MEV for Non-Quants — Sandwiches, PBS, and How Users Can Reduce Losses

MEV for Non-Quants: Sandwiches, PBS, and Practical Ways to Reduce Losses

You don’t need a PhD in market microstructure to keep more of your swaps. This guide explains what MEV actually is, how it shows up as worse prices and failed transactions, what Proposer-Builder Separation (PBS) changed after Ethereum’s Merge, and most importantly the simple wallet settings, RPCs, and app choices that help ordinary users avoid getting sandwiched or sniped.

We include plain-English theory, concrete examples, and verified links to specs, docs, and tools: Flashbots, MEV-Boost, CoW Protocol, 1inch Fusion/0x RFQ, UniswapX intents, private mempool RPCs like MEV Blocker and Flashbots Protect, and credible explainers from Ethereum.org.

TL;DR:
  • MEV = value extracted from your transaction by reordering/including/excluding it in a block. Common retail pain: sandwich attacks on AMM swaps and liquidation sniping around lending protocols. Primer: Ethereum.org — MEV.
  • PBS changed the supply chain: builders craft blocks, proposers pick the best bids via relays (MEV-Boost). This professionalized extraction but also enabled user protection routes via private orderflow. Docs: Flashbots — MEV-Boost.
  • Quick wins for users: use private or protected RPCs for swaps (e.g., MEV Blocker RPC, Flashbots Protect RPC); prefer CoW Swap batch auctions, 1inch Fusion RFQ, or UniswapX intents; avoid huge slippage; set reasonable deadlines; and double-check approvals/permit usage.
  • Advanced knobs: “deadline” or “TTL” minutes, per-trade slippage, use RFQ/intents for large trades, and split size across time. On L2s, similar concepts apply but mempool access and builder markets differ by chain.

1) What is MEV? (Plain English)

MEV stands for Maximal (or Miner/Maximal) Extractable Value. It’s the profit someone can make by choosing how your transaction sits in the block before or after other transactions or by deciding to exclude it entirely. Think of a block as a short line at a ticket counter. The person arranging the line can decide to put themselves (or their friends) in front of you, or split your party to get slightly better seats for themselves. On Ethereum, after the Merge, the “person arranging the line” is usually a block builder working with searchers and relays, while the proposer (validator) chooses the best block they’re offered.

From a user’s wallet, MEV shows up as:

  • Worse swap price than what your screen showed (sandwich attack or routing against you).
  • Failed or reverted transactions that still cost gas (arbitrage you “created” disappeared, or you lost a race).
  • Being liquidated right at the edge when prices dip or an oracle updates.

Key background: Ethereum.org MEV overview • Flashbots research & blogs.

2) Where does MEV come from? (AMMs, liquidations, NFTs, oracles)

MEV isn’t “one trick.” It’s a family of opportunities around public mempools, deterministic AMM math, lending liquidations, oracle update timing, NFT mints, and cross-domain bridges. The common ingredient is information asymmetry: somebody can see or anticipate your order and take a step that extracts value before you can react.

  • AMM sandwiches: A searcher buys before your swap (pushing price up), lets your order execute at a worse price, then immediately sells after you, pocketing the slippage you paid. Classic on Uniswap-style pools. Background: Uniswap Docs, “Economics of MEV” papers.
  • Arbitrage & backruns: Prices differ across pools/venues; your trade creates a profitable spread that a bot closes immediately behind you.
  • Lending liquidations: Protocols like Aave/Compound reward liquidators when a borrower’s health factor falls below 1; bots compete to be first. See: Aave liquidations FAQ, Compound Docs.
  • Oracle/keeper timing: Updates by Chainlink or TWAPs create predictable moments when prices “jump.” Docs: Chainlink Data Feeds.
  • NFT primary drops: First-come mints cause gas wars; snipers reorder to grab the supply. See: OpenSea Docs, Blur resources.

3) What Proposer-Builder Separation (PBS) changed (post-Merge)

Before the Merge, miners both built and proposed blocks. After the Merge, Ethereum moved to proof-of-stake and opened the door to PBS: specialized builders assemble the most profitable blocks; relays deliver them; and the randomly chosen proposer (validator) picks the highest-paying block without seeing its contents. In practice, PBS today is implemented via MEV-Boost, client software that lets validators receive block bids from multiple builders. Intro & guide: Flashbots — MEV-Boost.

Why users should care:

  • Professionalization: Sophisticated builders and searchers capture MEV more efficiently.
  • New protection paths: The same plumbing that routes lucrative bundles can also carry private orderflow that shields you from public mempool bullying.
  • Relay diversity & policy risk: Different relays/builders have different policies; healthy diversity reduces censorship and reliability risks. Relay lists: mevboost.org.

4) Retail pain points in the wild: how the attacks feel

4.1 Sandwich attacks on swaps

Symptom: you click “swap,” see a preview price, and the execution is worse—even though the chart barely moved. Behind the scenes, your on-chain transaction sat in the public mempool long enough for a bot to simulate: if they buy before you and sell after you, can they harvest your slippage? If yes, they submit a bundle to a builder that positions their transactions around yours. You pay the gas, they eat your lunch.

4.2 Liquidation sniping and grief

If your loan is near liquidation, MEV bots monitor oracles and mempools to liquidate the instant your health factor dips. They’ll even manipulate orderflow around oracle updates to make the window cleaner. The user experience: your collateral is sold faster than you can supply more margin.

4.3 NFT gas wars

On hyped mints, bots bid sky-high priority fees and strategically bundle to maximize fills, leaving retail stuck with failed transactions and big gas burn. Project teams increasingly use allowlists, VRF randomness (e.g., Chainlink VRF), or batching to soften this, always read the mint’s rules first.

5) Mitigation that actually works: RPCs, apps, and settings

You can’t “turn off” MEV across the network, but you can route your orderflow and shape your transactions to remove most retail-level leakage. Here are the highest-leverage options:

A. Use MEV-aware orderflow by default (one-tap protection)
  • CoW Swap (cow.fi): batch auctions & solvers execute off-mempool and settle atomic matches; designed to avoid sandwiches by design. Docs: docs.cow.fi.
  • 1inch Fusion (fusion): RFQ/auction model where Resolvers fill intents; reduces mempool exposure vs direct AMM posts.
  • UniswapX (overview): off-chain intents filled by fillers; settlement chooses best price/route while minimizing MEV exposure.
B. Send swaps through a private/“protected” RPC (no public mempool)
  • MEV Blocker RPC — community anti-MEV relay stack that routes orderflow privately and shares backrun revenue with users. Site: mevblocker.io.
  • Flashbots Protect RPC — sends your tx as a private bundle to participating builders; not visible in public mempools. Docs: Flashbots Protect.
  • bloxroute Protect — private transaction endpoints via BDN. Docs: docs.bloxroute.com.
  • Taichi Private RPC (Amber Group) — private gateway historically used by arbitrageurs; check availability for your region. Info: taichi.network.

Trade-off: private RPCs may have different inclusion guarantees than public mempool. Read each provider’s docs for failure modes and “refund” policies (if any).

C. Shape the transaction you send
  • Reduce slippage (e.g., 0.3–1.0% on majors; only go higher if the pair is illiquid). Slippage is literally the “budget” a sandwich harvests.
  • Set short deadlines (“transaction expires in” or “TTL”): 60–180 seconds is often plenty for liquid pairs.
  • Split big trades across time or venues; use RFQ/intents for size. For AMM-heavy pairs, check Matcha (0x) or 1inch to compare quotes.
  • Use tokens’ permit flows/Permit2 where available (reduces persistent unlimited approvals). Docs: Permit2.

6) Step-by-step recipes (MetaMask & friends)

6.1 Route trades via CoW Swap (zero config)

  1. Open cow.fi and connect wallet. Review fee model and supported networks in docs.
  2. Enter token pair/size. Click settings and set slippage (e.g., 0.5%).
  3. Submit intent. Your order goes to solvers; settlement is batched to avoid public mempool sandwiching.

6.2 Use a private RPC (MEV Blocker or Flashbots Protect)

MetaMask → Settings → Networks → Add network manually

Provider RPC URL (example) Docs
MEV Blocker See “Get RPC” button at mevblocker.io (they provide the exact endpoint string). Website
Flashbots Protect Per docs: Quick start Docs
bloxroute Protect Create endpoint via account dashboard; see docs. Docs

Tip: Keep your public RPC as backup. Some private relays won’t include obviously toxic orderflow; failed inclusion can still happen. Re-submit via a protected DEX if needed.

6.3 1inch Fusion / UniswapX intents (good for bigger sizes)

  1. Open 1inch → enable Fusion. Docs: Fusion explainer.
  2. Or use Uniswap with UniswapX enabled. Docs: UniswapX overview.
  3. Set slippage but expect better effective execution than a raw AMM post during volatile minutes.

6.4 Sensible defaults for everyday swaps

  • Slippage: 0.3–0.8% for liquid majors; 1–2% for mid-liquidity; illiquid pairs only if you can monitor.
  • Deadline/TTL: 1–3 minutes for majors; don’t leave unbounded.
  • Approvals: prefer per-trade or Permit2 approvals; remove unlimited approvals periodically (e.g., via revoke.cash).
  • When in doubt: route through CoW/1inch/UniswapX first, not a random front-end that posts to public mempool by default.

7) DeFi habits that quietly save you money

Liquidity & timing
  • Trade when pool depth is healthy and spreads are tight; avoid very thin hours around major news.
  • Consider TWAP-style splitting for large sizes; many aggregators let you simulate impact.
  • Check pool analytics on Dune or token pair pages on your DEX.
Security & approvals
  • Use Permit2 or per-trade approvals when available.
  • Regularly revoke stale allowances: revoke.cash.
  • Prefer front-ends with audits and public docs (CoW/1inch/Uniswap core sites).
Liquidations (borrowing users):
  • Keep buffer health factor (e.g., target 1.5+ on volatile assets). See Aave liquidation FAQ.
  • Set alerts and consider auto-repay bots that act before liquidators do.
  • Understand your oracle sources. If it’s Chainlink, review update cadence: Chainlink feeds.

8) MEV on L2s (Arbitrum, Optimism, Base, others)

Layer-2s share the same fundamentals who sees your order before it finalizes, and who can reorder it, but the plumbing can differ:

  • Mempool visibility: Some L2s don’t expose a public mempool in the same way, but sequencer-level reordering and “private orderflow” equivalents can still exist. Always check the L2’s docs.
  • Sequencer role: The L2 sequencer often acts like “single builder.” Research their MEV policy and whether private orderflow endpoints exist (e.g., “protected” APIs from aggregators).
  • Same mitigations apply: Use RFQ/intents (CoW/1inch/UniswapX where supported), keep slippage realistic, and prefer reputable front-ends.

Examples & readings: Arbitrum Docs • Optimism Docs • Base Docs.

9) Builders’ corner: shipping MEV-aware products

If you run a wallet, DEX, or aggregator UI, give users protection by default:

  • Offer private relay routing (MEV Blocker / Flashbots Protect) as a toggle-on default, with clear messaging.
  • Prefer intents/RFQ for large trades; only post to AMM when necessary.
  • Simulate sandwichability in the UI (pre-trade analysis) and warn when the slippage window is too wide.
  • Use permit flows to minimize standing approvals; surface revoke links after completing a swap.
  • Measure effective price vs quote, not just gas. Show “you saved from MEV” receipts where appropriate.

Core references for implementers: MEV-Boost • Protect RPC • CoW Docs • 1inch Fusion • UniswapX • Ethereum.org MEV.

10) FAQ

Does EIP-1559 stop MEV?

No. EIP-1559 changed how users bid for inclusion (base fee burn + priority tips). It didn’t remove the underlying incentives to reorder within a block. That’s why PBS/MEV-Boost and private orderflow exist.

Will a higher gas tip protect me?

Sometimes you’ll just move faster into the trap. Paying more can help you beat other users, but it doesn’t stop a builder from arranging bundles around your trade. Routing via protected relays or intents is usually better for price protection than simply paying more.

Are private RPCs safe?

They reduce exposure to the public mempool, which cuts sandwich risk. But they have operational trade-offs (different inclusion guarantees, possible regional outages). Always keep a fallback RPC and understand the provider’s policy. Docs: Flashbots Protect, MEV Blocker, bloxroute.

Do L2s eliminate MEV?

No. They often centralize the “builder” role into the sequencer, but many of the same incentives exist. Some L2s are exploring fair ordering/auction schemes. Read each L2’s docs (e.g., Arbitrum, Optimism, Base).

What about “MEV burn” or smoothing?

You’ll see proposals/research about redistributing builder profits back to users/validators (e.g., “MEV burn,” smoothing). Treat these as long-term protocol discussions; they don’t replace the immediate protections you can use today. Use the references in the research hubs (Flashbots, Ethereum R&D) to follow updates.

References & further reading (official docs & explainers)

Note: Product policies, RPC endpoints, and relays evolve. Always check the linked docs before relying on a specific protection route, and keep your wallet/software up to date.