opinion, not research.
trustless payments might make ethereum more reliant on trust
disclosure: nice to meet you, alex, co-founder of ultra sound, we run a block relay. trustless payments may end relays. at least in their current form. i am therefore obviously biased. i try to be more truthful still. i’ll leave the judgement to you (:
EIP-7732 buys us something real: pipelining. decoupling beacon and execution blocks gives builders more time, supports larger blocks and higher gas limits, and avoids increased hardware requirements for solo stakers.
bundled into that change is “trustless” builder payment. sounds like a strict improvement over trusting relays. may not be that simple.
today: public relay auctions
most blocks come from a small set of builders (Titan ~50%, BuilderNet ~35%, Quasar ~15%), and relays (ultra sound ~40%, bloXroute ~30%, Titan ~30%). out-of-protocol blocks pay low hundreds of millions per year. protocol rewards are some ~4x - 10x higher still.
relays run open first-price auctions and have invested over many months to enable any anonymous builder, to submit any bid they’re able to pay, without collateral, to any proposer open to an out of protocol bid.
as these auctions are fast and incremental, builders can place thousands of tiny bid updates per slot, quickly converging to roughly a second price, landing the highest value block at the second price. caveat: insofar as top builders are competitive.
fair to note, relays are trusted intermediaries. far from ideal, and frankly against ethereum’s core values. this trust is spread across multiple operators, under strong reputational and economic pressure, and currently open and relatively easy to enter for new builders.
what trustless payments change
trustless payments introduce a new option. a relay publishes an auction and reveals the best public bid. a selfish builder watches that and sends the proposer a slightly higher, trustless bid directly, and privately. the proposer rationally takes the higher bid; the relay and the present-day builder lose.
had that block gone through a relay, dominant builders would likely have outbid it. with public bids plus private, trustless outbidding, those dominant builders are now punished for revealing their valuations.
the rational response is to stop revealing. builders start bidding privately, trustlessly, directly to proposers instead of via relays.
in principle, this defensive behaviour could be limited to “slots where a relay is bypassed.” in practice, this may be hard to predict. so incentive spills over: stop revealing everywhere. the result looks like a first-price sealed-bid auction direct to proposers and the set of builders equipped to compete blind, instead of an open public auction on relays.
some therefore expect: worse price discovery worse, harder entry for new builders, and increased edge in trusted relationships with proposers.
it is fair to note this equilibrium collapse is possible today. eg titan runs a competitive relay and promises to keep it open today, but we should assume tomorrow is hard to predict. rumors exist some relays may attempt to offer builders sealed bid functionality. this warrants a larger discussion but imo the current equilibrium feels defensible. certainly more so without trustless payments.
to offer another fair counter, coordination failures can be overcome. eg builders and relays may be able to coordinate around sealed-bid second-price (Vickrey) auctions with strict policing of bids at the proposer level. note this feels complicated, brittle, and socially hard to enforce.
trustless != low-trust, diverse
trustless payments fix a narrow problem: the trust we place in relays. they do not fix the broader forces that concentrate block building: winner-take-most dynamics, private orderflow, capital intensity, infra moats, and speed that compounds.
so while trustless payments make it possible for any builder to pay any proposer, they don’t make it likely that meaningful new builders will appear or thrive. if dominant builders switch to sealed-bid private strategies, the system may end up more concentrated and more reliant on trust between proposers and a few builders, not less.
mechanical downsides of trustless bids
on top of the game theory, trustless bids have real operational limitations:
can’t be cancelled when sent over p2p.
must be rate-limited when sent over p2p, or they become a DoS vector.
require deposited collateral (iiuc the sum of all promised payments in an epoch).
builders must pay the full sum on missed slots.
to alleviate some of these downsides, builders may encourage proposers to whitelist them to fetch sealed bids directly. an advantage for well connected builders. worse, all downsides above may be alleviated with tight integration and deeper trust still, allowing builders to offset risk and offer more competitive bids.
that’s a worse version of the uncomfy structure we worry about today: a handful of highly optimized builders, in tight loops with proposers, dominating a sealed-bid game that’s difficult for new entrants to crack. meanwhile, relays lose meaningful bid flow as those builders avoid exposing bids publicly and the relay layer withers.
compelling upside
there is a strong upside story for trustless payments: imagine Ethereum under heavy attack, local block building is no longer feasible, existing builders and relays are censored or compromised, and new builders can only appear pseudonymously over the p2p network.
to some extent, we’re already partway there in practice: a lot of valuable flow is private, and many blocks are built downstream of opaque orderflow pipelines rather than by proposers themselves.
in the extreme case, trustless direct bidding from anonymous builders could help Ethereum keep producing blocks even when “official” builders and relays are toast. you might also imagine “pirate relays” that spring up in emergencies to coordinate this.
i do question is how likely that world is, and whether we want to risk re-shaping today’s relatively healthy equilibrium for it.
concluding
pipelining is great, ship it. imo the risk of increased missed slots may well be worth the scaling, and can be pushed back on both in-protocol, and even through trusted relays ha.
trustless payments otoh may come at a high cost. the worry of current out-of-protocol actors is that they may:
weaken the incentive to participate in open, public relay auctions
push dominant builders toward private, sealed-bid relationships with proposers
counter-intuitively risk collapsing a multi-relay, multi-builder landscape into something smaller and more trusted, not broader and more trustless.
i don’t pretend to know the right course here but do personally feel the above perspective has been insufficiently explored. i for one can’t say landing trustless payments makes a trade-off which is understood and supported by a consensus view where differences of opinion have been heard and considered.
may sharing this view make a small contribution in helping ethereum forward 
–
notes i couldn’t help leave out 
:
if staked builders are dominant, they could be punished for exercising the free option inherent in pipelining. perhaps builders / relays should still be staked. requires further consideration imo.
i’m aware of many fair counter arguments which impact the balance above but complicate the discussion and imo don’t change much about the core game theory.
i spoke with a lot of folks to condense the above but also feel takes from core consensus client developers are still sorely lacking. i have many calls scheduled to try and improve this. i imagine the trustless payment upside section especially is still lacking. perhaps some readers may already contribute below.
we should be careful giving out-of-protocol actors veto-like influence.
i was too impatient to wait for review, expects some edits / corrections in the next few hours while i ask people smarter than me what i got wrong.
