Ethereum (ETH) Validators Would Cast Two Votes Under Researchers' 4-Minute Finality Draft

Two Votes Instead of One

Validators securing Ethereum (ETH) would eventually have to cast two separate votes instead of one under a draft improvement proposal that Ethereum researchers filed on Wednesday, October 21. The document targets finality, the point at which a transaction becomes effectively impossible to reverse, and aims to cut the average wait from roughly 16 minutes to about 4 minutes. The mechanism it proposes to change sits at the core of the network's proof of stake design, the Casper protocol that replaced Ethereum's mining-based consensus, under which validators post staking collateral to participate. Today a single vote from each validator does double duty: it helps select the next block and, at the same time, counts toward finalizing blocks from the past. The draft splits those duties into two distinct vote types, which allows finality votes to run on their own dedicated cadence rather than being chained to block selection. The proposal does not alter transaction execution itself; it changes how the network's votes are organized. The motivation is economic, and the research behind the proposal spells it out. Operators who move customer funds onto a Layer 2 chain typically advance the money first and collect repayment only once the underlying transaction reaches finality; in practice they front many concurrent transfers at once, so the exposure compounds by the hour. The Ethereum (ETH) Foundation's research treats the funds held back while awaiting repayment as an operator cost. An operator that prepays $1 million per hour carries roughly $267,000 in unreimbursed funds under the current 16-minute delay, according to the analysis. Shortening finality to 4 minutes would cut that figure to about $67,000, a drop of roughly three quarters. The draft itself carries no direct read for the Ethereum (ETH) price, but the cost arithmetic underneath it is unusually concrete for a research-stage document.

Quantum-Resistant Groundwork and Open Questions

Two further goals ride along in the same draft, and both bear on the network's long-term security posture. The first is groundwork for a future move to quantum-resistant signatures. Once block selection and finalization are voted on separately, each vote type can migrate to a new signature scheme on its own schedule, which the researchers present as a structural benefit of the split. The draft does not choose a signature scheme, and the researchers are explicit that the selection sits outside the proposal's scope. The second goal is resistance to a specific withholding tactic: an attacker who conceals votes during a round and reveals them later can sway which block gets selected, and separating the votes is meant to narrow the window that trick depends on. The document is candid about what it leaves unfinished. The voting cadence for the separate finality votes is not fixed. The penalty rules that would apply to validators who miss their new duties are not written. Parts of the safety analysis, the argument that a split design preserves the guarantees the chain offers today, remain incomplete. Until those pieces are settled, the proposal stays a research artifact rather than a schedulable change, and no activation hard fork is named. Any activation date, testnet rehearsal or client milestone would be set only after that analysis is complete. Node operators face no immediate action. Nothing in the draft alters client behavior today, and operators do not need to prepare a configuration change or a testnet migration on any timeline the text provides. The obligation sits with the researchers to finish the specification. Ethereum's upgrade history suggests the distance from draft to mainnet is measured in years, and changes to finality touch every client, every validator and every application that assumes a settled transaction stays settled.

What Still Has to Be Built

In our reading, the cost model is what makes this draft more than an academic exercise. A three-quarter cut in capital tied up per bridging operator, from about $267,000 to roughly $67,000 at a $1 million-per-hour prepayment rate, turns finality latency into a payments-margin question rather than only a consensus-engineering one. If the design survives review, the first beneficiaries are the payment and bridging firms doing the prepaying, and their economics give the researchers a concrete reason to finish the safety proofs. The next step, completing the voting cadence, the penalty design and the safety analysis, belongs to the proposal's authors; validators pick up obligations only after that work is done.