# Paper Proposes Leader-and-Backup Procurement for Blockchains

Published 2026-10-21 · Crypto.world Team · Crypto.world

> A paper on blockchain procurement finds that a randomly selected primary prover with a backup committee can be optimal in many modeled settings. It also reports that staking’s modeled effect varies with the assumed honest share and failure penalty.

> “Specifically, in many regimes the optimum designates a single, random node as the primary worker and a committee as a fallback, which is reminiscent of leader-based consensus mechanisms.”
>
> Maryam Bahrani, Michael Neuder and S. Matthew Weinberg, authors of the paper

A revised paper from Maryam Bahrani, Michael Neuder and S. Matthew Weinberg sets out a mechanism for paying blockchain participants to complete expensive work, such as generating a cryptographic proof, when some potential suppliers may act against the protocol. ArXiv records the latest version on October 20, updating a paper first submitted on May 7. The authors frame the problem as a trade-off: pay enough to secure the work, but avoid spending too much when multiple providers respond.

## One primary prover, with backups on call

The paper models a protocol that needs one valid proof from a pool of pseudonymous provers. Each prover faces the same normalized cost of producing it; at least a specified number are assumed honest, while the remaining participants may behave adversarially. The protocol must account for both a liveness failure, when nobody delivers, and overprocurement, when it pays for more proofs than needed.

In the proposed designated mechanism, the protocol selects one prover at random and requests a proof from that primary worker. It also selects a committee of backups, each asked to work with a set probability. The protocol can adjust the committee size and backup probability to the assumed number of honest provers and the economic penalty for missing a proof.

The authors find that this design is optimal in many parameter regimes and, across the model, comes within one unit of proof-production cost of the optimal mechanism. That guarantee is a theoretical result, not evidence that a deployed prover market has adopted the mechanism. Its appeal is the familiar structure: one assigned worker, with redundancy added to limit the chance that the task goes unfinished.

## Why missed proofs change the calculation

The model measures the liveness penalty as *C* times the cost of producing one proof. When that penalty rises, the protocol has stronger reason to pay for redundancy; when adversarial control rises, securing a delivery also becomes harder. For a designated mechanism, the authors show that worst-case loss scales as O(log(*C*)/τ), where τ is the honest share of the prover pool.

The logarithmic relationship means the modeled loss grows more slowly than the missed-proof penalty itself, while the honest share affects the scale of that loss. The paper’s result concerns its stated assumptions: one task, one round, a common proving cost and a known lower bound on the number of honest suppliers. It does not establish the same guarantee for a live market with variable task sizes, provider costs or deadlines.

## Staking can reduce losses, but the effect varies

The paper also analyzes negative payments, the mechanism’s term for slashing collateral when participants fail to meet expected behavior. In this model, confiscated stake can compensate the protocol for damage from a liveness fault. The authors describe staking as especially useful when the potential failure penalty is high, though the gains depend on how large that penalty is relative to available collateral.

One example uses 100 provers, including 33 adversarial participants, for an honest share of two-thirds; with a liveness penalty of 10,000 proof-cost units and aggregate stake of 500 units, the paper reports a 47.5% reduction in modeled loss. A second example uses 200 provers, half adversarial, a penalty of 20 and aggregate stake of one unit; it reports a 4.5% reduction. The paper’s table states that these percentage reductions are relative to the symmetric mechanism’s cost with no stake. They are parameterized examples, not measurements from an operating blockchain.

## The paper’s boundary is a single round

The October 20 revision presents procurement as a mechanism-design problem for systems that outsource work but can cheaply verify its result. The authors point to zk-EVM proving as a central motivation and also discuss how the model may inform other forms of permissionless delegation. Their analysis does not cover repeated rounds or multiple simultaneous tasks, where protocols could adapt assignments after seeing which work remains incomplete.

The paper also leaves open whether the best implementable mechanism is always either a designated primary-and-backup scheme or a symmetric committee. The authors state that their analysis does not settle that question, even as they recommend the designated form under the assumptions they consider. For now, the paper’s sharpest practical prescription is also its narrowest: one costly proof, one procurement round, and backups chosen to balance redundant work against the cost of failure.

## Sources

1. Original source · Preprint [Adversarial procurement in blockchains](https://arxiv.org/html/2605.05559v2) arXiv:2605.05559

## Related stories

- [China Targets a National Blockchain Network by 2030](https://crypto.world/news/china-national-blockchain-network-2030-target)
- [Netanyahu Requests Blockchain Plan for Possible Banking Sanctions](https://crypto.world/news/netanyahu-blockchain-plan-banking-sanctions)
- [PQ-Vault Proposes Quantum-Safe Ethereum Custody Without L1 Changes](https://crypto.world/news/pq-vault-quantum-safe-ethereum-custody)
- [SEACAST Delivers Ethereum’s 1 MiB Payload to 99% in 0.64 Seconds](https://crypto.world/news/seacast-ethereum-broadcast-design-simulation)
- [Ethereum Delays Glamsterdam’s Hoodi Upgrade Over Contract Risks](https://crypto.world/news/ethereum-delays-glamsterdam-hoodi-upgrade)

## Explore

[More articles](https://crypto.world/news)

---

Canonical: https://crypto.world/news/paper-leader-backup-blockchain-procurement
