Strategic Block Withholding in Blockchain Consensus

552 concepts166 papers1 proposal

In a standard longest-chain blockchain protocol, miners are expected to publish newly discovered blocks immediately. This real-time broadcast allows the network to converge on a single, shared transaction history. However, game-theoretic analysis reveals that immediate publication is not always the most profitable strategy for a rational actor.

By strategically withholding valid blocks, a malicious or profit-maximizing miner can maintain a private chain fork. This private branch is selectively revealed to the public network to invalidate ("orphan") the blocks discovered by honest miners. To understand why this strategic block withholding works, one must look beyond immediate absolute payouts and examine how consensus rules, network topology, and shifting reward structures alter the mathematical incentives of distributed ledgers.

The Mechanics of Relative Reward Maximization

The fundamental intuition behind strategic block withholding—often generalized as selfish mining—rests on a counterintuitive premise: an attacker is willing to intentionally waste their own computational work to force the rest of the network to waste even more.

When an attacker discovers a block and keeps it secret, they gain a head start on mining the subsequent block. If the honest network eventually discovers a block on the public chain, the attacker can release their private block. This creates a temporary fork or tie. Because honest miners split their computational power trying to extend competing branches, the attacker effectively dampens the honest network's efficiency.

Why does this yield higher profits if the attacker occasionally loses their own withheld blocks in the process? The answer lies in the protocol's dynamic difficulty adjustment. As Roi Bar Zur et al. demonstrate in Efficient MDP Analysis for Selfish-Mining in Blockchains, the difficulty adjustment mechanism shifts a rational participant's objective from maximizing absolute rewards to maximizing their relative share of the total network output.

Because the protocol adjusts difficulty based on the total number of blocks produced over a given timeframe, the intentional orphaning of honest blocks tricks the network into thinking the total hashpower has decreased. The protocol responds by lowering the mining difficulty. As analyzed by Qianlan Bai in A Deep Dive into Blockchain Selfish Mining, this strategy operates as a "long game." The attacker initially sustains operational losses due to discarded blocks, but once the difficulty drops, their fixed hashpower captures a disproportionately larger share of the newly adjusted, easier-to-mine blocks.

Exploiting Network Topology and Latency

Strategic block withholding does not occur in a mathematical vacuum; its profitability is heavily bounded by the physical realities of network propagation and graph topology.

When a tie occurs between a public honest block and a newly released private block, the fraction of the honest network that chooses to mine on top of the attacker's block determines the attack's success. Yang Xiao et al. show in Modeling the Impact of Network Connectivity on Consensus Security of Proof-of-Work Blockchain that an attacker's betweenness centrality within the network graph drastically lowers the hashpower threshold required for profitable exploitation. If an attacker occupies high-traffic peering locations, they can propagate their withheld blocks faster than honest nodes, capturing the majority of subsequent forks.

This interplay between network positioning and consensus manipulation can be weaponized further. Kartik Nayak et al. introduce a broader family of deviations in Stubborn Mining: Generalizing Selfish Mining and Combining with an Eclipse Attack. By combining block withholding with network-level eclipse attacks (which isolate specific nodes), an attacker can reveal only the bare minimum number of blocks required to maintain a tie. This forces isolated honest nodes to inadvertently spend their energy extending the attacker's private branch, creating a symbiotic relationship where the victim has no immediate economic incentive to restore honest connectivity.

Honest Chain:   [Block 1] -> [Block 2 (Honest)] -> [Wasted Power]

Private Chain:  [Block 1] -> [Block 2 (Hidden)] -> [Block 3 (Hidden)]
                                 └─► Released to orphan Honest Block 2

In Proof-of-Stake (PoS) systems, the vulnerability manifests through the lens of time and predictability. Jonah Brown-Cohen et al. establish in Formal Barriers to Longest-Chain Proof-of-Stake Protocols that when block validity depends on recent history, strategic actors can mine on multiple forks simultaneously. Crucially, this multi-fork mining can be structured to be mathematically indistinguishable from natural network latency. Because an outside observer cannot differentiate between a malicious block-withholder and an honest node experiencing propagation lag, the protocol cannot safely deploy slashing penalties without risking the punishment of legitimate users.

The Instability of Fee-Dominated Economies and MEV

The classical security models for block withholding assume a static, predictable reward structure (i.e., a fixed subsidy per block). However, as block subsidies decline and networks transition toward transaction fees and Maximal Extractable Value (MEV), the underlying game theory shifts dramatically.

Stochastic, high-variance reward spikes alter the threshold at which withholding becomes viable. Maryam Bahrani et al. analyze this dynamic in Selfish mining under general stochastic rewards, demonstrating that the presence of occasional high-value MEV spikes lowers the profitable selfish mining threshold to hashrates as low as 10%. When a block contains an extraordinarily high-value transaction fee, the incentive to withhold that block and protect it from being front-run or overridden increases exponentially.

Conversely, fee-dominated environments introduce a unique tension. When transaction fees are high and accrue over time, a miner holding a private block faces a compounding risk: the longer they withhold their block, the higher the probability that the honest network will discover a competing block and claim those lucrative fees.

This environment also gives rise to "undercutting" attacks. As explored by Xiongfei Zhao et al. in Dynamic Transaction Storage Strategies for a Sustainable Blockchain, a miner might choose to ignore the current tip of the public chain not to hide a longer private chain, but to deliberately mine an older block height. By leaving high-fee transactions unconfirmed in the mempool, the undercutting miner incentivizes subsequent miners to build on their new fork rather than the honest tip, destabilizing consensus convergence.

Multi-Attacker Dynamics and Stealth Thresholds

Early literature modeled block withholding as a single monolithic adversary competing against a unified, honest majority. Real-world deployment, however, involves multiple independent, rational actors competing simultaneously.

In a multi-attacker landscape, the clean mathematical boundaries of selfish mining begin to blur. Hanqing Liu et al. model these interactions in On the Strategy and Behavior of Bitcoin Mining with N-attackers, introducing threshold-based block publication strategies. In these scenarios, an attacker maintains a private chain until it reaches a specific length threshold, then strategically publishes blocks to recruit honest miners before rival attackers can override them. Without explicit coordination, independent attackers frequently fall into "blind auctions," inadvertently wasting their private blocks against one another rather than the honest chain, which can paradoxically stabilize the network's overall security.

Furthermore, attackers can fine-tune their withholding behavior to avoid detection entirely. While aggressive block withholding causes visible spikes in the network's orphan rate, Maryam Bahrani proves in Undetectable Selfish Mining that a miner controlling up to roughly 38.2% of the total network hashrate can execute a modified withholding strategy. This strategy remains strictly more profitable than honest mining while remaining statistically indistinguishable from honest behavior to an outside observer, neutralizing defenses that rely on monitoring anomalous fork rates.

Designing Incentive-Compatible Defenses

To neutralize the threat of strategic block withholding, protocol designers must alter the structural rules of consensus so that honest behavior remains the unique Nash equilibrium.

One prominent approach focuses on punishing the concurrent block production that withholding relies upon. Danny Dolev et al. propose a structural solution in Colordag: An Incentive-Compatible Blockchain by organizing blocks into a colored Directed Acyclic Graph (DAG). The protocol enforces a rule where all blocks mined at the same depth within a specific color sub-graph are denied rewards entirely. This symmetric penalty ensures that an attacker attempting to invalidate honest blocks incurs a disproportionately higher loss in relative revenue than the honest majority, rendering withholding economically unviable.

Symmetric Penalty Rule (e.g., Colordag):
Depth H:  [Honest Block]  vs  [Withheld Block (Released)]
               │                      │
               └───────────┬──────────┘
                           ▼
               Both Blocks Denied Rewards

Another avenue of defense involves breaking the information asymmetry that gives hidden blocks their power. Yi-Chun Feng evaluates this mechanism in Modeling and Verification of CKB Consensus Protocol in UPPAAL (S), analyzing a two-step confirmation process. By decoupling transaction proposal from transaction commitment through a mandatory block height interval, the protocol ensures that transactions must be visible to the public network long before they can be finalized. This structural delay prevents miners from leveraging secret, pre-mined blocks to front-run or manipulate transaction ordering, effectively neutralizing the economic drivers behind strategic block withholding.

Key Foundational Work

On Power Splitting Games in Distributed Computation: The Case of Bitcoin Pooled Mining (2015) — contributed 8 concept elements to this direction.
Stubborn Mining: Generalizing Selfish Mining and Combining with an Eclipse Attack (2015) — contributed 8 concept elements to this direction.
On Power Splitting Games in Distributed Computation: The Case of Bitcoin Pooled Mining (2015) — contributed 6 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_strategic_block_withholding_bb9e108e,
  title = {strategic block withholding},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/strategic-block-withholding/rdr_bb9e108e4c4c4c3cae87ba9afa61ef4b},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 1, frozen text):

@misc{lacuna_strategic_block_withholding_bb9e108e_v1,
  title = {strategic block withholding},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/strategic-block-withholding/rdr_bb9e108e4c4c4c3cae87ba9afa61ef4b/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake