arXiv Science⌕ Search

arXiv · 2610.24697

Blockchain Participation Games with Upfront Contribution Commitment

Abstract

Blockchain participation games model costly entry, random selection, and threshold success; their retraction variants charge completion costs only to selected contributors. We study committee tasks requiring operators to reserve computing or communication capacity before assignment, so preparation costs are incurred even when capacity remains unused. Users choose abstention, bare registration without preparation, or binding commitment to contribute if selected. Success requires sufficiently many selected contributors, creates public value, and may trigger separate rewards for registration and verified contribution. We characterize all pure Nash equilibria under homogeneous selection and analyze ordering under heterogeneous selection. The results distinguish free-riding on public value from registration that earns rewards without productive commitment; more frequently selected users can choose bare registration while less frequently selected users contribute. Our main result characterizes every homogeneous minimum-outlay scheme supporting full contribution, where every user commits, as a Nash equilibrium. When positive rewards are required, these schemes form a line segment. There is always an optimum with no registration reward. It eliminates every equilibrium containing bare registrants at minimum outlay while retaining the full-contribution equilibrium; threshold-coordination failures may remain. Under heterogeneous selection, a two-user instance has a unique optimum using both rewards, cheaper than every zero-registration-reward scheme supporting full contribution. Finally, if full contribution is a positive-welfare social optimum, the homogeneous zero-registration-reward design attains price of stability one. Under arbitrary independent selection probabilities, every full-contribution Nash equilibrium is also a strong equilibrium.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Zhenhe Zhang, Jing Chen. 2026-10-17. Blockchain Participation Games with Upfront Contribution Commitment. https://arxiv.org/abs/2610.24697

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

Sealed-bid Auctions on Blockchain with Timed Commitment Outsourcing

Sealed-bid auctions play a crucial role in blockchain ecosystems. Previous works introduced viable blockchain sealed-bid auction protocols, leveraging timed commitments for bid encryption. However, a crucial challenge remains unresolved in these works: Who should bear the cost of decrypting these timed commitments? This work introduces a timed commitment outsourcing market as a solution to this challenge. We first introduce an aggregation scheme for timed commitments, which combines all bidders' timed commitments into one while ensuring security and correctness and allowing a varying number of bidders. Next, we remodel the utility of auctioneers and timed commitment solvers, developing a new timed commitment competition mechanism and combining it with the sealed-bid auction to form a two-sided market. The protocol includes bid commitment collection, timed commitment solving, and payment. Through game-theoretical analysis, we prove that our protocol satisfies Dominant Strategy Incentive Compatibility (DSIC) for bidders and Bayesian Incentive Compatibility (BIC) for solvers, and achieves optimal revenue for the auctioneer among a large class of mechanisms. Finally, in the forced-opening-only model in which bidders cannot self-open, we prove that no mechanism can achieve positive expected revenue for the auctioneer while satisfying DSIC and Individual Rationality (IR) for both bidders and solvers, together with ex-post IR for the auctioneer.

cs.GT↗

Improved Weighted Maximin Share Guarantees for Indivisible Chores

We consider the problem of assigning indivisible chores to agents with different entitlements in the maximin share (\MMS) context. While constant-\MMS\ allocations are guaranteed to exist for both goods and chores in the symmetric setting, the situation becomes much more complex when agents have different entitlements. For the allocation of indivisible goods, it has been proven that an $n$-\WMMS\ (weighted \MMS) guarantee is the best one can hope for. For indivisible chores, however, it was recently proved that an $O(\log n)$-\WMMS\ assignment is guaranteed to exist. In this work, we improve this upper bound to a constant-\WMMS\ guarantee. We prove the existence of a $\tfrac{54}{5}$-$\WMMS$ assignment using a combinatorial layered moving-knife algorithm.\footnote{For concurrent and independent work by Li, Wang, and Xing~\cite{li2025constant}, see the \hyperref[remark:parallel-work]{remark on parallel work} in the introduction.} For every fixed $ε>0$, this yields a $(\tfrac{54}{5}+ε)$-$\WMMS$ assignment in polynomial time. Using a faster estimate of the weighted maximin shares instead gives a $\tfrac{108}{5}\cdot (1+ε)$-$\WMMS$ assignment in almost linear time. We also develop a chore-oblivious analysis that guarantees\linebreak $\WMMS$ approximation factors of approximately $2.06447$, $2.475855$, and $2.732051$ for three, four, and five agents, respectively.

cs.GT↗

Envy Cycle Elimination with Strategic Agents: Best Responses and Fairness Guarantees

With strong evidence in the literature showing that fairness and truthfulness are incompatible, there is a recent line of work focusing on the fairness properties of equilibria of simple fair division mechanisms, especially Round-Robin. We consider the Envy Cycle Elimination (E-C-E) procedure of Lipton et al. EC 2024, one of the most versatile tools in fair division. While this simple and intuitive algorithm achieves allocations that are envy-free up to one item (EF1) for any number of agents and general monotone valuation functions, surprisingly little is known about its behavior when agents act strategically. We demonstrate how the presence of incentives, although highly natural and relevant for the majority of applications, completely removes the intuitive clarity in the algorithm's execution, even for a few agents and very simple valuation functions. Additionally, while in the standard algorithmic setting there is great flexibility in how the details of E-C-E are implemented, here additional specifications are needed before the procedure is clearly defined, the choice of which has a potentially huge impact on the agents' behavior. Despite these obstacles, for various natural versions of E-C-E, we give the first results on the existence of Pure Nash Equilibria of the resulting mechanisms, and show there exist versions where fairness guarantees are approximately preserved for agents who play best responses\newx{, as long as there are at most three agents; for four or more agents, no such guarantee can hold for every best response, even at a pure Nash equilibrium. Our guarantees extend to monotone subadditive valuation functions. Moreover, best-responding additive agents always secure a constant fraction of their maximin share in the variants where the selected agent chooses her item.

cs.GT↗