LibraryConsensus2016Design paperCorpus record
On the Security and Performance of Proof of Work Blockchains
Proof-of-work security. Arthur Gervais, Ghassan O. Karame, Karl Wüst, Vasileios Glykantzis, Hubert Ritzdorf and Srdjan Capkun.
The paper builds a quantitative model of stale blocks, double-spends and selfish mining across proof-of-work parameters.
A reading of the public paper. Not a copy, not a benchmark, and not a claim about any later network.
A parameter change sold as 'more throughput' without a stale-block model has not answered this paper.
The five-minute read
The defect
Block interval, block size and network delay are argued about separately. An attacker experiences them together.
The proposal
The paper builds a quantitative model of stale blocks, double-spends and selfish mining across proof-of-work parameters.
Stale rate is the measured link between size, interval and delay.
A double-spend succeeds more often when the merchant waits less and the attacker sees stalls.
The bound
The 2016 traces are not the current relay network.
One action, walked through
- Fix a block interval and a size.
- Simulate propagation and the adversary.
- Read the stale rate and the double-spend probability off that model, not off a slogan.
- How many confirmations does the model need against a given attacker?
The argument, unpacked
What the paper is for
A parameter change sold as 'more throughput' without a stale-block model has not answered this paper.
What happened after
Compact blocks and relay networks exist partly because stale rate is this expensive. Cite the year of the measurement.
What has to be true
- The 2016 traces are not the current relay network.
- The paper does not set a universal safe confirmation count.
- It is not a valuation.
What happened after the paper
Compact blocks and relay networks exist partly because stale rate is this expensive. Cite the year of the measurement.
What to check before you use the idea
- What stale rate does the size and interval imply?
- How many confirmations does the model need against a given attacker?
- Which year's network delay was used?
Terms
- Stale block
- A valid block that lost the race and does not end up in the chain.
- Double-spend
- A payment that is reversed by a competing chain.
The problem the paper names
Block interval, block size and network delay are argued about separately. An attacker experiences them together.
What the design proposes
- Stale rate is the measured link between size, interval and delay.
- A double-spend succeeds more often when the merchant waits less and the attacker sees stalls.
- The numbers belong to the model and the traces they used.
How the mechanism is specified
- Fix a block interval and a size.
- Simulate propagation and the adversary.
- Read the stale rate and the double-spend probability off that model, not off a slogan.
What this page does not treat as proven
- The 2016 traces are not the current relay network.
- The paper does not set a universal safe confirmation count.
- It is not a valuation.
Why a venture studio still reads it
A parameter change sold as 'more throughput' without a stale-block model has not answered this paper.
This is Blockchain Lab's reading of a public design paper. It is not the paper, not a copy of it, and not an offer of tokens, equity, custody or a partnership. Later network behaviour can diverge from the text. Nothing here is investment, legal or technical advice.
Research status: Design paper. Last reviewed: 1 October 2026. This is a reading of a public paper, not investment, legal or security advice.