Pith. sign in

REVIEW 3 major objections 3 minor 1 cited by

Why Ethereum Needs Fairness Mechanisms that Do Not Depend on Participants' Altruism

T0 review · 3 major / 3 minor · reviewed 2026-08-02 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read An empirical study of 600,115 Ethereum block proposers concludes that at most 1.36 percent are plausibly altruistic, not the commonly read 9 percent; committee-based fairness mechanisms would need at least 128 members.

desk verdict Solid empirical upper bound on altruistic proposers, but the abstract oversells and the committee-size math outruns the measurements. read the letter →

arxiv 2603.05666 v2 pith:ZK4UZ7AV submitted 2026-03-05 cs.DC

Patrick Spiesberger , Nils Henrik Beyer , Hannes Hartenstein This is my paper
classification cs.DC
keywords Ethereumproposer-builderseparationmaximumextractablevalue(MEV)censorshipresistancefairnessmechanismsinclusionlistsaltruismmeasurementcommitteedesign
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

This paper tries to pin down how many Ethereum block proposers can be counted on to comply with a fairness mechanism even when compliance costs them money. Prior work's common reading of the 91% figure for centralized block construction is that the remaining ~9% are honest/altruistic; the paper argues this is wrong. Over the first four months of 2025 it observed 600,115 proposers and progressively excluded anyone with evidence of profit-seeking behavior — relay-mediated block construction, shared governance with relay users, builder payments, exclusive order flow, or non-standard transaction ordering. The residual is at most 1.36% of proposers (the abstract says 1.55%), an upper bound on altruism. If correct, 16-member fairness committees would have only a ~21% chance of containing an altruist, forcing committees of at least 128 members to reach ~83.5%, which raises serious doubts about the practical viability of altruism-dependent mechanisms.

What carries the argument

The load-bearing instrument is a six-stage exclusion pipeline that turns the unobservable moral trait 'altruism' into an observable negative: a proposer is (potentially) altruistic only if none of the six classifiers flag it. The classifiers remove, in order, proposers whose blocks were mediated by relay-based block construction services; proposers sharing a common reward-recipient (governance) cluster with a relay user or with a contract-account or builder-payment address; proposers including non-public (exclusive-order-flow) transactions; and proposers whose blocks are not strictly ordered by descending priority fees. The final condition is the subtle one: strict fee ordering is the defaul

What would settle it

Offer a permissionless bribe contract on a testnet fork of Ethereum that pays any proposer a premium for excluding a designated transaction from an inclusion-list challenge, then record how many of the 3,814 proposers classified as altruistic accept it. If a significant share accepts, or if those proposers later start using relay-mediated block construction once local construction becomes less profitable, the 1.36% upper bound is not a bound on bribe-resistant altruism.

Watch

Extended reading notes

Core claim

On the authors' own terms, the discovery is that altruistic proposers exist but are rare: 0.71% of observed proposers show no indication of any non-altruistic behavior, and even the most generous reading, the set of 'potentially altruistic' proposers, is 1.36% (the abstract rounds to 1.55%). After classifying all 859,313 blocks in the January–April 2025 window, the paper finds that 98.64% of proposers exhibit observable non-altruistic behavior of one of six kinds. It interprets the residual as an upper bound, not a point estimate, because borderline cases are classified conservatively. The paper's central consequence is that the commonly used 9% altruism estimate is almost an order of magnit

Load-bearing premise

The load-bearing premise is that absence of observable profit-seeking behavior equals willingness to forgo revenue and comply with a fairness mechanism; in particular, the paper counts strict descending priority-fee ordering — itself a profit-maximizing default heuristic of most clients — as the altruistic signal, so the 1.36% figure is an upper bound on detectable non-altruism, not a measure of bribe-resistant altruism.

Editorial extensions

If this is right

  • A 16-member inclusion-list or concurrent-proposer committee would contain at least one altruist with probability about 20.7%, not the 77.9% implied by the 9% reading; mechanisms proposed at that size are unlikely to deliver the promised censorship resistance.
  • Reaching an 83.5% probability of an altruistic member requires committees of at least 128, a size that dilutes individual influence and makes the mechanism harder to coordinate and reason about.
  • Because 98.64% of proposers show observable profit-seeking behavior, any mechanism that makes compliance costly will face near-universal defection unless fairness is made the financially rational strategy via rewards, penalties, or accountability.
  • Under enshrined proposer-builder separation, the rationality properties of proposers do not change: proposers who already chose rational delegation will not suddenly intervene, so the altruism bottleneck persists.
  • The altruistic pool skews toward validators registered before the proof-of-stake transition; as that cohort shrinks, the effective altruism share is likely to fall, so committee sizes may need to grow over time.

Reading between the lines

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

  • My reading: the paper's own Method VI weakens the estimate it claims. Strict descending priority-fee ordering is a profit-maximizing default heuristic of most consensus clients, so a rational proposer running default settings produces exactly the 'altruistic' signature. The 1.36% figure is therefore an upper bound on 'no observed non-altruism,' not on bribe-resistant altruism.
  • If the altruism signal is partly a default-configuration artifact, the true bribe-resistant share could be below 0.71%, and the committee-size calculation would be even more pessimistic.
  • A natural extension the authors leave implicit: use their classifier on a testnet with a live bribe challenge (a transaction that pays out if a certain address is excluded) to measure actual compliance among the 3,814 'altruistic' proposers.
  • Another extension: resample the same six classifiers at future dates; if the altruist share declines with validator turnover, mechanism designers should not treat 1.36% as a stable floor.
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

3 major / 3 minor

Summary. The paper estimates the prevalence of 'altruistic' Ethereum block proposers — proposers who, in the authors' definition, follow the protocol without deviation and could plausibly support committee-based fairness mechanisms such as Inclusion Lists and Multiple Concurrent Proposers. Using four months of Ethereum data (859,313 blocks, 600,115 proposers), the authors apply a cascade of classifiers to exclude proposers who use MEV-Boost relays, share governance with relaying proposers, are contract accounts, receive builder payments, include exclusive order flow, or do not order transactions by descending priority fees. They report that at most 1.36% of proposers (abstract: 1.55%) provide no clear evidence of non-altruistic behavior, and use this rate to argue that a 128-member committee would give only about an 83.5% probability of containing at least one altruistic member. The paper concludes that committee-based fairness mechanisms relying on altruism require substantially larger committees than currently proposed and that incentive- or penalty-based mechanisms are needed.

Significance. If the measurement is valid, the paper makes an important empirical contribution to Ethereum mechanism design: it challenges the commonly cited ~9% altruistic-proposer figure and provides a concrete, reproducible exclusion pipeline. The authors are appropriately careful in several places — they repeatedly frame 1.36% as an upper bound, acknowledge that de-anonymization could reveal cluster overlap, and make analysis scripts publicly available. The study is directly relevant to ongoing IL/MCP committee-size proposals. However, the central quantity the paper needs — the share of proposers who would comply with a fairness mechanism even at personal financial cost — is not directly measured, and one of the positive classifiers (strict priority-fee ordering) is, by the paper's own admission, the profit-maximizing default of most consensus clients. This weakens the quantitative committee-size claim, even though the qualitative conclusion that reliance on altruism is fragile appears robust.

major comments (3)
  1. [§3.3 (Method VI) and §5] Method VI labels strict descending priority-fee ordering as a positive signal of altruism, despite the paper's own statement that this ordering 'aims to maximize proposer profits' and is the default heuristic of most consensus clients. A rational proposer running an unmodified client produces exactly the 0.65% signature counted as altruistic. The 0.71% with non-strict but strongly correlated ordering are 'not clearly altruistic' in Figure 2, yet are included in the 1.36% used in Section 5 to compute 1−0.986^128 ≈ 83.5%. The 128-member figure is therefore an upper-bound probability under a generous definition, not a reliability guarantee; if the true bribe-resistant rate is lower, a 128-member committee may fall well below 83.5%. The qualitative 'larger committees are needed' conclusion survives, but the concrete 128/83.5% claim must be re-derived with a defensible parameter and clearly f
  2. [Abstract vs. body (§3, §4)] The abstract states 'full year 2025' and 'at most 1.55 percent of proposers', while the introduction and results consistently describe a four-month window (January–April 2025) and report 1.36% (0.65% strict altruistic and 0.71% potentially altruistic). These are materially different claims. The paper needs to reconcile the observation period and explain why the abstract's headline number differs from the body's 1.36%. This is not merely a typo, because the committee-size calculation in the abstract and Section 5 depends on the exact rate.
  3. [§4.1–§4.3] Many classifications rest on extremely thin individual evidence: the 47,216 non-relaying proposers have a median of exactly one proposed block, and the 3,814 'altruistic' proposers are classified from strict ordering in 'all proposed blocks' — which for a one-block proposer is a single observation. No sensitivity analysis is provided for the Spearman threshold (σ_rs > 0.7), the tie-handling epsilon, or the decision to exclude only proposers rather than clusters for classifiers V and VI. Since one default-configured block passes the strict-order test, the paper should report the block-count distribution for the 3,814 altruistic proposers and show how the 1.36% rate changes under reasonable threshold variations.
minor comments (3)
  1. [Figure 2] The figure labels the 0.71% share as 'not clearly altruistic' while the text in §4.3 says these proposers are 'also likely altruistic'. The terminology ('altruistic', 'potentially altruistic', 'not clearly altruistic') should be harmonized.
  2. [§4.3] The sentence 'Excluding all clusters containing at least one proposer with detected XOF activity would yield a value of 0.57 %' is presented without derivation; please define the baseline and clarify whether this is a cluster-level or proposer-level number.
  3. [§5, Validity under ePBS] The claim that proposers who previously outsourced block production 'would adopt this deviation' in the future is a behavioral inference, not a measurement. The paper should distinguish more sharply between observed past behavior and predicted future compliance with an incentive-compatible fairness mechanism.

Circularity Check

2 steps flagged · score 5.0 of 10

Headline altruism share is defined as the residual of the exclusion pipeline; Method VI further counts profit-maximizing default ordering as altruism, so the 83.5% committee reliability figure is partly an artifact of the classifier definition.

  1. self definitional [Section 3 Methodology / Section 6 Conclusion]
    "Proposers are classified as altruistic if they neither use MEV-Boost relays nor participate in shared governance with already excluded proposers, and if their blocks yield only protocol-intended rewards, i.e., block rewards and priority fees. ... We find that roughly 98.64 % of proposers behave non-altruistically, while only 1.36 % provide no clear evidence of such behavior."

    The classification defines 'altruistic' as the complement of a set of exclusion criteria, so the reported 1.36% is the residual of the pipeline by construction. The conclusion that '98.64% behave non-altruistically' is then just the complement of the not-excluded set, even though the paper's own Classifier VI explicitly 'does not imply non-altruism.' Thus the headline share is equivalent to the chosen operationalization rather than an independent measure of altruism.

  2. self definitional [Section 3.3, Method VI]
    "Such an ordering aims to maximize proposer profits and reflects a simple heuristic implemented by most consensus clients. Although this ordering is profit-maximizing, we classify the corresponding proposer as altruistic, since the heuristic depends solely on protocol-defined attributes."

    The positive signal for 'altruistic' is defined as the standard profit-maximizing client default (descending priority-fee ordering). A rational proposer using default client settings therefore automatically satisfies the altruism criterion, so the measured 'altruistic' set is populated by construction with profit-maximizing behavior. The later committee-size calculation (1 − 0.986^128 ≈ 83.5%) inherits this definitional inflation, making the reliability estimate an artifact of the classifier rather than an independent measure of bribe resistance.

full rationale

The paper is transparent about its methods and explicitly labels 1.36% as an upper bound, and the self-citations are not load-bearing. However, the central quantity is definitionally constructed: 'altruistic' is operationalized as 'not excluded by the pipeline,' and Method VI specifically counts the profit-maximizing default ordering as altruistic. Consequently, the headline '98.64% behave non-altruistically' and the '128 members / 83.5%' committee probability are restatements of the classifier assumptions rather than independent empirical findings. The paper's own caveats—Classifier VI does not imply non-altruism, and the method may overestimate altruism—confirm that the strong conclusion goes beyond what the measurement can support. This is a partial circularity, not a citation-chain or uniqueness-theorem problem.

Assumptions & free parameters 5 free parameters · 7 assumptions · 0 invented entities

No invented entities. The ledger shows the analysis is a chain of behavioral proxies; each proxy is a domain assumption, and the headline upper bound inherits uncertainty from all of them. The 'altruism' definition is effectively the paper's main free parameter, and the strict-fee-ordering axiom is the most fragile link because it labels a profit-maximizing default as the positive altruism signal.

free parameters (5)
  • Altruism definition (behavioral proxy chain) = no MEV-Boost use; no shared governance with relaying proposer; EOA-only cluster; no builder payments; no XOF; strict fee
    The central classifier chain; changing any step changes the headline number — the paper itself notes strict criteria give 0.65% and cluster-level XOF exclusion gives 0.57% (§4.3).
  • Spearman threshold (classifier VI) = 0.7
    Hand-chosen cutoff borrowed from 'common interpretations' [22] (a medical statistics reference); splits 0.65% 'altruistic' from 0.71% 'potentially altruistic'; not calibrated to Ethereum data, no sensitivity analysis.
  • Monitored relay set = 6 relays (Aestus, Agnostic, bloXroute ×2, Flashbots, Ultrasound, Titan)
    Proposers using unmonitored relays are classified as non-relaying, inflating the altruistic bound; completeness of the relay set is asserted, not verified.
  • Tie-handling epsilon in σ_rs = unspecified 'ε > 0 accounting for ties'
    The strict-ordering definition σ_rs = 1 − ε is not fully operationalized; different tie handling changes the 0.65%/0.71% split.
  • Altruism rate used in committee sizing = 1.4% (the potentially-altruistic share)
    The committee-size recommendation (128) uses the 1.36% upper bound; using the strict-altruistic 0.65% would require ~276 members — a factor-of-two design sensitivity not discussed in §5.
assumptions (7)
  • domain assumption Any use of MEV-Boost relay mediation is non-altruistic behavior (delegation of block contents to a profit-maximizing builder).
    Section 2.1; premise for classifier I, driving 92.13% of the exclusion; a proposer could in principle use relays without exploiting users, but the paper treats any relay use as non-altruistic.
  • domain assumption A proposer sharing a coinbase cluster with a relaying proposer is non-altruistic (guilt by governance association).
    Section 2.2, classifier II; cluster-level exclusion assumes the managing party's behavior transfers to all validators it controls. If a cluster is mixed, genuine altruists are misclassified as non-altruistic.
  • domain assumption All contract-account coinbase clusters are STaaS setups with non-altruistic managing parties.
    Section 3.2, Method III: 'manual inspection indicates that all identified CA clusters belong to STaaS setups' — a universal generalization from an unspecified manual procedure with no inter-rater check.
  • domain assumption Strict descending maxPriorityFeePerGas ordering evidences altruism.
    Section 3.3, Method VI; the paper itself says this ordering 'aims to maximize proposer profits' and is the default heuristic of most consensus clients — so the positive signal for altruism is a profit-maximizing default behavior, conflating rational default configs with altruism.
  • domain assumption Transactions absent from the Mempool Guru dataset are exclusive order flow (XOF) included for profit.
    Section 3.3, Method V; relies on the coverage of four monitoring nodes plus a dataset whose public access is not established; the paper acknowledges 'minor discrepancies' when cross-checked with Etherscan.
  • standard math Committee membership is effectively independent/random for the binomial probability 1 − p^n.
    Introduction and §5; assumes independence of altruism across committee draws; the paper's own clustering shows non-altruism is positively correlated across validators, so the variance of committee altruist counts is underestimated (the mean still follows the marginal rate).
  • standard math Spearman rank correlation is an appropriate content-independence / altruism metric.
    Section 3.3, Method VI, citing Zar [32]; standard statistic, but the mapping from ρ to a binary altruism label is the paper's choice.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Why Ethereum Needs Fairness Mechanisms that Do Not Depend on Participants' Altruism." pith.science (2026). https://pith.science/paper/ZK4UZ7AV

@misc{pith2026260305666,
  author       = {Pith},
  title        = {Pith review of: Why Ethereum Needs Fairness Mechanisms that Do Not Depend on Participants' Altruism},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZK4UZ7AV}},
  note         = {Machine review of arXiv:2603.05666}
}
read the original abstract

Ethereum's ideal of censorship resistance, together with related fairness properties, is undermined in practice, motivating fairness mechanisms that aim to restore these properties. Several of these mechanisms hand control over block contents to a committee of proposers under a 1-of-n honest assumption: at least one committee member complies with the mechanism even when deviating would increase personal revenue. We refer to such proposers as altruistic. Yet prior work shows that roughly 91 percent of blocks are constructed by centralized block-building services that demonstrably take user-adverse actions for financial gain; the responsible proposers sign these blocks blindly, without any means of intervention. A common reading of this figure is that 9 percent of proposers forgo these gains and act altruistically. Our empirical analysis of the full year 2025 shows that this share is far smaller: at most 1.55 percent of proposers can plausibly be regarded as altruistic, whereas the remaining 98.45 percent of proposers exhibit observable non-altruistic behavior. We interpret 1.55 percent as an upper bound on the prevalence of altruistic proposers. These results imply that committee-based fairness mechanisms that rely on altruistic members would require substantially larger committees than currently proposed. This raises concerns about their practical viability and motivates mechanisms in which fair behavior is the rational choice.

Figures

Figures reproduced from arXiv: 2603.05666 by the authors.

Figure 1
Figure 1. Methodology for identifying altruistic proposers: starting from all observed proposers, excluding evident MEV-Boost [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Analysis results – percentages denote shares of the [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Histogram of proposers sharing an EOA coinbase address, referred to as clusters. Grey dots denote clusters in which [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Histogram of proposers classified as potentially [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Accountable Transaction Inclusion Lists: Enhancing Ethereum's Censorship Resistance

    cs.DC 2026-07 reject novelty 6.0 of 10

    FairFIL makes multi-block censorship costly by requiring builders to disclose excluded transactions and forcing the next block to include them, with loss of the whole block reward for omissions.

Reference graph

Works this paper leans on

33 extracted references · 2 canonical work pages · cited by 1 Pith paper

  1. [1]

    Aiyer, Lorenzo Alvisi, Allen Clement, Mike Dahlin, Jean-Philippe Martin, and Carl Porth

    Amitanand S. Aiyer, Lorenzo Alvisi, Allen Clement, Mike Dahlin, Jean-Philippe Martin, and Carl Porth. 2005. BAR fault tolerance for cooperative services. In Proceedings of the twentieth ACM symposium on Operating systems principles (SOSP ’05). Association for Computing Machinery, New York, NY, USA, 45–58. doi:10.1145/1095810.1095816

    arXiv 2005

  2. [2]

    Matthew Weinberg

    Kaya Alpturer and S. Matthew Weinberg. 2024. Optimal RANDAO Manipulation in Ethereum. In6th Conference on Advances in Financial Technologies (AFT 2024), Vol. 316. doi:10.4230/LIPIcs.AFT.2024.10

    doi:10.4230/lipics.aft.2024.10 2024

  3. [3]

    Decentralized Systems Lab at Yale University. 2026. Mempool Guru. https: //mempool.guru//. [Online; accessed on February 16, 2026]

    2026

  4. [4]

    Nils Beyer and Patrick Spiesberger. 2026. Ethereum-Altruistic-Proposers. https: //github.com/kit-dsn/Ethereum-Altruistic-Proposers. [Online, accessed on February 13, 2026]

    2026

  5. [5]

    Vitalik Buterin. 2021. State of research: increasing censorship resistance of transactions under proposer/builder separation. https://notes.ethereum.org/ s3JToeApTx6CKLJt8AbhFQ [Online; accessed on August 14, 2025]

    2021

  6. [6]

    Vitalik Buterin, Eric Conner, Rick Dudley, Matthew Slipper, Ian Norden, and Abdelhamid Bakhta. 2019. EIP-1559: Fee market change for ETH 1.0 chain. https://eips.ethereum.org/EIPS/eip-1559 [Online; accessed on August 15, 2024]

    2019

  7. [7]

    Philip Daian, Steven Goldfeder, Tyler Kell, Yunqi Li, Xueyuan Zhao, Iddo Ben- tov, Lorenz Breidenbach, and Ari Juels. 2020. Flash Boys 2.0: Frontrunning, Transaction Reordering, and Consensus Instability in Decentralized Exchanges. InProceedings of the 2020 IEEE Symposium on Security and Privacy (SP). IEEE, 910–927. doi:10.48550/arXiv.1904.05234

    arXiv doi:10.48550/arxiv.1904.05234 2020

  8. [8]

    Ethereum Foundation. 2024. The Merge. https://ethereum.org/en/roadmap/ merge/. [Online; accessed on August 13, 2025]

    2024

Show all 33 references
  1. [9]

    Ethereum Foundation. 2025. Proof-of-Stake Consensus Specifications. https: //github.com/ethereum/consensus-specs. [Online; accessed on August 12, 2025]

    2025

  2. [10]

    Ethereum Foundation. 2025. Staking as a service. https://ethereum.org/en/ staking/saas/. [Online; accessed on August 14, 2025]

    2025

  3. [11]

    Flashbots. 2022. API specification for MEV-Boost PBS relays. https://flashbots. github.io/relay-specs/. [Online; accessed on August 15, 2025]

    2022

  4. [12]

    Flashbots. 2022. MEV-Boost. https://github.com/flashbots/mev-boost. [Online; accessed on August 25, 2025]

    2022

  5. [13]

    Pai, and Max Resnick

    Elijah Fox, Mallesh M. Pai, and Max Resnick. 2023. Censorship Resistance in On-Chain Auctions. In5th Conference on Advances in Financial Technologies (AFT 2023), Vol. 282. doi:10.4230/LIPIcs.AFT.2023.19

    2023 doi

  6. [14]

    Garimidi, J

    P. Garimidi, J. Neu, and M. Resnick. 2025. Multiple concurrent proposers: Why and how. (2025). arXiv:2509.23984 https://arxiv.org/abs/2509.23984

    2025

  7. [15]

    Vincent Gramlich, Dennis Jelito, and Johannes Sedlmeir. 2024. Maximal ex- tractable value: Current understanding, categorization, and open research ques- tions.Electronic Markets(2024). doi:10.1007/s12525-024-00727-x

    2024 doi

  8. [16]

    Barbara Guidi and Andrea Michienzi. 2025. Linking MEV attacks to further maximise attackers’ gains: evidence from the Ethereum blockchain.Blockchain: Research and Applications(2025). doi:10.1016/j.bcra.2025.100340

    2025

  9. [17]

    Lioba Heimbach, Lucianna Kiffer, Christof Ferreira Torres, and Roger Watten- hofer. 2023. Ethereum’s Proposer-Builder Separation: Promises and Realities. In Proceedings of the 2023 ACM on Internet Measurement Conference. ACM, Montreal QC Canada, 406–420. doi:10.1145/3618257.3624824

    2023

  10. [18]

    Lido Protocol. 2025. Lido – Liquid Staking Protocol. https://lido.fi/. [Online; accessed on November 19, 2025]

    2025

  11. [19]

    2024.Concurrent Block Proposers in Ethereum

    Michael Neuder and Max Resnick. 2024.Concurrent Block Proposers in Ethereum. https://ethresear.ch/t/concurrent-block-proposers-in-ethereum/18777 [Online; accessed on August 24, 2025]

    2024

  12. [20]

    Department of the Treasury

    U.S. Department of the Treasury. 2025. Tornado Cash Delisting. https://home. treasury.gov/news/press-releases/sb0057 [Online; accessed on February 21, 2025]

    2025

  13. [21]

    Danny Ryan. 2022. The Risks of LSD. Ethereum Research Notes. https://notes. ethereum.org/@djrtwo/risks-of-lsd# [Online; accessed on August 14, 2025]

    2022

  14. [22]

    Patrick Schober, Christa Boer, and Lothar Schwarte. 2018. Correlation Coef- ficients: Appropriate Use and Interpretation.Anesthesia & Analgesia(2018). doi:10.1213/ANE.0000000000002864

    2018 doi

  15. [23]

    Patrick Spiesberger, Jan Droll, and Hannes Hartenstein. 2025. A Policy-Based Mitigation of Transaction Exclusion in Ethereum [Work in Progress Paper]. In Proceedings of the 30th ACM Symposium on Access Control Models and Technologies. ACM, Stony Brook NY USA, 127–132. doi:10.1...

    2025

  16. [24]

    2024.EIP-7805: Fork-choice enforced Inclusion Lists (FOCIL)

    Thomas Thiery, Francesco D’Amato, Julian Ma, Barnabé Monnot, Terence Tsao, Jacob Kaufmann, and Jihoon Song. 2024.EIP-7805: Fork-choice enforced Inclusion Lists (FOCIL). https://eips.ethereum.org/EIPS/eip-7805 [Online; accessed on August 24, 2025]

    2024

  17. [25]

    Sarisht Wadhwa, Julian Ma, Thomas Thiery, Barnabe Monnot, Luca Zanolini, Fan Zhang, and Kartik Nayak. 2025. AUCIL: An Inclusion List Design for Rational Parties. Cryptology ePrint Archive. https://eprint.iacr.org/2025/194

    2025

  18. [26]

    Sarisht Wadhwa, Luca Zanolini, Aditya Asgaonkar, Francesco D’Amato, Chengrui Fang, Fan Zhang, and Kartik Nayak. 2024. Data Independent Order Policy Enforcement: Limitations and Solutions. InProceedings of the 2024 on ACM SIGSAC Conference on Computer and Communications Securit...

    2024

  19. [27]

    Anton Wahrstätter. 2024. MEV-Boost Dashboard. https://mevboost.pics/ [Online; accessed on August 13, 2025]

    2024

  20. [28]

    Anton Wahrstätter, Jens Ernstberger, Aviv Yaish, Liyi Zhou, Kaihua Qin, Taro Tsuchiya, Sebastian Steinhorst, Davor Svetinovic, Nicolas Christin, Mikolaj Barczentewicz, and Arthur Gervais. 2024. Blockchain Censorship(WWW ’24). Association for Computing Machinery, New York, NY, ...

    2024

  21. [29]

    Anton Wahrstätter, Liyi Zhou, Kaihua Qin, Davor Svetinovic, and Arthur Gervais

  22. [30]

    Sen Yang, Kartik Nayak, and Fan Zhang. 2025. Decentralization of Ethereum’s Builder Market. In2025 IEEE Symposium on Security and Privacy (SP). 1512–1530. doi:10.1109/SP61157.2025.00157

    2025

  23. [31]

    Sen Yang, Fan Zhang, Ken Huang, Xi Chen, Youwei Yang, and Feng Zhu. 2024. SoK: MEV Countermeasures. InProceedings of the Workshop on Decentralized Finance and Security (DeFi ’24). Association for Computing Machinery, New York, NY, USA. doi:10.1145/3689931.3694911

    2024

  24. [32]

    Jerrold Zar. 2005. Spearman Rank Correlation.Encyclopedia of Biostatistics. Vol. 5(2005). doi:10.1002/0470011815.b2a15150

    2005 doi

  25. [2023]

    arXiv:2305.16468 doi:10.48550/arXiv.2305.16468

    Time to Bribe: Measuring Block Construction Market. arXiv:2305.16468 doi:10.48550/arXiv.2305.16468

    doi arXiv

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.