You're a solo staker in Rotterdam. Thirty-two ETH on the line, a home server humming in the corner, and once every few days the network selects your validator to propose a block. You have maybe twelve seconds. Somewhere in the mempool, a large DEX swap is sitting there like an unlocked door, and you have neither the software nor the team to walk through it. Someone else does.

That someone is a block builder. What they do next is the whole story.

One Job Became Two

For most of Ethereum's proof-of-stake life, the validator who proposed a block also, implicitly, had the right to construct it. Order the transactions however you like, include what you want, leave out what you want. That's a serious amount of power tucked inside a role most people assumed was just rubber-stamping.

Proposer-Builder Separation, PBS in shorthand, splits that power in two. Specialized builders compete to assemble the most profitable block they can. Validators (the proposers) simply pick the highest bid from that auction and sign it. The proposer never sees the transaction order. They see a header and a payment.

The practical version already runs on Ethereum through MEV-Boost, a piece of middleware built by Flashbots. A majority of Ethereum blocks are now produced through MEV-Boost relays rather than assembled locally by validators. Enshrined PBS, baking the separation directly into the protocol rather than relying on off-chain software, remains an active area of research.

What Validators Used to Pocket

Maximal Extractable Value (MEV) is the profit a block producer can extract by choosing which transactions to include, exclude, or reorder. Arbitrage between DEXs, liquidations on lending protocols, sandwich attacks on large swaps: all of it flows from the simple fact that someone decides what goes in a block and in what order.

Before any separation existed, a validator running sophisticated software could capture that value directly. Most validators aren't sophisticated. The Rotterdam solo staker doesn't have a quant team writing transaction-ordering algorithms. A large staking pool operating thousands of validators does.

This created a structural advantage that compounded. The pools with better MEV extraction earned more ETH per block. More ETH meant more stake. More stake meant more blocks. The rich-get-richer dynamic was not subtle.

PBS interrupts that loop, at least partially. Under MEV-Boost, the Rotterdam validator and the large Singapore pool both receive bids from the same open relay market. Both sign the highest bid. Both capture roughly the same MEV revenue per block proposed, because the extraction work is done by builders, not proposers.

A Concrete Walk Through the Auction

Two validators: Ana, a solo staker with 32 ETH and a reliable home setup, and StakeFlow, a staking service operating 10,000 validators. Under the old model, StakeFlow's engineering team runs custom MEV bots, capturing an extra 0.05 ETH per block on average compared to Ana. Across thousands of blocks, that gap compounds into a yield difference that matters.

Under PBS with MEV-Boost:

  1. When Ana's validator is selected to propose a block, her MEV-Boost software queries a set of relays.
  2. Builders (separate, specialized firms) have already assembled candidate blocks stuffed with profitable transaction orderings, arbitrage, and liquidations.
  3. Each builder submits a bid: "I'll pay you 0.08 ETH to propose my block."
  4. Ana's software picks the highest valid bid, blindly signs the block header, and publishes it.
  5. Ana earns 0.08 ETH in MEV revenue she never could have extracted herself.

StakeFlow's validator, selected in the next slot, runs the same process and gets a similar bid. The yield gap, the one driven purely by MEV sophistication, shrinks toward zero.

Now ask yourself: if the proposer's edge just evaporated, where did the power go?

The Power Doesn't Disappear, It Moves

This is what most explanations quietly skip. PBS doesn't eliminate the incentive distortion in Ethereum's block production. It relocates it, the way rerouting a pipe moves the pressure rather than relieving it.

Builders now have enormous power. A builder who sees more of the mempool, maintains better relationships with large transaction sources (order flow), and runs faster infrastructure will consistently out-bid competitors. Over time, block building could concentrate into a small number of dominant firms, each with privileged access to private order flow from wallets and centralized exchanges.

Then there's the relay. In the current MEV-Boost architecture, a relay sits between builders and validators and is trusted to deliver the full block after the validator commits to a header. The relay sees everything. It must be honest about both the block contents and the payment. Flashbots and others operate relays today, and the trust assumption involved is explicit rather than hidden, but it is a trust assumption nonetheless.

Enshrined PBS attempts to remove that relay trust assumption through cryptographic commitments. The builder commits to a block; the proposer commits to proposing it based on the payment; neither can cheat the other without on-chain evidence. The mechanism design is genuinely hard. That's why it hasn't shipped.

What People Get Wrong About This

The common mistake is framing PBS as a fairness fix that's either complete or not worth doing. Both framings are wrong, and I'll say so plainly.

PBS meaningfully improves validator-level fairness. The yield compression between sophisticated and unsophisticated stakers is real and measurable. If solo stakers earn structurally less than large pools, rational capital flows toward pools, the validator set centralizes, and Ethereum's security model quietly degrades. That matters.

But PBS is not a MEV solution. MEV still exists. The value is still extracted from users, particularly the ones getting sandwiched on large swaps. PBS changes who extracts it and how the proceeds are distributed. Treating it as a user-protection mechanism is a category error. It is an incentive-alignment mechanism between Ethereum's own participants.

The right analogy: PBS is like separating the auctioneer from the bidders at an art sale. The auctioneer becomes neutral, takes a fair cut, stops having a thumb on the scale. The paintings still sell for whatever the market will bear, and the richest bidders still win the best lots.

One more thing people misread. The builder market isn't automatically competitive just because it's open. Order flow is the scarce resource. A builder with exclusive access to transaction flow from a major wallet app sees trades before anyone else does. That's a moat, and it doesn't erode just because anyone can technically run a builder. The economics of private order flow are, right now, the sharpest unsolved edge in this design.

PBS gives validators a fairer deal and gives the protocol a cleaner separation of concerns. What it doesn't give anyone, yet, is a world where the power to shape blocks has no center of gravity. That problem just got a different address.