Risk Analysis

Rollups are L2s that periodically post state commitments to Ethereum. These commitments are validated by either Validity Proofs or are accepted optimistically and can be challenged via Fraud Proof mechanism within a certain fraud proof window. Additionally L2 data is also posted to Ethereum, hence there are no additional trust assumptions introduced.

Columns

#	name	State Validation	Data Availability	Sequencer Failure	Proposer Failure

1	Base Chain Base Chain	Fraud proofs allow actors watching the chain to prove that the state is incorrect. Single round proofs (1R) prove the validity of a state proposal, only requiring a single transaction to resolve. A fault proof eliminates a state proposal by proving that any intermediate state transition in the proposal results in a different state root. For either, a ZK proof is used.	All of the data needed for proof construction is published on Ethereum L1.	In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 12h delay on this operation.	Anyone can be a Proposer and propose new roots to the L1 bridge.
2	Arbitrum One Arbitrum One	Fraud proofs allow actors watching the chain to prove that the state is incorrect. Interactive proofs (INT) require multiple transactions over time to resolve.	All of the data needed for proof construction is published on Ethereum L1.	In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 1d delay on this operation.	Anyone can be a Proposer and propose new roots to the L1 bridge.
3	OP Mainnet OP Mainnet	Fraud proofs allow actors watching the chain to prove that the state is incorrect. Interactive proofs (INT) require multiple transactions over time to resolve.	All of the data needed for proof construction is published on Ethereum L1.	In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 12h delay on this operation.	Anyone can be a Proposer and propose new roots to the L1 bridge.
4	Mantle Mantle	STARKs and SNARKs are zero knowledge proofs that ensure state correctness. STARKs proofs are wrapped in SNARKs proofs for efficiency. SNARKs require a trusted setup.	All of the data needed for proof construction is published on Ethereum L1.	In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 12h delay on this operation.	Only the whitelisted proposers can publish state roots on L1, so in the event of failure the withdrawals are frozen.
5	Lighter Lighter	SNARKs are succinct zero knowledge proofs that ensure state correctness, but require trusted setup.	All of the data (SD = state diffs) needed for proof construction is published onchain.	Users can force the sequencer to include a transaction by submitting a request through L1. If the sequencer censors or is down for for more than 14d, users can use the exit hatch to withdraw their funds.	Users are able to trustlessly exit by submitting a zero knowledge proof of funds.
6	Starknet Starknet	STARKs are zero knowledge proofs that ensure state correctness.	All of the data (SD = state diffs) needed for proof construction is published onchain.	Users can submit transactions to an L1 map, but can't force them. When users “complain” that their transaction is stuck on L1 and not picked up by the sequencer, the Security Council minority can bypass the sequencer by posting a state root that includes it.	Only the whitelisted proposer can update state roots on L1, so in the event of failure the withdrawals are frozen. The Security Council minority can be alerted to enforce censorship resistance because they are a permissioned Operator.
7	Ink Ink	Fraud proofs allow actors watching the chain to prove that the state is incorrect. Interactive proofs (INT) require multiple transactions over time to resolve.	All of the data needed for proof construction is published on Ethereum L1.	In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 12h delay on this operation.	Anyone can be a Proposer and propose new roots to the L1 bridge.
8	Linea Linea	SNARKs are succinct zero knowledge proofs that ensure state correctness, but require trusted setup.	All of the data needed for proof construction is published on Ethereum L1. Unlike most ZK rollups, transaction data is posted instead of state diffs.	There is no mechanism to have transactions be included if the sequencer is down or censoring. Eventually (after 6 months of no finalized blocks) the Operator role becomes public, theoretically allowing anyone to post data.	Only the whitelisted proposers can publish state roots on L1, so in the event of failure the withdrawals are frozen. Eventually (after 6 months of no finalized blocks) the Operator role becomes public, theoretically allowing anyone to propose state with valid proofs.
9	ZKsync Era ZKsync Era	STARKs and SNARKs are zero knowledge proofs that ensure state correctness. STARKs proofs are wrapped in SNARKs proofs for efficiency. SNARKs require a trusted setup.	All of the data (SD = state diffs) needed for proof construction is published onchain.	Users can submit transactions to an L1 queue, but can't force them. The sequencers cannot selectively skip transactions but can stop processing the queue entirely. In other words, if the sequencers censor or are down, they are so for everyone.	Only the whitelisted proposers can publish state roots on L1, so in the event of failure the withdrawals are frozen. There is a decentralized Governance system that can attempt changing Proposers with an upgrade.
10	Morph Morph	Fraud proofs allow actors watching the chain to prove that the state is incorrect. Single round proofs (1R) only require a single transaction to resolve. ZK proofs are used to prove the correctness of the state transition. The system currently operates with at least 5 whitelisted challengers external to the team.	All of the data needed for proof construction is published on Ethereum L1.	Users can force the sequencer to include a transaction by submitting a request through L1. If the sequencer censors or is down for 7d, new L1 batches must include at least 1 transaction from the queue.	Anyone can become a Proposer after 7d of inactivity from the currently whitelisted Proposers. This requires using the source-available prover to submit a zk proof of validity for the proposal.
11	Unichain Unichain	Fraud proofs allow actors watching the chain to prove that the state is incorrect. Interactive proofs (INT) require multiple transactions over time to resolve.	All of the data needed for proof construction is published on Ethereum L1.	In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 12h delay on this operation.	Anyone can be a Proposer and propose new roots to the L1 bridge.
12	BOB BOB	Fraud proofs allow actors watching the chain to prove that the state is incorrect. Single round proofs (1R) prove the validity of a state proposal, only requiring a single transaction to resolve. A fault proof eliminates a state proposal by proving that any intermediate state transition in the proposal results in a different state root. For either, a ZK proof is used.	All of the data needed for proof construction is published on Ethereum L1.	In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 12h delay on this operation.	The primary whitelisted proposer has an optimistic advantage, letting them win by default if no conflicting proposals are made. This privilege is dropped after 1mo of inactivity, and anyone can leverage the source available zk prover to prove a fault or a conflicting valid proposal to win against the privileged proposer and/or supply a bond and make a counter proposal at any time.
13	Katana Katana	STARKs and SNARKs are zero knowledge proofs that ensure state correctness. STARKs proofs are wrapped in SNARKs proofs for efficiency. SNARKs require a trusted setup.	All of the data needed for proof construction is published on Ethereum L1.	The self-sequencing delay is configured offchain and the node source and config are unverified. In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 12h delay on this operation.	Only the whitelisted proposers can publish state roots on L1, so in the event of failure the withdrawals are frozen.
14	Abstract Abstract	STARKs and SNARKs are zero knowledge proofs that ensure state correctness. STARKs proofs are wrapped in SNARKs proofs for efficiency. SNARKs require a trusted setup.	All of the data (SD = state diffs) needed for proof construction is published onchain.	Users can submit transactions to an L1 queue, but can't force them. The sequencers cannot selectively skip transactions but can stop processing the queue entirely. In other words, if the sequencers censor or are down, they are so for everyone.	Only the whitelisted proposers can publish state roots on L1, so in the event of failure the withdrawals are frozen. There is a decentralized Governance system that can attempt changing Proposers with an upgrade.
15	Scroll Scroll	STARKs and SNARKs are zero knowledge proofs that ensure state correctness. STARKs proofs are wrapped in SNARKs proofs for efficiency. SNARKs require a trusted setup.	All of the data needed for proof construction is published on Ethereum L1.	In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 7d delay on this operation. Proposing new blocks requires creating ZK proofs.	If the Proposer fails, users can leverage the source available prover to submit proofs to the L1 bridge.
16	Taiko Alethia Taiko Alethia	Every proposal range is verified by exactly two proofs chosen from SGX (Geth), SGX (Reth), SP1 and RISC0, with at least one SP1 or RISC0 proof required. Proof submission is gated by ProverWhitelist, which has 2 whitelisted provers. This can affect liveness but does not allow finalizing invalid state.	All of the data needed for proof construction is published on Ethereum L1.	Users can submit transactions to an L1 queue, but can't force them. The sequencers cannot selectively skip transactions but can stop processing the queue entirely. In other words, if the sequencers censor or are down, they are so for everyone. An inclusion becomes due after 9m 36s. From then on, a whitelisted proposer cannot publish another proposal without processing up to ten due inclusions.	Only the whitelisted proposers can publish state roots on L1, so in the event of failure the withdrawals are frozen. Proposing is gated by PreconfWhitelist, which selects a single active operator for the current epoch and has no permissionless fallback.
17	Arbitrum Nova Arbitrum Nova	Fraud proofs allow 10 WHITELISTED actors watching the chain to prove that the state is incorrect. At least 5 Challengers are external to the Operator. Interactive proofs (INT) require multiple transactions over time to resolve. There is a 6d 8h challenge period.	All of the data needed for proof construction is published on Ethereum L1.	In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 1d delay on this operation.	Anyone can become a Proposer after 28d of inactivity from the currently whitelisted Proposers.
18	ADI Chain ADI Chain	STARKs and SNARKs are zero knowledge proofs that ensure state correctness. STARKs proofs are wrapped in SNARKs proofs for efficiency. SNARKs require a trusted setup.	All of the data (SD = state diffs) needed for proof construction is published onchain.	Users can submit transactions to an L1 queue, but can't force them. The sequencers cannot selectively skip transactions but can stop processing the queue entirely. In other words, if the sequencers censor or are down, they are so for everyone.	Only the whitelisted proposers can publish state roots on L1, so in the event of failure the withdrawals are frozen.
19	Facet Facet	Actors watching the chain can challenge state proposals, and challenged proposals must provide ZK proofs. SNARKs are zero knowledge proofs that ensure state correctness, but require trusted setup.	All of the data needed for proof construction is published on Ethereum L1.	Users can self sequence transactions by sending them on L1. There is no privileged operator.	Anyone can propose blocks if accompanied by a validity proof. Only the whitelisted proposers can propose state roots for recent blocks optimistically. Anyone can propose optimistically for L2 blocks that are older than 14d.
20	INTMAX INTMAX	SNARKs are succinct zero knowledge proofs that ensure state correctness, but require trusted setup.	All data required for payments and withdrawals is self custodied by users.	In the event of a sequencer failure, users can force transactions to be included in the project's chain by sending them to L1. There can be up to a 7d delay on this operation. Proposing new blocks requires creating ZK proofs.	If the Proposer fails, users can leverage the source available prover to submit proofs to the L1 bridge.
21	Honeypot v2 Honeypot v2	Fraud proofs allow actors watching the chain to prove that the state is incorrect. Interactive proofs (INT) require multiple transactions over time to resolve.	All of the data needed for proof construction is published on Ethereum L1.	Users can self sequence transactions by sending them on L1. There is no privileged operator.	Anyone can be a Proposer and propose new roots to the L1 bridge.
22	Aztec Aztec	SNARKs are succinct zero knowledge proofs that ensure state correctness, but require trusted setup.	State diffs needed to reconstruct the L2 state are published in Ethereum blobs. Public transaction bodies and client CHONK proofs propagate offchain, so withholding them can prevent permissionless proving; the affected pending checkpoints expire and are pruned rather than finalized.	Users can permissionlessly become a sequencer by staking 200.00 K AZTEC to join the queue and wait to obtain committee-based block production rights. If the pseudo-randomly sampled committees censor proposals, anyone who bonds 332.00 M AZTEC will join the escape hatch candidate set. Every 2d 23h, a candidate is pseudo-randomly selected to propose and prove checkpoints fully autonomously. A candidate remains in the set until they are selected or leave voluntarily.	Checkpoint proposals come from the open sequencer set, with the escape hatch providing a bonded fallback if the sampled committees are censoring or unavailable. Anyone with access to the required hardware can submit epoch root proofs which finalize the proven checkpoints.
23	Ethscriptions Ethscriptions	Actors watching the chain can challenge state proposals, and challenged proposals must provide ZK proofs. SNARKs are zero knowledge proofs that ensure state correctness, but require trusted setup.	All of the data needed for proof construction is published on Ethereum L1.	Users can self sequence transactions by sending them on L1. There is no privileged operator.	Anyone can propose blocks if accompanied by a validity proof. Only the whitelisted proposers can propose state roots for recent blocks optimistically. Anyone can propose optimistically for L2 blocks that are older than 14d.

Search