Summary
The Fast Confirmation Rule (FCR) is a proposal that provides high-confidence transaction confirmation within approximately 13 seconds (one slot), reducing exchange deposit wait times by 80–98%. It operates above the existing consensus layer, requiring no protocol changes, by identifying conditions under which a block is effectively irreversible without waiting for epoch finality (~12.8 minutes). The main limitation is a <25% adversarial stake assumption.
Problem
Ethereum’s full finality takes ~12.8 minutes (2 epochs). Exchanges and bridges that want strong security must wait this long before crediting deposits or releases. In practice:
- Most exchanges wait 30–65 blocks (~6–13 minutes).
- Gnosis Bridge: 8 blocks, Circle CCTP: 65 blocks, LayerZero: 15 blocks.
- L1→L2 bridges: deposit times of 10–20 minutes are common.
This creates a terrible user experience for time-sensitive applications.
FCR Mechanism
The FCR fires when two conditions are both satisfied:
1. Dominance Check
The block’s attestation weight must dominate relative to competing blocks at the same height. Specifically:
weight(block) / total_attesting_stake > dominance_threshold
The block must have received attestations from a supermajority of the active validator set, with no competing fork having meaningful attestation weight. This ensures no honest validator could later switch to a competing block without equivocating.
2. Robustness Check
The network must be in a “robust” state — recent slots must not have shown evidence of high reorg risk (missed slots, equivocations, anomalous attestation patterns). This is a liveness-safety tradeoff parameter.
Timing
When both checks pass, the FCR typically fires within one slot (~12 seconds). Under normal network conditions, this is nearly always the case for recent blocks.
Security Assumptions
- Adversarial stake < 25%: if an adversary controls ≥25% of stake, they can potentially cause the FCR to fire on a block that later gets reorged.
- The 2025 paper providing formal analysis shows the FCR provides 99.9%+ confidence under <25% adversarial stake on real Ethereum networks.
- The assumption is weaker than the full finality guarantee (<33%), but much stronger than the probabilistic security of N-block confirmation rules.
Quantitative Impact
| Current approach | Wait time | FCR approach |
|---|---|---|
| Exchange: 30 blocks | ~6 min | ~13 seconds |
| Exchange: 65 blocks | ~13 min | ~13 seconds |
| Full epoch finality | ~12.8 min | ~13 seconds |
| L1→L2 bridge deposit | ~10-20 min | ~13 seconds |
Reduction: 80–98% depending on current confirmation practice.
Adoption Path
FCR is opt-in: exchanges and bridges adopt it client-side without any protocol change. The validator set provides FCR signal through attestation weights already published on-chain; the exchange just applies a different confirmation algorithm.
Proposed adoption sequence:
- Research publication + formal analysis
- Client libraries implement FCR detection
- Exchanges and L2 bridges adopt FCR as primary confirmation
- Protocol may later enshrine FCR natively (e.g., as a new safe head in the Engine API)
Relationship to Finality
FCR is NOT finality:
- Full finality = block cannot be reverted without ≥33% of stake being slashed
- FCR = block is extremely unlikely to be reverted given current attestation patterns, under the <25% adversarial assumption
FCR is a stronger confirmation rule than N-blocks but weaker than finality. Think of it as “probabilistic instant finality” that works for most use cases.
Rational Finality Stalls (Related Risk)
A separate but related finding: validators can economically profit by deliberately delaying finality without reverting blocks. A coalition can stall finality (by not attesting) when the bribery profit exceeds the inactivity leak penalty.
Key parameters:
- Estimated cost of a finality stall attack: ~4.61 ETH in bribery (very low)
- Duration before inactivity leak becomes prohibitive: variable
- Bridges acting pre-finality (i.e., using FCR) are not affected by finality stalls, since FCR does not depend on full finality.
- Bridges that wait for full epoch finality ARE vulnerable to finality stall-based griefing.
Implication: FCR adoption actually reduces bridge exposure to finality stall attacks, because finality stalls cannot prevent individual block confirmations from accumulating sufficient attestation weight.
Current Confirmation Practices (Reference)
| Protocol | Blocks waited | Approximate time |
|---|---|---|
| Gnosis Bridge | 8 | ~1.6 min |
| LayerZero | 15 | ~3 min |
| Exchanges (typical) | 30–65 | ~6–13 min |
| Circle CCTP | 65 | ~13 min |
| Full finality | ~64 (2 epochs) | ~12.8 min |
Open Questions
- Should FCR be standardized in the Engine API (new
safehead type, alongside the existingsafe= last finalized)? - How does FCR interact with sub-slot execution (TOOL/NuConstruct) proposals that change the attestation timing model?
- Can FCR thresholds be formally verified under the new Goldfish/Majorum consensus designs planned for future upgrades?
Related Pages
- Finality in Ethereum: Gasper, Gloas, and the Engine API — Full finality in Ethereum; Gloas nuances
- Decoupled Consensus: Goldfish, Majorum, and Dynamic Availability — Goldfish/Majorum; future consensus model
- Bridge Finality Risks and Pre-Finality Actions — Detailed analysis of bridge confirmation practices and finality stall exposure
- Sub-Slot Execution (TOOL / NuConstruct) — Sub-slots as an alternative to FCR for fast feedback
Key Sources
- The Fast Confirmation Rule (2026) — FCR mechanism, formal analysis, adoption path
- Rational Finality Stalls (2026) — economic analysis of finality stall attacks; bridge vulnerability data