Block Finality

The core property of finality. Once a state change is finalized, it cannot be reverted, altered, or forked away, except through an explicit, coordinated protocol upgrade (hard fork). This provides absolute certainty that a transaction is permanently settled. This is distinct from probabilistic finality, where the likelihood of reversion decreases over time but never reaches zero.

The time interval between a transaction's submission and its achievement of finality. This is a critical performance metric.

• Fast Finality: Protocols like Tendermint (Cosmos) or Avalanche achieve this in seconds.
• Delayed Finality: In Nakamoto Consensus (Bitcoin, Ethereum PoW), finality is probabilistic and requires waiting for multiple block confirmations (e.g., 6+ blocks for Bitcoin) for high assurance. Lower latency enables better user experience for exchanges and payment settlements.

A fundamental trade-off in consensus design related to finality.

• Safety: The guarantee that validators will not finalize conflicting blocks. High safety ensures the chain does not fork after finality.
• Liveness: The guarantee that the network can continue to produce new blocks and finalize transactions. Some protocols, under certain failure models (e.g., >1/3 Byzantine validators), must choose to either halt (preserve safety) or continue (preserve liveness), potentially compromising finality. This is formalized in the CAP theorem and FLP impossibility for distributed systems.

Auxiliary protocols that enhance a blockchain's native finality. They run alongside the main consensus to provide stronger guarantees.

• Example: Ethereum's Casper FFG: A finality gadget layered on top of the original Ethereum Proof-of-Work chain. It used a separate validator set to periodically finalize checkpoints, adding an extra layer of economic finality.
• Purpose: They can upgrade a chain with probabilistic finality to one with provable or absolute finality without a full consensus overhaul.

Finality enforced by cryptoeconomic incentives rather than pure cryptographic or algorithmic guarantees. It makes reversion financially prohibitive.

• Mechanism: Validators stake substantial capital as a security bond. If they attempt to finalize conflicting blocks (equivocate), their stake is slashed (destroyed).
• Example: In Proof-of-Stake networks like Ethereum, a finalized block can only be reverted by an attacker destroying at least one-third of the total staked ETH, a cost measured in billions of dollars. This provides practical, if not absolute, irreversibility.

Describes when the finality guarantee is provided.

• Instant Finality: The block is finalized as soon as it is created and accepted by the consensus protocol. Used in BFT-style protocols (e.g., Tendermint, Algorand).
• Eventual Finality: The block is considered provisionally valid upon creation, but finality is achieved only after a period of time or a number of confirmations, as in Nakamoto Consensus. The distinction is crucial for applications requiring immediate settlement certainty versus those that can tolerate short-term probabilistic assurance.

Transactions achieve finality probabilistically as more blocks are mined on top of them, making reorganization exponentially difficult. In Bitcoin, a common heuristic is to wait for 6 confirmations (about 1 hour) for high-value transactions, as the probability of a longer chain reversal becomes negligible. This is not absolute but provides extremely high confidence over time.

These protocols achieve instant, deterministic finality within a single block. Using Byzantine Fault Tolerant (BFT) consensus variants, once a supermajority of validators votes for a block, it is immediately finalized and irreversible. There is no waiting for confirmations, and chain reorganizations are impossible after finalization, providing strong security for real-time applications.

Ethereum's proof-of-stake uses a hybrid model. Casper the Friendly Finality Gadget (FFG) provides economic finality through validator votes. Finality is reached in epochs (every 32 slots, ~6.4 minutes). To reverse a finalized block, an attacker would need to burn at least one-third of the total staked ETH (a "slashing" condition), making attacks economically prohibitive.

These chains offer fast, optimistic finality with different models. Solana uses Proof of History (PoH) for sequencing and achieves finality in about 400ms once a supermajority of validators votes, though it can be reverted under extreme network partitions. Avalanche uses a novel Snowman consensus where nodes repeatedly sample peers, causing correct transactions to achieve finality in 1-3 seconds with high probability.

It's critical to distinguish block time (creation interval) from finality time (irreversibility). A chain may produce blocks every 2 seconds but require minutes to finalize them. Key metrics include:

• Time to Finality (TTF): The average time for a transaction to be irreversible.
• Finality Delay: The gap between block production and its finalization. Low TTF is essential for exchanges and cross-chain bridges to prevent double-spend risks.

Achieving faster finality often involves trade-offs:

• Decentralization: BFT protocols require fast, reliable communication between all validators, which can limit validator set size.
• Liveness vs. Safety: Protocols prioritize one under adversarial conditions. Probabilistic chains favor liveness (the chain always progresses). Instant finality chains favor safety (blocks are never reverted).
• Network Assumptions: Instant finality typically requires stronger synchrony assumptions about message delivery times.

Found in Proof-of-Work chains like Bitcoin, finality is not absolute but grows exponentially more certain with each subsequent block. An attacker with sufficient hash power can still reorganize the chain. Key concepts include:

• Reorg Attack: An attacker mines a longer, alternative chain to reverse transactions.
• Confirmation Depth: The number of blocks needed for a transaction to be considered settled (e.g., 6 blocks for high-value Bitcoin tx).
• 51% Attack: The theoretical attack vector where an entity controls majority hash power to force reorgs.

Achieved via consensus protocols where validators formally agree on a block, making it instantly irreversible. This is a core feature of Proof-of-Stake chains using BFT-style consensus (e.g., Tendermint, Ethereum's LMD-GHOST/Casper FFG).

• Safety over Liveness: These protocols prioritize agreement on a single chain, even if it means halting during network partitions.
• Slashing: Validators acting maliciously (e.g., double-signing) have their staked assets destroyed, economically securing finality.
• Instant Finality: Blocks are finalized in one round, eliminating reorg risks after the fact.

A hybrid model, most notably in Ethereum's PoS, where finality has both a cryptographic and an economic layer.

• Checkpoint Finality: Blocks are 'justified' and then 'finalized' every two epochs (~12.8 minutes) by a 2/3 supermajority of staked ETH.
• Slashing Conditions: Attacks on finality (e.g., surround voting or equivocation) are detectable and punishable by slashing.
• Cost of Attack: To revert a finalized block, an attacker would need to destroy at least 1/3 of the total staked ETH (tens of billions of dollars), making it economically infeasible.

A class of attacks targeting Proof-of-Stake chains where an attacker creates an alternative history starting from a point far in the past.

• Weak Subjectivity: To defend against this, nodes must periodically sync with a trusted 'weak subjectivity checkpoint' to identify the canonical chain.
• Stake Bleeding: A variant where an attacker slowly accumulates old validator keys to build a competing chain.
• Defense: Regular checkpoints and punishing slashing for historical violations mitigate this risk, requiring new nodes to trust recent state.

The failure to achieve finality is itself a security event. In BFT-based systems, if more than 1/3 of validators are offline or malicious, the chain cannot finalize new blocks.

• Liveness Failure: The network halts, preventing new transactions but preserving safety (no conflicting blocks are finalized).
• Network Partition: Can cause separate chain segments to finalize different blocks, requiring social coordination to resolve.
• Recovery: Often requires manual intervention via governance or a hard fork to restart the finality process.

Different finality models present distinct security assumptions and trade-offs.

• Probabilistic (PoW): High resilience to liveness failures, but susceptible to deep reorgs with enough hash power. Energy-intensive.
• Absolute (BFT PoS): Strong, instant cryptographic finality, but vulnerable to liveness halts if >1/3 of validators fail.
• Economic (Hybrid PoS): Balances safety and liveness, with extremely high economic cost to attack finalized blocks, but introduces complexity and weak subjectivity requirements.

A property where the likelihood of a block being reverted decreases exponentially as more blocks are added on top of it, making reversal computationally infeasible. This is the model used by Proof-of-Work chains like Bitcoin.

• Mechanism: Finality is not absolute but approaches certainty over time.
• Example: A Bitcoin transaction with 6 confirmations is considered settled, as the probability of a longer chain reorganization excluding it is astronomically low.

A property where a block is immediately and irreversibly finalized upon validation by a consensus mechanism. Once finalized, it cannot be reverted under any normal network conditions. This is characteristic of Proof-of-Stake chains using BFT-style consensus.

• Mechanism: Achieved through a voting or attestation round among validators.
• Example: In Ethereum's consensus layer, a block is finalized after being included in a chain that receives attestations from at least two-thirds of the total staked ETH.

A concept where reverting a transaction is possible but prohibitively expensive, as it would require the malicious actor to destroy a massive amount of staked capital or computational resources. It bridges probabilistic and absolute models.

• Mechanism: Security is tied to the cost of attack. In Proof-of-Stake, this is the slashing of validator stakes.
• Implication: An attacker could theoretically reorganize a chain, but the financial disincentive makes it irrational.

A property where transaction settlement is immediate and irreversible within a single consensus round, typically in the order of seconds. This is a goal of many high-throughput Byzantine Fault Tolerant (BFT) consensus algorithms.

• Mechanism: Used by networks like Solana (Proof-of-History) and Avalanche (Snowman++).
• Trade-off: Often requires stronger assumptions about network synchrony or a higher degree of centralization in validator sets to achieve speed.

The core protocol that enables network nodes to agree on the state of the ledger. It is the foundational process that produces finality. Different mechanisms define how finality is achieved.

• Proof-of-Work (PoW): Achieves probabilistic finality through computational race.
• Proof-of-Stake (PoS): Achieves absolute/economic finality through validator voting and slashing.
• Practical Byzantine Fault Tolerance (PBFT): Achieves instant finality through multi-round voting among known validators.

An event where the canonical chain switches to a different fork, potentially reversing previously assumed final transactions. Understanding finality is understanding the conditions under which a reorg can or cannot happen.

• In Probabilistic Systems: Reorgs of a few blocks are possible but become less likely with depth.
• In Absolute Finality Systems: Reorgs cannot include finalized blocks unless ≥1/3 of validators act maliciously (a slashing condition).