Skip to content
DeFi Reference

MEV burn, execution tickets, FOCIL, and timing games

1. TL;DR

Section titled “1. TL;DR”

Once block building is a market, the protocol can try to capture the auction’s proceeds (MEV burn), sell proposing rights explicitly (execution tickets and execution auctions), and constrain what builders may leave out (inclusion lists, FOCIL). All of these fight the same enemy: a single proposer who can wait, peek, or censor, which is why proposers rationally delay their blocks (“timing games”) and why leaderless and multi-proposer designs try to remove the “last look”. None is live on Ethereum mainnet yet; ePBS is scheduled for Glamsterdam and FOCIL for the following fork (as of 2026-08).

2. Explain it simply

Section titled “2. Explain it simply”

Analogy

Section titled “Analogy”

A town auctions a market stall every twelve seconds. Today the auctioneer keeps the winning bid. MEV burn says the town should keep most of it and let the auctioneer keep only a tip, so nobody fights to be auctioneer. Execution tickets say: sell raffle tickets in advance, draw one per slot, and let the winner run the stall, so the person who counts votes (the validator) is no longer the person selling the stall. Inclusion lists say: whoever runs the stall must display every product the town’s inspectors saw waiting outside. Timing games are what happens when the auctioneer learns that waiting until the last second brings in higher bids, and starts closing later and later.

ELI5

Section titled “ELI5”

Ethereum picks one validator every twelve seconds to publish a block, and that validator sells the block to the highest-bidding builder. Three problems follow. First, the validator pockets the bid, so big staking companies earn more and small ones less. Second, bids keep rising within the twelve seconds, so validators delay publishing to earn more, which makes the network wobble. Third, the builder can leave transactions out. The proposals here fix those: burn most of the bid so nobody cares who the validator is; sell the right to build in advance with a lottery; and force builders to include transactions that a committee saw waiting.

Step-by-step walkthrough

Section titled “Step-by-step walkthrough”

One slot under MEV burn (“a simple design”), with the attester floor parameter D=2D = 2 s:

  1. Before (t = −4 s): builders hold balances on chain (A: 40 ETH, B: 35 ETH). A sudden opportunity makes this slot worth about 30 ETH.
  2. Bidding (t = −4 to −2 s): bids specify a payload base fee (burned) and a payload tip (to the proposer). By t = −2 s the top observed base fee is A’s 27 ETH with tip 0.5 ETH; B has 26 ETH + 0.4 ETH.
  3. Floor freeze (t = −2 s): every honest attester records the top base fee it has seen, 27 ETH, as its subjective floor.
  4. Selection (t = 0): the proposer must include a bid whose base fee is ≥ the floor. A raises to base fee 28 ETH, tip 1 ETH; the proposer selects it (the highest tip among floor-satisfying bids).
  5. After: 28 ETH is burned regardless of whether A reveals a valid payload; the proposer’s balance is +1 ETH; A’s balance falls by 29 ETH and A recovers the ~30 ETH of MEV by executing its payload, netting ~1 ETH. Without MEV burn the proposer would have received ~29 ETH.

Common misconceptions

Section titled “Common misconceptions”
  • Myth: MEV burn destroys MEV. Reality: the MEV is still extracted by the builder; what changes is that the builder’s payment goes to ETH holders via burn instead of to a randomly selected proposer.
  • Myth: Execution tickets are a form of PBS. Reality: they are orthogonal: tickets create a primary market for future proposing rights; a just-in-time block auction (MEV-Boost or ePBS) can still run as a secondary market once a ticket is assigned.
  • Myth: Inclusion lists force builders to include everything. Reality: FOCIL lists are built from transactions committee members saw in the mempool and are conditional on block space; a full block remains valid.
  • Myth: Timing games are cheating. Reality: the honest spec says publish at t = 0, but delaying is individually rational and is indistinguishable from slow home-staker latency; a median proposer already receives its header 735 ms late without trying (as of 2024-02, secondary).
  • Myth: Timing games create new MEV. Reality: they are zero-sum across proposers: slot N’s extra value is taken from slot N+1’s auction.

If you only remember one thing

Section titled “If you only remember one thing”

Every design on this page attacks the proposer’s monopoly on when and whether things get included; burn removes the reward, tickets move it, inclusion lists constrain it, and leaderless auctions try to eliminate the last look.

3. How it works

Section titled “3. How it works”

Timing games

Section titled “Timing games”

Ethereum slots are 12 s; attesters vote at t=4t=4 s for the block they have seen, and a block that reaches the deadline late loses attestations. Proposer boost gives the current proposer 40% of the committee’s weight, and honest-reorg rules let the next proposer reorg a block with under 20% attestation weight. Block value V(t)V(t) increases within the slot as more MEV accrues, so a proposer choosing publication time tt maximizes

E[payoff(t)]=V(t) (1−pmiss(t))\mathbb{E}[\text{payoff}(t)] = V(t)\,\big(1 - p_{\text{miss}}(t)\big)

where pmiss(t)p_{\text{miss}}(t) is the probability the block is orphaned. Schwarz-Schilling and Neuder show the all-delay equilibrium is no better for any validator than honesty, but there is “money to be made on the path to the equilibrium”, so honesty is not an equilibrium. Practical constraints are the sidecar’s 950 ms getHeader timeout and relay cutoffs (getHeader by 3 s, getPayload by 4 s, as of 2023-12). Mitigations range from relay-enforced deadlines to missed-slot penalties, an earlier attestation deadline, or sub-slot proposer rotation.

The last look and leaderless auctions

Section titled “The last look and leaderless auctions”

Paradigm’s Leaderless Auctions generalize the problem: any auction with a leader gives that leader a free option to bid last. With 3f+13f+1 participants of whom 2f+12f+1 are honest, each threshold-encrypts a bid (decryptable by f+1f+1), gossips it in round 1 and its bid set in round 2; a bid in f+1f+1 bid sets is provably on time, giving no latecomers and no free option. Paradigm’s timing-advantage paper shows a later mover earns a strictly positive expected profit equal to an exchange option’s value, with “timing pressure” equal to its theta.

MEV burn (a simple design)

Section titled “MEV burn (a simple design)”

Drake’s design is an add-on to ePBS. Bids carry a payload base fee β\beta (burned) and a payload tip τ\tau (to the proposer), with β+τ≤\beta + \tau \le the builder’s on-chain balance. Attesters impose a subjective floor, the top β\beta observed DD seconds before proposal time, and only vote for proposals with β≥floor\beta \ge \text{floor}; the burn thus approximates the second-highest builder valuation. To avoid a “builder race to infinity”, the required pre-execution balance is capped at MM ETH (e.g. 32) and payloads that cannot cover β\beta post-execution are invalidated. Claimed benefits: lower reward variance, weaker reorg and equivocation incentives, no “rugpool” risk, a censorship-resistance forcing function via the floor, and ETH supply falling ~2.5× faster since the Merge (~270K vs ~110K ETH burned, as of 2023-05).

Execution tickets and execution auctions

Section titled “Execution tickets and execution auctions”

Execution tickets (Drake & Neuder, 2023-12) split each slot into a beacon round and an execution round. The beacon proposer (a validator) publishes a beacon block containing only an inclusion list; the execution proposer, drawn at random from holders of protocol-sold tickets, publishes the execution payload. Tickets are priced by a 1559-style controller targeting a fixed supply; the full ticket price is burned. A back-of-envelope estimate: with 2,628,000 slots per year and 890,000 validators a validator proposes about three times a year, so the implied “price of being a proposer” is ~0.11 ETH (as of 2023-12), while MEV-Boost payments were stable near 800 ETH/day. Execution auctions (Burian & Crapis, 2024-06) replace the lottery with a deterministic slot auction held e.g. 32 slots in advance; both remove MEV from validator rewards and blunt timing games for the beacon proposer, but shift them to execution proposers.

Inclusion lists and FOCIL

Section titled “Inclusion lists and FOCIL”

Fork-Choice enforced Inclusion Lists (Thiery, Monnot, D’Amato, Ma, 2024-06): each slot a committee of mm validators each publishes a local list of pending mempool transactions (sender, gas limit); the proposer includes the deterministic aggregate Agg(IL1,…,ILm)=⋃iILi\text{Agg}(\text{IL}_1,\dots,\text{IL}_m) = \bigcup_i \text{IL}_i; attesters compute their own aggregate and reject the block if the proposer’s aggregate omits transactions they saw (Eval\text{Eval} returns false), subject to conditionality (a full block is still valid). The example timeline releases local lists at slot n−1n-1, t=6t = 6 s, freezes views at t=9t = 9 s, and checks at slot nn, t=4t = 4 s. Motivation: two of the top three builders were filtering sanctioned addresses while 90% of validators were not (as of 2024-06). FOCIL became EIP-7805.

4. Worked numeric example

Section titled “4. Worked numeric example”

Continue the §2 slot. Opportunity value ≈30\approx 30 ETH; builder A values the block at vA=30v_A = 30, builder B at vB=29v_B = 29 (B lacks one bundle).

Under MEV-Boost. Ascending auction: winning bid ≈vB+ε=29.0\approx v_B + \varepsilon = 29.0 ETH to the proposer; A keeps 1 ETH.

Under MEV burn (D=2D = 2 s). At t=−2t = -2 s the top observed base fee is 27 ETH, so the floor is 27. A submits β=28,τ=1\beta = 28, \tau = 1; the proposer selects it (floor satisfied, highest tip). Burned: 28 ETH. Proposer: 1 ETH. Builder A: 30−29=130 - 29 = 1 ETH. Proposer income falls from 29 to 1 ETH; ETH holders gain 28 ETH of supply reduction.

Timing-game expected value. Let V(t)=0.05+0.004 tV(t) = 0.05 + 0.004\,t ETH for a normal slot and pmiss(t)=0p_{\text{miss}}(t) = 0 for t≤2t \le 2, rising by 0.1 per 0.5 s after that. Then

E[t=0]=0.050\mathbb{E}[t=0] = 0.050, E[t=2]=0.058\mathbb{E}[t=2] = 0.058, E[t=2.5]=0.060×0.9=0.054\mathbb{E}[t=2.5] = 0.060 \times 0.9 = 0.054, E[t=3]=0.062×0.8=0.050\mathbb{E}[t=3] = 0.062 \times 0.8 = 0.050.

The proposer delays to t≈2t \approx 2 s, gaining 16% on this slot at the expense of the next proposer. Under execution tickets the beacon proposer’s block has no execution payload, so V(t)V(t) is flat and the incentive disappears for validators (it moves to ticket holders).

Execution ticket pricing. A risk-neutral buyer pays up to the expected per-slot execution reward μR\mu_{\mathcal{R}} per ticket; with ~800 ETH/day across 7,200 slots, μR≈0.11\mu_{\mathcal{R}} \approx 0.11 ETH, matching Drake & Neuder’s estimate (as of 2023-12).

5. Where it’s used

Section titled “5. Where it’s used”

Ethereum

Section titled “Ethereum”
  • MEV-Boost timing games — relay cutoffs getHeader ≤ 3 s, getPayload ≤ 4 s (as of 2023-12); median header delay 735 ms (as of 2024-02, dataalways, secondary). See /mev/pbs-and-epbs/.
  • EIP-7805 (FOCIL) — headliner for the Hegotá fork after Glamsterdam (as of 2026-08, secondary).
  • MEV burn — research design layered on ePBS (EIP-7732); Wahrstätter’s simulations (“In a post MEV-Burn world”, 2023-10) estimate its effect on issuance.
  • Execution tickets / auctions — research track (Attester-Proposer Separation); interacts with based preconfirmations, see /mev/based-rollups-and-preconfs/.
  • Leaderless auctions — designed for Angstrom’s top-of-block auction; suggested for ePBS, UniswapX/MEV-Share, and L2 sequencers.

Solana

Section titled “Solana”
  • MEV burn — n/a as a builder-bid burn: there is no builder auction to tax. The related mechanism is fee burning: 50% of base and priority fees were burned, 50% paid to the leader (Helius, as of 2024); Jito tips are not burned.
  • Execution tickets — n/a; the stake-weighted leader schedule, known an epoch ahead, resembles the “perpetual tickets” the ET post ascribes to today’s Ethereum.
  • Inclusion lists — n/a; the leader has full discretion; BAM’s TEE attestations (2025-07) are the emerging way to prove a leader followed a sequence.
  • Timing games — different shape: leaders get four consecutive ~400 ms slots and searchers race on latency rather than proposers on delay. See /mev/solana-mev/.

6. Risks, attacks, and incidents

Section titled “6. Risks, attacks, and incidents”
  • Timing games in production (2023–2024). Relay bid data showed a spike of winning headers between 2,000 and 2,500 ms into the slot and some operators combining low latency with late blocks (ethresear.ch/17612 discussion, as of 2023-12); sophisticated proposers at t≥3t \ge 3 s versus home stakers at t<1t < 1 s threaten the level playing field (Aestus, as of 2023-12).
  • April 2023 relay patches and reorgs. Latency added to relay publication after the unbundling exploit interacted with honest reorgs to produce a 5× spike in reorged blocks (13 blocks, 4.3%, in the worst hour, as of 2023-04), a live demonstration of how sub-slot timing affects consensus.
  • Rugpool risk. Whenever a staking operator’s collateral is worth less than a single MEV spike, stealing the spike from the pool is rational (Drake); burn or tickets remove the spike from the validator.
  • Builder race to infinity. Without the balance cap MM, ePBS with burn lets the best-capitalized builder win every spike above the second-largest balance (Drake, 2023-05).
  • Splitting and late-bidding attacks on MEV burn. A dishonest proposer can split attesters on whether the floor was met; builders could bid only inside the DD-second window to zero the floor. Drake argues late bidding is irrational unless all capable builders collude; Wahrstätter (2023-11) proposes rewards for early bids.
  • Inclusion-list bribery. Proposers can be bribed to publish empty lists; FOCIL’s committee is the response, but “Uncrowdable inclusion lists” (2024-04) shows tension between lists, preconfirmations, and proposer commitments.

7. Open problems

Section titled “7. Open problems”
  • Which primary market: tickets or auctions? Burian & Crapis compare random tickets with deterministic slot auctions; Terence compares block-auction ePBS with tickets, noting ET pushes the execution deadline into the second half of the slot, “a significant mental shift for applications” (as of 2024-04).
  • Does MEV burn need ePBS first? The “simple design” needs attesters to observe bids in protocol; the ET variant burns the ticket price instead and loses the “cost of censorship” property (2023-12).
  • Sustainable timing-game mitigation. Relay deadlines are “unsustainable and messy” social coordination (17612); missed-slot penalties and earlier attestation deadlines trade liveness for fairness; random attestation deadlines do not help because committees are large enough to be split reproducibly.
  • Multi-proposer designs. Paradigm’s timing-advantage paper motivates multiple concurrent proposers, FOCIL, threshold encryption, and leaderless auctions as complementary; how to combine them without bloating the slot is unresolved (as of 2025-05).

8. Ethereum vs Solana

Section titled “8. Ethereum vs Solana”
AspectEthereumSolana
Slot12 s, one proposer, attestation deadline at 4 s~400 ms, four consecutive slots per leader
Proposer’s timing leverdelay getHeader up to relay cutoffsnone comparable; leader already controls its window
Who gets the auction proceedsproposer (today); burn/tickets proposedleader + stakers via Jito tips (6% Jito fee)
Censorship constraintinclusion lists (FOCIL, EIP-7805) proposednone in protocol; BAM attestations, social pressure
Fee burnEIP-1559 base fee; MEV burn proposed50% of fees burned; tips not burned
Lookaheadproposer known ~1 epoch ahead; ET proposes longerleader schedule known a full epoch ahead

Ethereum’s designs are attempts to keep a large, unsophisticated validator set out of the MEV game while still letting the protocol price the auction; the details (burn floors, ticket lotteries, committee lists) all exist because the proposer is a single trusted-for-12-seconds leader. Solana starts from the opposite premise, a fast, professional leader with four consecutive slots, and so has not pursued burn or tickets; its equivalents are stake-weighted access, Jito’s off-chain auction, and now TEE-attested sequencing.

9. Reference doc

Section titled “9. Reference doc”

The reference

Section titled “The reference”

MEV burn—a simple design — Justin Drake, 2023-05-15. https://ethresear.ch/t/mev-burn-a-simple-design/15590

Summary of the reference

Section titled “Summary of the reference”

Part 1 recaps enshrined PBS as the base: builders hold on-chain balances in EOAs, gossip bids containing a payload commitment, a payload tip, a pubkey and signature; the proposer selects a bid within a window; the winning builder reveals the payload within a subsequent window; attesters enforce timeliness of both; the tip is paid to the proposer even if the payload is invalid or late.

Part 2 adds the burn: each bid also specifies a payload base fee no larger than the builder balance minus the tip; it is burned unconditionally. Honest attesters set a subjective payload base fee floor equal to the top base fee they observed DD seconds (e.g. 2) before honest proposal time, and honest proposers maximize the base fee. The post then fixes the “builder race to infinity” by relaxing the balance requirement to a maximum upfront base fee MM (e.g. 32 ETH) and invalidating payloads whose post-execution balance cannot pay, so a malicious builder can force an empty slot for at most MM but cannot steal spikes.

Part 3 covers technical remarks: prior art (D’Amato’s committee-driven smoothing; Domothy’s burn via proposer auctions), unbounded burn, a liveness proof under synchrony, efficient gossip (bids fit in one 1,500-byte packet; nodes keep only the top bid), the proposer’s second optimization game (guessing the floor), splitting attacks, late-bidding analysis via a prisoner’s dilemma, and inclusion lists as the censorship/preconf tool. It draws the analogy to EIP-1559: both rely on an honest attester majority, both are partial burns, both create on-chain oracles (congestion vs MEV).

Part 4 lists security benefits of smoothing: micro consensus stability (less reorg/equivocation/DoS incentive), macro stability (a rollup-hack-sized spike landing on one proposer), lower reward variance, rugpool protection, censorship resistance via the floor, and removal of the “toxic MEV whitewashing” dilemma. Part 5 lists economic benefits mirroring EIP-1559: lower staking participation and issuance, more “pristine economic bandwidth”, stable USD-denominated staking returns, diversified revenue streams, tax efficiency, and scarcity (supply would have fallen ~2.5× faster since the Merge). Part 6 gives the mental model: inclusion causes congestion priced by EIP-1559; ordering causes contention priced by MEV burn.

Key quotes

Section titled “Key quotes”

“We describe a simple enshrined PBS add-on to smooth and redistribute MEV spikes. Spike smoothing yields security benefits. Redistribution yields economic benefits like EIP-1559.” (TLDR)

“Honest attesters set their payload base fee floor to the top builder base fee observed D seconds (e.g. D = 2) prior to the time honest proposers propose.” (Part 2)

“for exceptionally large MEV spikes the most capitalised builder has the power to capture all MEV above the second-largest builder balance.” (Part 2, builder race to infinity)

“Competition for inclusion leads to congestion, and competition for ordering leads to contention.” (Part 6)

How to read the original

Section titled “How to read the original”

Background: ePBS bid/reveal flow, EIP-1559, and what attesters do. Read the TLDR, Parts 1–2, and Part 6 first; Parts 4–5 are argument, not mechanism. The hardest paragraph is the floor definition: the floor is subjective (each attester’s own observation DD seconds before), so the proposer can only be sure of satisfying every honest attester under a synchrony assumption that bids propagate within DD; this is why the liveness proof in Part 3 needs synchrony.

What changed since

Section titled “What changed since”
  • 2023-10: Wahrstätter’s simulations quantified burn amounts and issuance effects; 2023-11: a variant adding rewards for earlier bidding.
  • 2023-12: Execution tickets proposed burning the ticket price instead of a per-slot floor.
  • 2024-06: Execution auctions compared deterministic vs lottery allocation.
  • 2025–2026: ePBS (EIP-7732) scheduled for Glamsterdam without a burn component; MEV burn remains research (as of 2026-08).

Secondary references

Section titled “Secondary references”
  1. “Execution Tickets” (Drake & Neuder, 2023) — read if you want the alternative that moves MEV off validators entirely.
  2. “Timing Games: Implications and Possible Mitigations” (Schwarz-Schilling & Neuder, 2023) — read if you want why proposers delay and the mitigation menu.
  3. “FOCIL” (Thiery et al., 2024) — read if you want the censorship-resistance piece.
  4. “Leaderless Auctions” (Paradigm, 2024) — read if you want the last-look problem solved cryptographically.
  5. “Committee-driven MEV smoothing” (D’Amato, 2021) — read if you want the prior art the burn design builds on.

The reference

Section titled “The reference”

Execution Tickets — Justin Drake, Mike Neuder (with Francesco D’Amato and Barnabé Monnot), 2023-12-23. https://ethresear.ch/t/execution-tickets/17944

Summary of the reference

Section titled “Summary of the reference”

The “High-level view” distills the design into three parts: an in-protocol market for tickets that entitle the owner to a future block proposal right, priced dynamically to target a supply; the existing lottery for beacon proposers, whose beacon block now carries an inclusion list instead of an execution payload; and a second lottery drawing a ticket to select the execution proposer, who posts collateral per ticket and is slashed for equivocation or absence. A back-of-envelope estimate prices today’s implicit ticket at ~0.11 ETH. Tickets are declared orthogonal to PBS (a JIT auction may still run as a secondary market) and motivated by “firewalling off” the validator set from MEV’s centralizing forces.

“Definitions” and “Design” lay out the slot as a beacon round followed by an execution round, each with its own attesters, and walk through three slots including a missed execution block (the next beacon proposer builds on the previous beacon block; the inclusion list may execute instead). Key design remarks: two lotteries instead of one; a 1559-style primary market possibly implemented as an EVM opcode; separation of the validator set from ticket holders, who will be “preconfirmers for L2s, hedge funds, builders, and HFT firms”; lookahead lengths; and fork-choice robustness with two committees.

“Analysis” covers validating and staking (simpler validator tasks, reduced timing-game incentive since beacon block value no longer rises, MEV removed from validator rewards, safer delegation and no rugpools), roadmap compatibility (full ticket price burned, losing the cost-of-censorship benefit of the floor-based burn; compatibility with preconfirmations, with the caveat that execution proposers as preconfirmers “closely mirrors the mev-geth regime”; MEV-Boost payments stable at ~800 ETH/day), the secondary market (fungible unassigned tickets vs non-fungible assigned ones), and multi-slot MEV with five mitigations (randomness, honest inclusion lists, block or list maximization, encrypted mempools, upfront missed-slot fees). It closes with open questions on pricing mechanics, fork-choice implications, and preconfirmations.

Key quotes

Section titled “Key quotes”

“Tickets entitle the owner to future block proposal rights. Using a dynamic pricing mechanism, we can target a certain number of tickets in circulation” (§High-level view)

“Execution tickets aim to detangle the MEV rewards from the beacon block production to ‘firewall off’ the decentralized validator set from these centralizing forces.” (§High-level view)

“Importantly, the incentive to play proposer timing games is greatly reduced, because the value of the beacon block should not increase as time passes.” (§Analysis, Validating & staking)

“Historical data for prices paid in the mev-boost auctions are also extremely stable at around 800 ETH/day.” (§Roadmap compatibility)

How to read the original

Section titled “How to read the original”

Background: PBS/MEV-Boost, inclusion lists, and the beacon chain’s proposer lottery. Read the high-level view and the three-slot figure discussion, then the multi-slot MEV section; skip the secondary-market figure on first pass. The hardest paragraph is the fork-choice discussion: if execution blocks carry beacon attestations, a dominant ticket holder could censor them; if they do not, attesters might vote late, so where votes live determines who can grief whom.

What changed since

Section titled “What changed since”
  • 2024-04: “Block-auction ePBS versus Execution Ticket” compared the two on timing games and deadlines.
  • 2024-06: “Execution Auctions as an Alternative to Execution Tickets” proposed deterministic slot auctions.
  • 2024–2025: Attester-Proposer Separation research continued alongside based preconfirmations; no ticket mechanism is scheduled for a fork (as of 2026-08).

Secondary references

Section titled “Secondary references”
  1. “Execution Auctions as an Alternative to Execution Tickets” (Burian & Crapis, 2024) — read if you want the economic comparison with formulas.
  2. “Block-auction ePBS versus Execution Ticket” (Terence, 2024) — read if you want the implementation-level differences.
  3. “Timing Advantages in Onchain Auctions” (Paradigm, 2025) — read if you want a model of how much a late mover can extract.

10. Sources

Section titled “10. Sources”