Summary

OP Mainnet’s stake-based transaction ordering system replaces pure gas-price priority with a two-tier approach: Phase 1 gives all accounts staking ≥100,000 OP tokens equal “fast lane” access (flat FIFO within stakers), while Phase 2 introduces a multiplier formula that scales priority proportionally to stake amount via a square-root curve, capped at 3× base priority. No token lockups are required. The PolicyEngineStaking contract manages stake positions and distributes priority multipliers. This is the first production deployment of stake-based sequencer ordering on a major L2.

Background: The Ordering Problem on L2s

L2 sequencers currently order transactions by:

  1. Gas price (most chains): higher tip = earlier inclusion, same as Ethereum mainnet
  2. FCFS (some chains): first-come-first-served based on network receipt time
  3. Private mempool: centralized sequencer has full discretion

Problems with gas price ordering on L2s:

  • Near-zero gas costs → ordering competition is won by infrastructure latency rather than economic merit
  • Priority fees paid by users capture no value for the ecosystem; they go to the sequencer operator
  • High-frequency bots can spam microsecond-precision submissions to win FCFS races

Stake-based ordering introduces a new dimension: transaction priority is partly determined by the sender’s on-chain relationship with the network (token stake), not just their willingness to pay in the current block.

Phase 1: Flat Priority Access

Design

  • Threshold: accounts staking ≥ 100,000 OP tokens receive “fast lane” priority
  • Ordering: within the fast lane, transactions are ordered FCFS (by network receipt time)
  • No stake gradation: staking 100,001 OP gives the same priority as staking 1,000,000 OP
  • Non-stakers: ordered after all fast-lane transactions, by gas price

Rationale

Phase 1 is intentionally simple: it tests whether stake-based ordering changes submission behavior without introducing complex multiplier math. The 100K OP threshold (~$100K-200K at typical prices) is high enough to require commitment but not so high that it excludes mid-sized market participants.

No lockup: stakes can be withdrawn at any time. This prevents the stake from acting as a long-term capital lockup (which would disadvantage smaller participants) while still creating a meaningful on-chain commitment.

Phase 2: Continuous Multiplier Formula

Effective Priority Gas

Phase 2 replaces the binary fast-lane with a continuous multiplier:

effectivePriorityGas = priorityGas × m_total(S, t)

Where:

  • priorityGas = the gas tip set by the transaction (standard priority fee)
  • m_total(S, t) = total priority multiplier for account with stake S at time t
  • effectivePriorityGas = the value used for ordering (not the actual fee paid)

Key: effectivePriorityGas is used only for ordering; the actual fee paid is still priorityGas. Stakers do not pay more or less in fees — they receive ordering priority.

Multiplier Formula

m_total(S, t) = 1 + m_stake(S) × f(t)

Where:

  • m_stake(S) = stake-based component (see formula below)
  • f(t) = time-decay factor (reduces multiplier for long-held stakes; prevents indefinite accumulation)

Square-root curve for stake component:

m_stake(S) = sqrt(S / S_ref)

Where S_ref is a reference stake amount. This gives:

  • Diminishing marginal returns to staking (doubling stake increases multiplier by √2, not 2×)
  • Progressive scaling: small stakers get proportionally more benefit from each additional token than large stakers

Maximum multiplier: capped at 3× base priority. An account staking unlimited OP cannot achieve more than 3× the ordering priority of an unstaked account with the same gas tip.

Cap Rationale

The 3× cap prevents stake from being the dominant ordering criterion:

  • An unstaked transaction with 4× the gas tip still beats a staked transaction
  • Stake provides a meaningful edge in competitive situations (where tips are similar) but doesn’t allow stakers to outbid everyone regardless of gas price
  • Maintains the price-discovery role of priority fees

PolicyEngineStaking Contract

The PolicyEngineStaking contract manages the stake system:

Functions:

  • stake(amount): deposit OP tokens, immediately active (no lockup period)
  • unstake(amount): withdraw OP tokens, immediately returned (no lockup period)
  • getMultiplier(address): return the current m_total for an address
  • updatePolicy(params): governance-controlled parameter updates

Composability: the contract is readable by the sequencer client. The sequencer queries getMultiplier(sender) for each incoming transaction to compute effectivePriorityGas.

Decentralization path: in a future multi-sequencer design, all sequencers read from the same PolicyEngineStaking contract → consistent ordering decisions across sequencers → ordering disputes resolvable by re-running the formula.

MEV Implications

Reduced Gas Wars

Under pure gas-price ordering, searchers competing for the same MEV opportunity bid up gas prices. With stake-based ordering:

  • A searcher with sufficient stake has a priority edge without bidding up gas
  • Multiple searchers competing for the same opportunity may all stake rather than all bidding gas
  • Net effect: MEV extraction redistributes from gas fees (burned) to stake (which accumulates multipliers)

Sequencer MEV

The sequencer operator (currently OP Labs) sees transaction content before ordering. Stake-based ordering does not prevent the sequencer from extracting MEV directly. However:

  • If the sequencer’s ordering follows the published formula, it cannot arbitrarily reorder transactions
  • Verifiable ordering: external parties can verify that the sequencer respected multipliers by checking on-chain stake positions
  • Future proof: PoS-style sequencer selection (multiple sequencers) could rotate ordering rights among stakers

Protocol-Captured MEV

A portion of ordering priority accrues to OP token stakers rather than validators (as on Ethereum mainnet). This captures MEV value at the L2 ecosystem level:

  • Stakers (long-term OP holders) benefit from ordering priority
  • The protocol can direct a fraction of this value to public goods (Optimism RetroPGF)
  • MEV becomes partially a protocol resource rather than purely a builder/validator resource

Relationship to L2 Sequencer Debates

From the Gogol analysis (Insight #6): partial centralization increases predictability without compromising trust. The stake-based ordering system is a concrete instantiation:

  • Centralized sequencer ordering: the sequencer uses the formula, but users cannot verify in real-time
  • Deterministic formula: the formula is public and on-chain → ordering is auditable post-hoc
  • History-rewriting: impossible because L1 settlement is canonical
  • Censorship: possible (sequencer could ignore formula) but detectable via stake position audit

Open Questions

❓ What is the empirically observed effect of Phase 1 on transaction ordering and MEV extraction rates? (Data not yet published as of April 2026.)

❓ Does the square-root curve produce the intended distribution of stakers, or does it incentivize concentration at the 3× cap threshold?

❓ How does the stake-based ordering interact with FCFS-based frontrunning? (A staker with 3× priority still loses to an unstaked account with 4× the gas tip, but beating a staked competitor requires 3× more gas.)

❓ What is the minimum stake to make a meaningful difference in competitive MEV extraction? (Depends on the distribution of gas tips in the blocks where MEV opportunities arise.)

Timeline

  • 2026-02 — Phase 1 update published (February 2026 1TS update)
  • 2026 (planned) — Phase 2 deployment on OP Mainnet

See Also