Pith. sign in

REVIEW 4 major objections 5 minor 34 references

Resonance: Transaction Fees for Heterogeneous Computation

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read A broker competition mechanism makes blockchain fees efficient and margin-free.

desk verdict A novel broker-based fee mechanism with a real but fixable proof gap in its headline efficiency theorem; worth reviewing after the authors repair Theorem 5.2. read the letter →

arxiv 2411.11789 v2 pith:A273T43X submitted 2024-11-18 cs.GT

classification cs.GT MSC 91A1091B2691B03
keywords transactionfeemechanismtwo-sidedmarketbrokercompetitionpureNashequilibriumsurplusmaximizationheterogeneouscomputationblockchaindesign
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Resonance is a transaction fee mechanism for blockchains in which users and nodes are fully heterogeneous: each user has an arbitrary private value for inclusion and each node an arbitrary private cost for any bundle of transactions, with arbitrary validity constraints on allocations. The paper's central claim is that delegating pricing and allocation to competing brokers — sophisticated agents who know everyone's types and propose complete routings — yields efficient outcomes in equilibrium. Specifically, with at least two brokers, every pure Nash equilibrium is welfare-maximizing among all valid allocations and surplus-maximizing among all weakly budget-balanced routings, with zero broker margin, so users and nodes capture all generated value. The authors also prove that simpler multi-dimensional fee markets with shared unit prices cannot achieve this: even given perfect pricing and allocation oracles, such markets can lose almost all surplus once nodes have heterogeneous costs (Theorem A.3). Resonance is budget-balanced, individually rational, and computationally light, and it minimizes the need for strategization by users and nodes.

What carries the argument

The load-bearing object is the broker proposal combined with a surplus-maximizing selection rule. A routing specifies a valid allocation $\alpha : T \to 2^N$, a user payment rule $\pi$, and a node payment rule $\varphi$, and its margin is the net cash flow $\sum_{t\in T} \pi(t) - \sum_{n\in N} \varphi(n)$. Brokers compete by proposing routings; the mechanism selects the highest-surplus budget-balanced proposal and transfers the margin to the winning broker. Competition among at least two brokers drives margins to zero, because any positive margin can be undercut by a rival who reduces one transaction's payment slightly, raising surplus and winning the auction. That zero-margin property converts surplus maximization into welfare maximization and is the key identity behind Theorem 5.2.

What would settle it

One concrete test: in a small instance with two brokers, two transactions, and two nodes, run a full search over both agents' action spaces and verify that every pure Nash equilibrium has zero margin and maximum surplus. Finding any equilibrium with strictly positive margin or suboptimal surplus would refute Claim 5.3 and Theorem 5.2. A second, easier test is to endow one broker with slightly noisy information about a node's cost; the paper's assumptions no longer hold, so an equilibrium with positive margin would show where the efficiency guarantee breaks.

Watch

Extended reading notes

Core claim

The central discovery is that the two hard problems in heterogeneous fee markets — computing a good allocation and discovering individualized prices — can be offloaded to a new class of agents called brokers. Each broker submits a routing, meaning a valid allocation plus a payment from every transaction and a payment to every node; the mechanism keeps the budget-balanced routing with the highest surplus according to the reported types, provided it gives every transaction and node non-negative utility, and pays the winning broker the routing's margin. The paper proves (Theorem 5.2) that with at least two brokers, at any pure Nash equilibrium the winning routing has zero margin and is surplus-maximizing among all weakly budget-balanced routings, hence also welfare-maximizing. With a single broker the outcome is welfare-maximizing but surplus is zero, and Resonance is dominant-strategy incentive-compatible for users and nodes; with multiple brokers it satisfies the weaker notion 'DSIC barring B', meaning users and nodes need not strategize against each other when brokers play a natural equilibrium.

Load-bearing premise

The theorems require brokers to know the exact private valuations and cost functions of every transaction and node, and to bear zero cost for proposing and operating routings; if broker information is incomplete or costly, the efficient zero-margin equilibrium is not delivered.

Editorial extensions

If this is right

  • If Resonance is used as a blockchain fee market, users and nodes can be told to report their valuations and costs truthfully; strategizing is relegated to brokers, whose competition replaces a centralized price oracle.
  • The mechanism supports arbitrary validity constraints, so it can express state-conflict rules, capacity limits, and multi-node transaction types such as multi-party computation in one framework.
  • Theorem A.3 implies that any fee mechanism that charges a shared vector of unit prices per resource can lose almost all surplus when nodes have heterogeneous costs, providing a formal limit on the EIP-1559-style approach and a direct motivation for per-transaction, per-node pricing.
  • Because Resonance is weakly budget-balanced and individually rational, it can be deployed on-chain without subsidy, and its per-proposal computations scale linearly in the number of agents and brokers.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The zero-margin equilibrium result depends on brokers having complete and free information about all types; in practice, broker specialization and estimation costs would leave positive margins, so the paper's efficiency guarantee should be read as an idealized benchmark rather than a literal prediction.
  • The DSIC-barring-B notion is weaker than it might appear: it guarantees users and nodes need not strategize against each other only when brokers coordinate on a common welfare-maximizing allocation with at least two zero-margin proposals, which the paper argues is the outcome of best-response dynamics but does not prove for all equilibria.
  • A testable extension would be to simulate broker best-response dynamics in a setting with noisy type estimates and see whether margins converge to the cost of information acquisition rather than to zero; the paper's claim would then become that competition reduces margins to broker costs, not to zero.
  • The grievance-vector attacks identified in Section 6 (single-agent rejection and private order flow) suggest that practical deployment needs additional safeguards; the paper states modifications exist but does not analyze them, so the mechanism in its stated form may not be robust to a single malicious low-reporting agent.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes Resonance, a broker-based transaction fee mechanism for a two-sided market with fully heterogeneous users (transactions with private valuations) and nodes (with private cost functions), subject to arbitrary allocation constraints. Brokers submit routings (allocation plus transaction and node payments); the mechanism selects the budget-balanced proposal with highest surplus under the reported types, checks individual rationality, and pays the winning broker the routing's margin. The paper claims budget balance, individual rationality for all parties, welfare/surplus efficiency at pure Nash equilibria (Theorems 5.1 and 5.2), a new incentive property termed DSIC barring B (Theorem 5.3), and computational tractability. An appendix analyzes limitations of multidimensional fee markets, arguing that individualized prices are necessary in heterogeneous settings.

Significance. If the main results were correct, Resonance would be a significant design contribution: it replaces on-chain price oracles and allocation algorithms with competition among sophisticated brokers, while keeping the user/node interface strategically simple. The general two-sided model with arbitrary validity constraints is well motivated by prover networks, AI/ML compute markets, and specialized blockchain infrastructure. The appendix's separation of pricing from allocation and its impossibility results for shared base fees are also useful. The paper is clearly written and self-contained, and Section 6 candidly discusses practical issues such as griefing, private order flow, broker specialization, and collusion. However, the central efficiency theorem (Theorem 5.2) is false as stated, and the proof contains additional gaps; these issues are load-bearing and prevent acceptance in the current form.

major comments (4)
  1. [Section 5.3, Theorem 5.2] Theorem 5.2 is false as stated. Consider T={t} with v_t=1, N={n} with c_n(·)=0, and B={b1,b2}, with b1 first in the tie-breaking order. Let both brokers propose R=(α(t)={n}, π(t)=2, φ(n)=2). This routing is budget-balanced (margin 0) and has surplus S(R|θ)=(1-2)+(2-0)=1, so it is a highest-surplus budget-balanced proposal. The mechanism selects b1 but then, because the transaction has negative utility under the truthful reported type, the IR check fails and the mechanism outputs the empty routing. The profile (θ,R,R) is a PNE: b1 deviating to the IR routing (π=1, φ=1) gets the same surplus and zero margin, so its payoff stays 0; any positive-margin deviation has surplus below 1 and does not win; b2 cannot win any deviation with surplus 1 because of the fixed tie-break, and deviations with surplus below 1 lose; a unilateral report change by the transaction either leaves the output empty or makes R pass IR only if the transaction reports at least 2, yielding true utility -1, while the node alone cannot make R pass. Thus the empty routing is output at a PNE, yet its welfare is 0 while the maximum welfare is 1, contradicting both bullets of Theorem 5.2. The proof's deviation arguments do not account for the fact that a losing broker's higher-surplus proposal may be blocked by another broker's non-IR proposal that wins the surplus comparison but is later rejected.
  2. [Section 5.3, proof of Theorem 5.2, Case 2] The proof of the equal-margin case constructs R'' from R' by decreasing the transaction payment to π'(t)-δ and asserts that R'' has strictly higher surplus and strictly higher margin than R. This is arithmetically incorrect: margin is defined as Σπ−Σφ, so lowering a transaction payment lowers the margin. Since R' has zero margin, R'' has margin −δ<0 and is not weakly budget-balanced, so the mechanism would not select it. A correct deviation would need to increase a payment or decrease a node payment, with a separate argument preserving individual rationality and surplus dominance; the proof as written does not provide this.
  3. [Section 5.3, proof of Theorem 5.2, Case 1] The proof assumes that if an alternative budget-balanced routing R' has higher surplus and higher margin than the incumbent R, then the winning broker can deviate to R' and collect margin γ(R'). This ignores the mechanism's individual-rationality check: if R' gives any transaction or node negative utility under the true reported types, the mechanism outputs the empty routing and the deviating broker receives zero, not γ(R'). The proof never shows that R' (or a suitable modification of it) is individually rational, so the claimed contradiction is not established even apart from the counterexample in the previous comment.
  4. [Section 5.4, Theorem 5.3] The proof constructs 'a surplus-maximizing routing' for the brokers without specifying the feasible set over which surplus is maximized. If the maximum is taken over all routings, surplus is unbounded because node payments can be increased without limit; if it is taken over weakly budget-balanced routings, the resulting routing need not be individually rational, in which case the constructed profile (θ,σ) can be the bad equilibrium described in the counterexample to Theorem 5.2. The assertion that σb' is 'welfare-maximizing with zero margin' is therefore unsupported. The definition of the broker strategy in the proof and the verification of both bullets of Definition 5.3 need to be made precise.
minor comments (5)
  1. [Section 4, Mechanism definition] The mechanism defines b* as the arg max over budget-balanced proposals, but does not specify what happens when no budget-balanced proposal exists (the arg max is over an empty set). The text should state explicitly that the mechanism then outputs the empty routing, as is implicitly assumed in Lemma 5.2.
  2. [Section 5.4, proof of Theorem 5.3] The sentence 'all routings in σ have the same allocation rule, and therefore the same surplus under any type vector' is not literally correct if the routings have different payment rules; the surpluses then differ by the respective margins. What matters for the argument is that the relative ranking of the proposals is independent of the reported types because all allocations coincide, and this should be stated accurately.
  3. [Section 5.3, proof of Theorem 5.2] Typo: 'budged-balanced' should be 'budget-balanced'.
  4. [Section 2, notation] The sentence 'We will denote by A the set A of all transactions and nodes' is redundant; it should read 'We denote by A the set of all transactions and nodes.'
  5. [Section 6, first paragraph] The text says 'As shown in Section 3, Resonance satisfies...', but the formal properties are established in Sections 4 and 5; the reference should be updated.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: Resonance's equilibrium results are derived from the mechanism's selection rule and agent incentives rather than assumed into the conclusions.

full rationale

I walked the derivation chain and found no load-bearing step that reduces to its own inputs. The mechanism is defined as selecting the highest-surplus budget-balanced broker proposal, and Theorems 5.1 and 5.2 prove efficiency by deviation arguments: Claim 5.2 constructs a budget-balanced routing whose margin equals the welfare of any target allocation, and Claim 5.3 proves that with at least two brokers an equilibrium must have zero margin by exhibiting a winning deviation for a losing broker. These are ordinary equilibrium arguments, not definitions of the conclusions. Definition 5.3 introduces a weak incentive notion, and Theorem 5.3 proves existence of the required broker action profile by explicit construction; the theorem does real work and is not a renaming of the definition. Appendix A's impossibility results are self-contained worst-case constructions with no fitted parameters and no dependence on the paper's own prior work. The only self-referential material is contextual (Ritual as a motivating application), and it is not load-bearing. I also noted the internal proof error in Theorem 5.2's equal-margin case: reducing a transaction payment by delta makes the constructed routing fail weak budget balance, so the written proof does not establish the surplus-maximality claim. That is a correctness risk, not a circularity, so it does not affect the circularity score.

Assumptions & free parameters 0 free parameters · 6 assumptions · 1 invented entities

The central claim rests on a small set of domain assumptions about the model, the strongest being broker omniscience and zero cost. No free parameters are fitted to data. The only ad hoc element is the informal equilibrium selection argument used to motivate DSIC barring B.

assumptions (6)
  • domain assumption Brokers know the exact private types (valuations and cost functions) of all transactions and nodes.
    Stated in Section 4 (Information Structure). This is necessary for brokers to submit optimal proposals; without it, the equilibrium results do not follow.
  • domain assumption Brokers have zero operating costs and are purely revenue-maximizing.
    The model defines broker utility as payment received; Section 6 (Broker Compensation) acknowledges nonzero costs are left to future work, so the efficiency theorems depend on this.
  • domain assumption The set of valid allocations V includes the empty allocation and is otherwise arbitrary.
    Section 2; this is a modeling choice that makes the mechanism general.
  • domain assumption Users and nodes have quasilinear utilities and report private types; no collusion is considered in the main analysis.
    Section 2 and Section 6 (Collusion and Sybil Resistance) notes collusion guarantees are impossible in the strictest form.
  • domain assumption Agent types have constant description size, so evaluating a routing's surplus takes O(|A|*|B|) time.
    Section 5.5 tractability claim relies on this; arbitrary cost functions might be expensive to evaluate.
  • ad hoc to paper A specific broker equilibrium (all brokers submitting the welfare-maximizing allocation with at least two zero margins) is reachable via natural competition.
    Section 5.4 justifies this with an informal best-response dynamics argument, not a theorem; it is needed for the DSIC barring B guarantee to be meaningful.
invented entities (1)
  • Brokers
    purpose: Sophisticated agents who submit routings (allocations + payment rules) to the mechanism, offloading allocation and pricing complexity from the on-chain auctioneer.
    Brokers are a theoretical construct; the paper assumes they are omniscient and costless, with no empirical or independent support for their existence or capabilities.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Resonance: Transaction Fees for Heterogeneous Computation." pith.science (2026). https://pith.science/paper/A273T43X

@misc{pith2026241111789,
  author       = {Pith},
  title        = {Pith review of: Resonance: Transaction Fees for Heterogeneous Computation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/A273T43X}},
  note         = {Machine review of arXiv:2411.11789}
}
read the original abstract

Blockchain networks are facing increasingly heterogeneous computational demands, and in response, protocol designers have started building specialized infrastructure to supply that demand. This paper introduces Resonance: a new kind of transaction fee mechanism for the general two-sided market setting (with users on one side and nodes on the other), where both sides of the market exhibit a high degree of heterogeneity. We allow users submitting transactions to have arbitrary valuations for inclusion, nodes responsible for executing transactions to incur arbitrary costs for running any bundle of transactions, and further allow for arbitrary additional constraints on what allocations are valid. These constraints can, for example, be used to prevent state conflicts by requiring transactions that utilize the same part of the network's state to not be executed in parallel. They also enable support for new transaction types, such as transactions that require multiple nodes for execution (e.g. to run multi-party computation for better transaction privacy). Resonance's design utilizes competition among sophisticated brokers to find individualized prices for each transaction and node. We show that at pure Nash equilibria, Resonance finds an efficient outcome and minimizes the need for strategization by users and nodes. It is also budget-balanced, individually rational for all parties, and computationally tractable.

Figures

Figures reproduced from arXiv: 2411.11789 by the authors.

Figure 1
Figure 1. Nonexistence of Collusion-Resistant Routings. Consider the above example where there are two transactions and two nodes. The first transaction (shown on the top left) requires both nodes to execute, and the second transaction (shown on the bottom left) only requires one node, and can be executed with either node. Nodes however only have capacity to execute one of the two transactions. There are therefore three valid… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 28 canonical work pages

  1. [1]

    https://docs.zksync.io/build

    zksync docs. https://docs.zksync.io/build

  2. [2]

    https://www.gelato.network/raas, 2022

    The all-in-one ethereum rollup as a service platform. https://www.gelato.network/raas, 2022

  3. [3]

    https://docs.gensyn.ai/litepaper, 2022

    The hyperscale, cost-efficient compute protocol for the world ’s deep learning models. https://docs.gensyn.ai/litepaper, 2022

  4. [4]

    https://ethresear.ch/t/rollup-as-a-service-opportun ities-and-challenges/13051 , 2022

    Rollup as a service: Opportunities and challenges. https://ethresear.ch/t/rollup-as-a-service-opportun ities-and-challenges/13051 , 2022

  5. [5]

    Multidimension al blockchain fees are (essentially) optimal

    Guillermo Angeris, Theo Diamandis, and Ciamac Moallemi. Multidimension al blockchain fees are (essentially) optimal. arXiv preprint arXiv:2402.08661 , 2024

  6. [6]

    The anatomy of proof generation

    Andy Arditi and Ye Zhang. The anatomy of proof generation. https://scroll.io/blog/proof-generation, 2022

  7. [7]

    Chainlink 2.0: Next steps in the evolution of decentralized oracle netw orks

    Lorenz Breidenbach, Christian Cachin, Benedict Chan, Alex Cove ntry, Steve Ellis, Ari Juels, Farinaz Koushanfar, Andrew Miller, Brendan Magauran, Dan iel Moroz, et al. Chainlink 2.0: Next steps in the evolution of decentralized oracle netw orks. Chainlink Labs, 1:1–136, 2021

  8. [8]

    Multidimensional eip 1559

    Vitalik Buterin. Multidimensional eip 1559. https://ethresear.ch/t/multidimensional-eip-1559/ 2022

Show all 34 references
  1. [9]

    Multidimensional gas pricing

    Vitalik Buterin. Multidimensional gas pricing. https://vitalik.eth.limo/general/2024/05/09/mul 2024

  2. [10]

    Eip-4844: Shard blob t ransac- tions

    Vitalik Buterin, Dankrad Feist, Diederik Loerakker, George Kad ianakis, Matt Garnett, Modi Taiwo, and Ansgar Dietrichs. Eip-4844: Shard blob t ransac- tions. https://eips.ethereum.org/EIPS/eip-4844, 2022. [Online serial]. Available: https://eips.ethereum.org/EIPS/eip-4844

  3. [11]

    The multip le subset sum problem

    Alberto Caprara, Hans Kellerer, and Ulrich Pferschy. The multip le subset sum problem. SIAM Journal on Optimization , 11(2):308–319, 2000. 18

  4. [12]

    Programmable boot strapping enables efficient homomorphic inference of deep neural networks

    Ilaria Chillotti, Marc Joye, and Pascal Paillier. Programmable boot strapping enables efficient homomorphic inference of deep neural networks. In Cyber Security Cryptogra- phy and Machine Learning: 5th International Symposium, CSC ML 2021, Be’er Sheva, Israel, July 8–9, 2021, Pr...

  5. [13]

    Collusion-resilience in transaction fee mechanism design

    Hao Chung, Tim Roughgarden, and Elaine Shi. Collusion-resilience in transaction fee mechanism design. arXiv preprint arXiv:2402.09321 , 2024

  6. [14]

    Foundations of transaction fee mecha nism design

    Hao Chung and Elaine Shi. Foundations of transaction fee mecha nism design. In Pro- ceedings of the 2023 Annual ACM-SIAM Symposium on Discrete A lgorithms (SODA) , pages 3856–3899. SIAM, 2023

  7. [15]

    Designing mul- tidimensional blockchain fee markets

    Theo Diamandis, Alex Evans, Tarun Chitra, and Guillermo Angeris. Designing mul- tidimensional blockchain fee markets. In 5th Conference on Advances in Financial Technologies (AFT 2023) . Schloss Dagstuhl-Leibniz-Zentrum f¨ ur Informatik, 2023

  8. [16]

    Barriers to collusion-resistant tra nsaction fee mechanisms

    Yotam Gafni and Aviv Yaish. Barriers to collusion-resistant tra nsaction fee mechanisms. arXiv preprint arXiv:2402.08564 , 2024

  9. [17]

    On coalition incentive c ompatibility

    Jerry Green and Jean-Jacques Laffont. On coalition incentive c ompatibility. The Review of Economic Studies , 46(2):243–254, 1979

  10. [18]

    Matic whitep aper

    Jaynti Kanani, Sandeep Nailwal, and Anurag Arjun. Matic whitep aper. Polygon, Ben- galuru, India, Tech. Rep., Sep , 2021

  11. [19]

    Cosmos whitepaper

    Jae Kwon and Ethan Buchman. Cosmos whitepaper. A Netw. Distrib. Ledgers , 27:1–32, 2019

  12. [20]

    Robust double auctions for resource allocation

    Arthur Lazzaretti, Charalampos Papamanthou, and Ismael H ishon-Rezaizadeh. Robust double auctions for resource allocation. Cryptology ePrint Archive , 2024

  13. [21]

    Optimalit y despite chaos in fee markets

    S Leonardos, D Reijsbergen, B Monnot, and G Piliouras. Optimalit y despite chaos in fee markets. financial cryptography and data secutiry’23, 2022

  14. [22]

    Optimal auction design

    Roger B Myerson. Optimal auction design. Mathematics of operations research, 6(1):58– 73, 1981

  15. [23]

    Efficient mechanism s for bilateral trading

    Roger B Myerson and Mark A Satterthwaite. Efficient mechanism s for bilateral trading. Journal of economic theory , 29(2):265–281, 1983

  16. [24]

    Bittensor: A peer-to-peer intelligence market

    Yuma Rao, Jacob Steeves, Ala Shaabana, Daniel Attevelt, and Matthew McAteer. Bittensor: A peer-to-peer intelligence market. arXiv preprint arXiv:2003.03917 , 2020

  17. [25]

    Transaction fee mechanism design

    Tim Roughgarden. Transaction fee mechanism design. ACM SIGecom Exchanges , 19(1):52–55, 2021

  18. [26]

    Modulus: An open modular des ign for interoperable and reusable machine learning

    Salvatore Vivona and Luca Vivona. Modulus: An open modular des ign for interoperable and reusable machine learning. 19

  19. [27]

    Mechanism design for zk-rollup prover markets

    Wenhao Wang, Lulu Zhou, Aviv Yaish, Fan Zhang, Ben Fisch, and B enjamin Livshits. Mechanism design for zk-rollup prover markets. arXiv preprint arXiv:2404.06495 , 2024

  20. [28]

    Polkadot: Vision for a heterogeneous multi-chain fr amework

    Gavin Wood. Polkadot: Vision for a heterogeneous multi-chain fr amework. White paper, 21(2327):4662, 2016

  21. [29]

    Multidimensional Blockchain Fees are (Essentially) Optimal

    Guy Zyskind, Yonatan Erez, Tom Langer, Itzik Grossman, and Lior Bondarevsky. Fhe- rollups: Scaling confidential smart contracts on ethereum and bey ond. A On Multi-Dimensional Fee Mechanisms This section provides a formal discussion on whether the multi-dimen sional fee market ...

  22. [30]

    In an inclusion-maximal allocation, the mechanism allocates a maximal set of transactions to nodes

    Inclusion-Maximal Allocations. In an inclusion-maximal allocation, the mechanism allocates a maximal set of transactions to nodes. That is, any alloca tion that allocates a strict superset of the transactions that are allocated under an inclusion-maximal 20 allocation must be ...

  23. [31]

    In a fee-maximal allocation, the mechanism allocates trans- actions that maximize the total revenue from base fees

    Fee-Maximal Allocations. In a fee-maximal allocation, the mechanism allocates trans- actions that maximize the total revenue from base fees. Computin g a multi-dimensional fee-maximal allocation involves solving the multiple subset-sum problem, which is NP- Hard, and further h...

  24. [32]

    An oracle-maximal allocation maximizes surplus given full knowledge of transaction valuations and node costs

    Oracle-Maximal Allocations. An oracle-maximal allocation maximizes surplus given full knowledge of transaction valuations and node costs. It is uncle ar exactly how the mechanism would obtain such knowledge (though priority fees perha ps provide a coarse approximation), hence ...

  25. [33]

    If p > 1, then only the empty allocation lies in Vp, and all fee- maximal allocations attain 0 surplus

    It follows that the surplus-maximizing allocation includes the valuable transaction tk+1 and k − 1 of the other transactions, whence we find that OPT = 1 2 + (k − 1) · 1 2(k − 1) = 1 Next, let p be any unit base fee. If p > 1, then only the empty allocation lies in Vp, and all ...

  26. [34]

    lies in Rp), attaining total fees of kp

    Finally, if 0 ≤ p ≤ 1 2(k−1) , then any allocation that allocates exactly k of the k + 1 transactions is both valid and maximizes fees ( i.e. lies in Rp), attaining total fees of kp. As the allocation that allocates t1, . . . , tk lies in this set as well, and the surplus atta...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.