The Physics of MEV: The Blockspace Auction
Why your transaction is being watched. A deep analysis of the Dark Forest, Proposer-Builder Separation (PBS), and the industrialization of arbitrage.
🎯 What You'll Learn
- Deconstruct the 'Dark Forest' mempool model
- Trace the MEV Supply Chain (User → Searcher → Builder → Validator)
- Calculate the math of a Sandwich Attack
- Analyze Proposer-Builder Separation (PBS)
- Evaluate Censorship Resistance in a Flashbots world
Introduction
Blockchains promise transparency. But for a few milliseconds every 12 seconds, the most important market in the world is completely opaque.
This is the market for Transaction Ordering.
If I know you are about to buy $10M of ETH, I can buy it right before you and sell it to you at a higher price. This is risk-free profit. This value, extracted by reordering transactions, is called Maximal Extractable Value (MEV). It is the hidden tax that funds the entire validator industry.
The Dark Forest: The Public Mempool
The moment you broadcast a transaction, you enter the Dark Forest (a concept from Cixin Liu, adapted by Dan Robinson). Thousands of “Searcher” bots simulate your transaction in real-time.
- If you left money on the table (Arb), they take it.
- If you created slippage (Sandwich), they eat it.
- If you made a mistake (Liquidation), they kill you.
The Physics of the Sandwich
A Sandwich Attack is Atomic Volatility Creation.
- Victim Tx: “Buy 1000 ETH with 1% slippage.”
- Attacker Tx 1 (Frontrun): Buy ETH, pushing price up by 0.99%.
- Victim executes: Pushes price up another 0.01%.
- Attacker Tx 2 (Backrun): Sell ETH immediately.
Math: The victim pays for the attacker’s profit plus the attacker’s gas fees.
The Industrial Revolution: PBS (Proposer-Builder Separation)
In the old days (2020), miners ran the bots. It was chaotic. Today, the supply chain is specialized.
1. The Searcher (The Sniper)
- Role: Writes highly optimized Rust/C++ code to find specific MEV opportunities.
- Input: Public Mempool.
- Output: A “Bundle” (Ordered list of txs).
2. The Builder (The Factory)
- Role: Aggregates bundles from thousands of Searchers to build a full block.
- Input: Bundles.
- Output: A Block.
- Physics: They run massive simulations to find the combination of bundles that pays the highest total ETH.
3. The Proposer (The Validator)
- Role: The random ETH staker selected to propose the block.
- Input: Bids from Builders (“I will pay you 2.1 ETH to propose this block”).
- Output: Signing the block.
- Physics: They do nothing but auction off their “slot rights” to the highest bidder via MEV-Boost.
Code: Calculating an Arb
Here is the logic a simple arbitrage bot runs 10,000 times per second.
def check_arbitrage(token_in, amount_in, reserves_a, reserves_b):
# Calculate output from DEX A (Uniswap V2)
amount_out_a = get_amount_out(amount_in, reserves_a[0], reserves_a[1])
# Calculate output from DEX B (Sushiswap)
amount_out_b = get_amount_out(amount_out_a, reserves_b[1], reserves_b[0]) # Swap back
profit = amount_out_b - amount_in
gas_cost = ESTIMATED_GAS * GAS_PRICE
if profit > gas_cost:
print(f"ARB FOUND! Profit: {profit} wei")
execute_bundle()
The race is not about math. It is about Latency. Checkers that run closer to the Builder’s server (often in the same AWS datacenter) win.
Deep Dive: The Censorship Risk
Because Builders construct 90% of blocks, they are central points of failure. If a major Builder decides to censor transactions from “Address X” (e.g., Tornado Cash), those transactions might be delayed.
The Defense:
- Censorship Resistance Lists (crLists): Proposals to force builders to include valid transactions.
- Encrypted Mempools: Builders can’t censor what they can’t see (until it’s too late).
Practice Exercises
Exercise 1: The Bribe (Beginner)
Scenario: You find a 1000 should you bid to the Validator to ensure your bundle gets included? (Hint: It’s an auction. What is the next highest bidder bidding?)
Exercise 2: Private Transactions (Intermediate)
Scenario: You want to swap $1M USDC. Task: Why should you send this via “Flashbots Protect” RPC instead of Infura/Alchemy? What is the trade-off in latency?
Exercise 3: Builder Centralization (Advanced)
Scenario: 3 Builders produce 80% of Ethereum blocks. Task: Analyze the impact on the network if these 3 builders collude to ignore a specific DeFi protocol.
Knowledge Check
- What is a Bundle?
- Why can’t you just broadcast a sandwich attack to the public mempool?
- Who pays the MEV reward ultimately?
- What is the role of the Relay in PBS?
- Is MEV solvable?
Answers
- An ordered group of transactions. “My Tx + Victim Tx + My Tx”. It executes atomically (all or nothing).
- You will be sandwiched. A generalized frontrunner bot will see your attack and attack you. You must use a private relay.
- The User. Through worse execution price (slippage) or higher gas fees (congestion).
- Trust. The Relay ensures the Builder pays the Validator and the Validator proposes the Builder’s block without stealing the MEV.
- No. MEV is fundamental to permissionless systems. You can only minimize it (fair ordering) or democratize it (distribute profits).
Summary
- MEV is inevitable. Wherever there is ordering, there is value.
- PBS industrializes it. It separates the “sophisticated work” (Building) from the “decentralized work” (Validating).
- The user pays. Use private RPCs to protect yourself.
Want to go deeper?
Weekly infrastructure insights for engineers who build trading systems.
Free forever. Unsubscribe anytime.
You're in. Check your inbox.
Questions about this lesson? Working on related infrastructure?
Let's discuss