that searcher’s revenue. If a validator chooses only certain relays, it creates a market of inclusion preferences.
In theory, validators still have the final say. In practice, the economics push them to choose the most profitable block available. That usually means choosing the best-paying builder, who is in turn fed by the fastest, smartest searchers. Ordering becomes not a function of protocol rules, but of off-chain coordination and latency races.
Transaction ordering in proof-of-stake Ethereum isn’t deterministic. It’s a marketplace. And that marketplace is governed by hardware, bandwidth, incentive design, and a constantly shifting field of actors vying to be first.
To understand who controls ordering in Ethereum, you have to understand who builds blocks — and who builds the incentives behind them.
Chapter 4: Ethereum Core Concepts for MEV
Ethereum isn’t just a blockchain — it’s a programmable financial engine. And to understand how MEV arises from this engine, we need to understand what it’s made of.
Every transaction on Ethereum modifies global state. That could mean moving tokens, interacting with a smart contract, minting NFTs, or initiating a flash loan. But under the hood, all of these are variations of the same structure: calldata, gas, execution, and state transition.
When you send a transaction, you’re not sending tokens. You’re submitting instructions — calldata — that tell the Ethereum Virtual Machine what to do. Those instructions have a cost, paid in gas. And the amount of gas they consume depends on how complex the operation is: reading storage, writing new data, looping over arrays, etc.
Gas is what prevents spam and abuse, but it’s also what introduces economic constraints. MEV actors must compete not just on logic but on cost-efficiency. When a searcher identifies an arbitrage opportunity, it only matters if they can fit it into a transaction that is profitable after gas is paid.
That’s where the art of MEV begins: manipulating calldata, compressing logic, using low-gas opcodes, and avoiding expensive writes. It’s not enough to be right — you have to be fast, small, and cheap.
But MEV doesn’t come from Ethereum’s engine alone. It comes from composability. From the fact that all DeFi protocols live on the same ledger, share the same execution model, and speak the same language. A token swap on Uniswap can be instantly paired with a flash loan on Aave, a vault withdrawal on Yearn, and a liquidation on Compound — all in one atomic transaction. No other system in the world allows that kind of economic stitching.
This composability is what creates MEV. Not just the existence of value, but the ability to extract it. The permissionless, combinatorial nature of Ethereum is both its strength and its vulnerability.
To understand MEV is to understand Ethereum itself: its transaction model, its gas market, its block structure, and its incredible capacity for modular execution. The game is played at the opcode level — but the incentives are shaped by protocol design.
That’s why every new contract matters. Every vault, pool, bridge, or peg adds a new surface for MEV. Adds a new route, a new edge, a new profit path. This is the terrain that searchers map every block, every second, across every shard and rollup.
You don’t just need to know how Ethereum works. You need to know how to weaponize it.
Chapter 5: Who Actually Extracts MEV?
It’s tempting to think that validators are the ones extracting MEV. After all, they’re the ones who propose the blocks. But in reality, the vast majority of MEV is discovered, structured, and delivered by a different class of actor altogether: the searchers.
Searchers are the hidden brains of the MEV ecosystem. They run specialized infrastructure that monitors Ethereum’s mempool in real time, analyzing the flow of transactions as they emerge and identifying opportunities where reordering, insertion, or bundling can create profit. These opportunities aren’t hypothetical. They’re extremely concrete: a pool imbalance, a liquidation threshold, a mispriced asset between two venues. The searcher’s job is to spot it, quantify it, and act on it — before anyone else does.
To do that, they rely on speed, precision, and code. A typical searcher bot will observe thousands of transactions per second. It will simulate the state of the chain if each one executes and then simulate how inserting another transaction might affect the result. It’s a constant, high-stakes loop: watch, simulate, act. The faster you are, the more likely your bundle is to land first. The slower you are, the more likely you are to be front-run.
Get 𝕮. 𝕰. 𝕳𝖎𝖗𝖘𝖈𝖍𝖆𝖚𝖊𝖗’s stories in your inbox
Join Medium for free to get updates from this writer.
But even finding a profitable bundle isn’t enough. Searchers can’t include their transactions directly into blocks. Instead, they submit them to builders — along with a bid that reflects how much they’re willing to share with the validator who ultimately proposes the block. This is where Flashbots comes in. Flashbots is a private relay that lets searchers send bundles to builders without exposing them to the public mempool. It protects against copycats, preserves privacy, and reduces failed transactions. But it also centralizes order flow, and introduces dependencies on off-chain infrastructure.
Because the MEV game is competitive, searchers often end up giving away a large portion of their profit just to win the block slot. In volatile conditions, a searcher might bid away 90% or more of the value in their bundle to beat another bot by a millisecond. This creates a ruthless economic environment where the margins are razor thin and the cost of delay is total failure.
Yet searchers persist — because when it works, it works. A well-crafted arbitrage bundle can net thousands of dollars in a single block. A well-timed liquidation can recover a trove of collateral. A properly constructed sandwich attack can siphon slippage from an unsuspecting trader. And when these profits are aggregated across hundreds of blocks per day, they form the backbone of MEV extraction across the Ethereum network.
So, who actually extracts MEV? Not just validators. Not just protocols. The real extractors are the searchers: anonymous, high-frequency, latency-optimized operators running bots in the dark forest of Ethereum. They are the snipers, the surgeons, the ones who turn protocol design into financial outcome.
They are the reason MEV isn’t just theoretical. They are the reason it’s real.
Chapter 6: MEV Techniques and Attacks
Every block on Ethereum is a battlefield. The weapons are bundles. The ammunition is gas. And the combatants are searchers, racing to claim slices of value before someone else does. But not all MEV is extracted in the same way. Over time, a handful of core techniques have emerged — each exploiting the openness of the mempool, the atomicity of transactions, or the structure of DeFi protocols.
The first and most notorious of these is the sandwich attack. Imagine a user submitting a large trade on a decentralized exchange. A searcher sees the pending transaction in the mempool, calculates the expected price impact, and reacts instantly. They submit two transactions: one that front-runs the user by buying the asset before them, and another that sells the same asset back immediately after the user’s trade executes. The user pays a worse price, while the searcher profits from the price movement they helped create. It’s called a sandwich because the victim’s trade is squeezed between two slices of predation.
Then there’s arbitrage. This is the cleanest form of MEV — the one most aligned with market efficiency. If a token trades at different prices on Uniswap and Curve, a searcher can buy low on one and sell high on the other. In doing so, they help normalize the price across platforms. But even here, speed is everything. By the time an arbitrage opportunity appears, multiple bots are already simulating paths to capture it. The winner is often the one with the best latency, the most optimized calldata, or the best tie-breaking bribe to the builder.
Liquidations are another rich vein. In lending markets like Aave or Compound, borrowers must stay above a collateral ratio. When they fall below it, anyone can liquidate them — repaying their debt and seizing collateral at a discount. Searchers monitor these ratios constantly, looking for accounts on the brink. The moment a borrower tips over, bots race to execute the liquidation and capture the spread. Here, the MEV isn’t from price impact but from protocol design: a structural bounty for whoever moves first.
And then there’s backrunning. Subtler than the rest, backrunning involves identifying a profitable state change caused by another transaction and following up immediately to capitalize on it. This might mean trading after a large swap that pushes the price, or rebalancing a pool after a rebasing token adjusts its supply. Unlike front-running, which can harm users, backrunning often operates in the wake of value creation rather than in advance of it.
What unites these techniques isn’t just their speed or complexity — it’s their dependence on Ethereum’s architecture. MEV doesn’t exist in a vacuum. It arises because of composable protocols, transparent mempools, and the deterministic nature of execution. Searchers don’t guess. They simulate. They test every path, every routing, every gas configuration, then fire the one that’s most likely to land.
To a casual observer, these transactions are invisible — just hashes on a block. But to those in the game, they are tactical maneuvers, coded and launched with precision. They are not just transactions. They are exploits. They are edge.
The more open Ethereum becomes, the more fertile the ground for MEV. And every new mechanism — whether AMM, oracle, vault, or bridge — offers both functionality and fragility. It is in that fragility that searchers find their advantage.
Chapter 7: Gas Auctions and Bidding Wars
Every block is an auction. Not just for the right to be included — but for the right to be first. This is the world of gas auctions: a relentless, high-speed economy where searchers bid for execution priority, and where the smallest edge in price or latency can separate profit from failure.
To understand gas auctions, you have to understand one thing: Ethereum processes transactions in order of their effective gas price. The higher the price, the earlier the inclusion — if there’s space. This creates a brutally competitive arena in which searchers who identify the same MEV opportunity must outbid one another to ensure their bundle lands before anyone else’s. It’s not just a race of algorithms. It’s a race of budgets.
These bidding wars are especially intense during times of market volatility. A mispriced token or a liquidation cascade can trigger dozens of searchers to simultaneously submit competing bundles. Each bundle contains a slightly higher gas price than the last, a slightly different ordering, a slightly different bribe to the validator. The one that wins is the one that pays the most — up to the edge of profitability.
This is why MEV extraction often isn’t as lucrative as it appears. While some searchers may identify profitable opportunities, the majority of the margin is eaten by competition. It’s common for the winner to surrender 80 to 90 percent of the gross MEV value in gas costs or validator tips. The validator profits. The network burns gas. And the searcher walks away with a sliver of what looked like a gold mine.
To make matters worse, failed bundles are expensive too. If two bundles compete and neither wins, the losing transactions may revert, wasting gas entirely. This is why many searchers route their bundles through private channels like Flashbots, where they can specify conditions for inclusion and avoid front-running.
But even within Flashbots, the gas auction still exists — just hidden. Searchers embed bribes in the form of direct payments to validators, calculated to beat competitors without exposing themselves to public mempool risks. These private bribes are just as aggressive, just as thin-margin, and just as punishing.
The gas auction is a fundamental feature of MEV. It ensures that only the most efficient, aggressive, and well-capitalized searchers survive. It rewards those who can simulate faster, compress calldata better, and make tighter bids. It penalizes inefficiency. It filters the field.
In this game, you don’t just win by seeing the opportunity. You win by outbidding everyone who saw it too. You win by paying more — and keeping less.
Chapter 8: Gas Golfing — Optimization Mastery
There is an art to winning MEV. It’s not enough to find the opportunity. You have to deliver the transaction. You have to make it fit. You have to make it fast. And most of all, you have to make it cheap. This is gas golfing: the practice of squeezing every possible byte, every cycle, every opcode down to its smallest, fastest, leanest form.
In MEV, efficiency isn’t cosmetic — it’s existential. The cost of a transaction isn’t just a rounding error. It’s the difference between landing first and landing never. When multiple searchers are competing to exploit the same arbitrage or execute the same liquidation, the one with the most gas-efficient bundle can bid more aggressively and still turn a profit. That margin — sometimes just a few thousand gas — decides who wins the block.
Gas golfing begins at the most fundamental level: calldata. Every byte of calldata has a price. Passing a long string of token addresses? That’s gas. Using human-readable selectors instead of raw function hashes? That’s gas. A skilled searcher compresses calldata down to its minimal expression — shortened paths, tightly packed arrays, precomputed values. They remove redundancy the way a racer strips bolts from a car.
But it doesn’t stop at calldata. The next frontier is contract logic. A well-designed MEV bot avoids storage whenever possible. Writing to state costs gas. Reading from storage costs gas. The most elite strategies execute entirely on stateless logic — calculations done on the fly, results discarded once the transaction completes. They avoid emitting events, avoid fallback handlers, avoid anything that isn’t strictly necessary.
Gas golfing is also about opcode selection. The EVM rewards simplicity. Complex math can often be replaced by bitwise tricks or lookup tables. Repeated calls can be batched. External calls can be consolidated. Even deployment itself is fair game: some searchers deploy fresh contracts mid-bundle that self-destruct after use, removing the need for persistent addresses.
This level of optimization creates a strange arms race. The most successful MEV bots aren’t just smart — they’re small. They are precision weapons, coded in assembly, debugged at the bytecode level, deployed with surgical intent. They exist only to execute a single maneuver, perfectly, at the lowest possible cost.
Gas golfing isn’t a style. It’s a survival skill. In the mempool, there is no second place. Either your transaction gets in first, or it doesn’t get in at all. And if your transaction costs too much gas, you can’t afford to bid high. If you can’t bid high, you don’t get included. If you don’t get included, you don’t get paid.
In MEV, gas is gravity. It pulls every strategy downward — toward simplicity, toward elegance, toward brutal necessity. The best searchers know that profit doesn’t just come from opportunity. It comes from execution. And execution begins with cost.
Chapter 9: Real-World Case Studies
Theory is clean. Execution is messy. And nowhere is that mess more instructive than in the real-world battlefields of MEV, where millions are won and lost one block at a time. These stories don’t just show what’s possible — they reveal the edge cases, the failures, the adaptations, and the invisible rules of a game most people never see.
Consider a sandwich attack that targeted a large stablecoin trade on Uniswap. The user submitted a transaction to swap several million dollars worth of USDC for ETH. Before it could execute, a searcher caught the trade in the mempool, calculated the slippage, and crafted a precise bundle: a buy before the trade, a sell after, both compressed into a few hundred bytes of calldata. The bundle was submitted privately to a Flashbots relay. It landed. The searcher netted over $100,000. The user got a worse price. The validator collected a tip. And on-chain, the entire maneuver looked like three unrelated trades — just data.
But not all stories end in profit. In another case, a high-profile liquidation opportunity triggered a bidding war between searchers. A large position on a lending platform had fallen below collateral threshold. The race began: dozens of bots simulated the liquidation, bundled their bids, and began outbidding each other via bribes and gas prices. At the last second, one searcher submitted a malformed calldata payload that reverted the entire bundle. Others followed suit. The block closed with no liquidation. The opportunity evaporated. Hours later, the position was manually closed — and the protocol took a loss.
Then there are the true monsters: the bundles that span protocols, that use flash loans across three or four chains, that initiate cross-DEX arbitrage, trigger a vault withdrawal, and sweep profit back through a bridge — all within one atomic transaction. These mega-bundles aren’t just trades. They’re entire strategies encoded in a single breath. One such bundle in 2023 earned a searcher over $1.4 million in profit. It combined a Balancer-curve arbitrage with a collateral flash mint, a rebasing token exploit, and a synthetic oracle manipulation. It was live for less than 20 seconds before being executed and then buried forever in a finalized block.
These aren’t edge cases. They are the standard for elite searchers. They demonstrate how far the field has come — and how unforgiving it remains. Mistakes are expensive. Gas is blood. Timing is law. And only those who understand the mechanics, the mempool, the network latency, the calldata budget, and the validator preferences can consistently extract value.
In MEV, there are no second chances. These case studies are a window into that pressure. Into the black-box reasoning of bots competing at nanosecond timescales. Into the silent, invisible warfare that defines the bleeding edge of DeFi.
These are not just stories. They are blueprints. They are warnings. They are proof that the game is real — and that it’s being played with sharper tools every day.
Chapter 10: Toxic vs. Non-Toxic MEV
Not all MEV is created equal. Some forms of value extraction push markets toward efficiency, closing price gaps and increasing arbitrage parity. Others leech value from unsuspecting users, exploiting slippage, latency, and information asymmetry to steal from those least prepared to defend themselves. This is the ethical spectrum of MEV — from neutral to malignant, from optimization to outright exploitation.
Toxic MEV is value taken at someone else’s expense. The clearest example is a sandwich attack: a user submits a trade, and a searcher injects their own buy and sell orders around it, manipulating the price so the user receives worse execution. The searcher profits directly from the user’s loss. No value is added to the system. Only value transferred, forcibly and invisibly, from one wallet to another.
There’s a deeper problem with toxic MEV: it degrades trust. If users believe they’ll be front-run or sandwiched on every trade, they stop participating. If DeFi becomes a casino where only insiders win, the entire vision of permissionless finance collapses. Protocols lose users. Liquidity dries up. The system cannibalizes itself.
On the other side is non-toxic MEV. These are forms of value extraction that improve efficiency without harming counterparties. A searcher who performs cross-DEX arbitrage to equalize prices is contributing to market health. A liquidation bot that promptly clears risky debt is safeguarding lender solvency. These actions align with the protocol’s intended function. They provide services under competitive pressure, and in doing so, they justify their own margins.
But the line is not always clear. Is a fast-reacting oracle arbitrage toxic? What about capturing MEV from a poorly designed rebase function? In many cases, the toxicity depends on context: how informed the counterparty is, how predatory the execution feels, how unavoidable the loss becomes for the average user.
The MEV community has spent years grappling with these questions. Flashbots, for instance, has proposed frameworks for classifying MEV, distinguishing between extractive behaviors that benefit protocols and those that exploit protocol weaknesses. Some suggest a sort of MEV taxonomy — much like malware classification — where we label attacks, simulate damage, and assign moral weight.
Others argue that ethics are irrelevant in a permissionless system. That if code is law, then outcomes are neutral. But this view ignores social context. DeFi is built by humans, used by humans, and governed by communities. Ethics aren’t optional — they’re structural. If we don’t encode fairness into our systems, we’ll build tools that empower predators.
The most promising solutions involve shifting design. Batch auctions, encrypted mempools, and intent-based architectures aim to reduce the surface area for toxic MEV. They level the playing field without eliminating competition. They allow searchers to earn — but not by harming others.
Toxic MEV isn’t just a technical challenge. It’s a moral one. And the way we answer it will define whether DeFi becomes a tool for liberation — or just another engine of extraction.
Chapter 11: Protocol-Level Mitigations
The war for MEV doesn’t just unfold in the mempool — it’s increasingly being fought at the design level. For years, developers treated MEV like a side effect, a shadow price of doing business on a composable blockchain. But today, the conversation has changed. Protocols are no longer ignoring MEV. They’re engineering against it.
One of the most effective tools in this space is the batch auction. Unlike the standard mempool, where transactions are processed in the order of their gas price, a batch auction collects all orders over a fixed time window and executes them simultaneously. This removes the benefit of front-running because no one can guarantee they’ll be first. Projects like CowSwap and Seaport have deployed variations of this model to neutralize sandwich attacks and reduce toxic order flow.
Another powerful concept is encrypted or private mempools. In these systems, users can submit their transactions in a way that keeps the content hidden until after block inclusion is finalized. Flashbots’ SUAVE initiative, for example, proposes a decentralized network of execution environments that protects transaction privacy and routing. The goal is to prevent searchers from accessing sensitive trade data before it executes — essentially cutting off their visibility into user intent.
Then there’s the idea of intent-based architecture. Instead of submitting a fully formed transaction, users declare what they want to happen — like converting one token into another or repaying a loan — and let a network of solvers compete to fulfill that intent in the most efficient, least extractive way. This flips the execution model on its head. Users no longer author specific calldata — they define outcomes. And in doing so, they limit how much information can be used against them.
Even at the consensus level, changes are emerging. Proposer-Builder Separation (PBS), which formalized the relationship between validators and builders, is now being considered for inclusion directly into Ethereum’s protocol. By building PBS into the base layer, Ethereum can ensure a fairer, more transparent builder marketplace and reduce the power of centralized relays.
These aren’t fringe experiments. They’re real shifts in architecture, adopted by protocols that understand the existential threat of unchecked MEV. They acknowledge a hard truth: if the game rewards predators, users will flee. And if users flee, the value layer collapses.
Mitigating MEV doesn’t mean eliminating profit. It means reshaping incentives so that value is created, not extracted. It means leveling the field so searchers earn by helping, not harming. And it means building systems that remember what blockchains were meant to be — open, fair, and free.
The future of MEV isn’t purely technical. It’s architectural. And the winners won’t just be those who extract value. They’ll be those who design systems where value flows in ways everyone can live with.
Chapter 12: MEV on L2s and Rollups
As Ethereum has scaled beyond its mainnet, the battleground for MEV has expanded with it. Layer 2s and rollups promise lower fees and faster transactions, but they also come with their own flavor of MEV — different in form, but no less intense in consequence. If Layer 1 is the grand arena, Layer 2 is the side alley where the same fights play out under a different set of constraints.
Each rollup has its own execution environment, its own sequencer, and its own block production flow. That means MEV is not just about global Ethereum anymore. It’s about local context — who controls ordering, how finality works, and whether the mempool is public or private. And in most cases, the answer is the same: the sequencer is king.
Sequencers on rollups act as temporary block producers. They receive transactions, order them, execute them, and eventually submit the resulting state root to Ethereum for settlement. But this power comes with risk. A centralized sequencer can reorder transactions at will. It can front-run, censor, or bundle its own MEV strategies before any user transaction sees the light of day. This isn’t theoretical. It’s structural.
On Arbitrum, Optimism, zkSync, and others, sequencers today are mostly operated by the core teams. They promise not to exploit their power — but there is little transparency or accountability. As usage grows, so do the incentives. If mainnet MEV drives searchers to build private relays and bribe validators, what happens when a single party controls the ordering logic for an entire L2?
Still, some rollups are experimenting with alternatives. Shared sequencing, decentralized builders, and proof-based ordering models are being explored. The goal is to bring some of the fairness guarantees of PBS to the Layer 2 world. But none of these are mature. None are widespread. And in the meantime, MEV on rollups remains a dark pool.
There’s another wrinkle: cross-domain MEV. A searcher might spot an arbitrage between a token on Ethereum mainnet and its wrapped version on a rollup. Or a liquidation opportunity triggered by a state update that hasn’t yet been finalized across bridges. These opportunities are rare, but massive. And they require bots that understand latency not just within a chain, but across them.
MEV on Layer 2 isn’t a separate domain. It’s an extension of the same game. The players are the same. The tools are similar. But the rules have shifted. The geography has changed. And the future of MEV will be shaped as much by what happens off mainnet as on it.
Ethereum’s vision is modular. But modularity doesn’t eliminate MEV. It fragments it. It reassigns power. It changes the topology of extraction. And if we’re not careful, it will centralize what mainnet spent years trying to decentralize.
The next wave of MEV isn’t just a fork of the old system. It’s a multi-domain, multi-layer, multi-chain battlefield. And every rollup is its own terrain.
Chapter 13: Governance and Power
Beneath the flash loans, calldata, and arbitrage graphs lies a deeper question: who decides? Who controls the infrastructure, the rules, and the flows that determine who gets to extract MEV and who gets excluded from the game? In a system driven by permissionless access and open execution, power doesn’t disappear. It shifts.
In Ethereum’s MEV economy, the most visible players are the searchers, the validators, and the builders. But behind them, often less discussed, are the relays, the protocol designers, and the infrastructure operators. These actors don’t just respond to incentives. They shape them. They set the terms of inclusion, the norms of competition, and the boundaries of acceptable behavior.
Relays like Flashbots decide which builders get through. Builders decide which bundles make it into blocks. Validators select from builder submissions, but their choices are constrained by marketplace dynamics. A validator who picks the lowest-paying builder earns less. A relay that censors political transactions can starve activists of block space. And a searcher with exclusive access to a builder can dominate an entire class of MEV.
Governance isn’t just about votes and DAOs. It’s about the architecture of coordination. It’s about the invisible assumptions that define who has the right to participate and who merely watches from the outside. The current MEV stack is largely off-chain. That means it’s governed not by on-chain mechanisms, but by reputation, access, latency, and alignment.
And yet these power structures matter. If a single relay becomes dominant, it can censor bundles, bias inclusion, or sell access. If a builder cartel forms, it can suppress competition and extract rent from searchers. If validators collude with builders off-protocol, they can reroute value behind closed doors.
This isn’t dystopian speculation. It’s the natural outcome of high-value coordination in a world with no central enforcement. And the only counterbalance is transparency, decentralization, and the active engineering of competitive, accountable alternatives.
The future of MEV governance is uncertain. Some advocate for protocol-enforced PBS to constrain validator discretion. Others explore decentralized relays, permissionless builder markets, and open-source simulation tooling. Still others argue for redistributive systems that redirect a portion of MEV back to users or the protocol treasury.
What’s clear is that MEV is no longer just a technical exploit. It’s a structural force. And the question of who controls it — who defines the rails, who shapes the market, who owns the edge — will determine the kind of financial system Ethereum becomes.
This is the politics of MEV. Not a matter of ideology, but of infrastructure. Not a question of fairness alone, but of power. And power, like value, always finds a way to concentrate unless we actively resist it.
Chapter 14: Toward Ethical MEV
The existence of MEV is inevitable. But the way we design for it — the way we shape who gets to participate, who gets to profit, and who gets protected — is still up for grabs. The question isn’t whether MEV will exist. It’s what kind of MEV ecosystem we want to build. And that starts with ethics.
In a permissionless system, it’s tempting to think that ethics don’t matter. That the network is neutral, that profit is the only valid signal, and that anything that fits within the rules is fair play. But Ethereum was never just code. It was coordination. It was a social layer wrapped around an execution engine. And that social layer defines what we tolerate.
Ethical MEV is about aligning incentives with outcomes. It’s about designing systems where searchers profit from maintaining market efficiency, not preying on user ignorance. It’s about ensuring that validators earn for contributing security, not for auctioning off the right to extract value from retail traders. It’s about creating an environment where the value produced by blockspace is not just harvested, but redistributed.
That might mean redistributing a portion of MEV back to users — either through rebates, treasury capture, or protocol-integrated profit sharing. It might mean strengthening the separation of powers between builders and proposers, ensuring that no single actor can dominate block construction. It might mean defaulting to private mempools, where users can safely transact without fear of being sandwiched. It might mean educating users to recognize transaction risks, or building wallets that automatically route through protective relays.
There is no one-size-fits-all solution. But the goal is clear: a system where MEV extraction is not synonymous with exploitation. A system where participation doesn’t require predation. A system where the arms race is one of efficiency and ingenuity — not opacity and exclusion.
Ethical MEV is also about defaults. About building tooling that nudges participants toward better outcomes, even when they’re unaware of the risks. About surfacing the costs of transaction reordering, making them legible, and giving users the ability to opt out. About designing protocols with MEV in mind, rather than retrofitting them after exploitation.
Ethereum doesn’t need to eliminate MEV to be ethical. It needs to contain it, shape it, guide it. To channel its energy into directions that serve the network rather than undermine it. Because MEV isn’t going away. But it can be made sustainable. It can be made fair. It can be made something we’re proud to defend, not just clever enough to exploit.
Ethical MEV isn’t about idealism. It’s about resilience. It’s about designing a system that can scale, include, and endure.
