Analysis of the Blockchain Protocol in Asynchronous Networks

The five-minute read

One action, walked through

1. State the delay the proof allows.
2. Show how many blocks sit outside the unsafe tail.
3. Do not treat a six-block rule from one deployment as the theorem.
4. How does the confirmation depth depend on that delay?

The argument, unpacked

What has to be true

• This does not price bitcoin.
• It does not cover a stake chain.
• An implementation with a different delay has a different bound.

What happened after the paper

Later backbone papers keep the same vocabulary: prefix, quality, growth, and the delay.

What to check before you use the idea

• What network delay is assumed?
• How does the confirmation depth depend on that delay?
• Is honest majority still required?

Terms

Asynchronous

Here: the proof's timing model is weaker than lockstep rounds.

Confirmation depth

How many blocks you wait before treating a prefix as stable.

The problem the paper names

The backbone proofs that assume a synchronous round do not automatically survive a network where messages can be late.

What the design proposes

• Synchrony is a model, not a property of the internet.
• Confirmation depth is the knob the proof turns.
• Honesty of hash power is still assumed.

How the mechanism is specified

• State the delay the proof allows.
• Show how many blocks sit outside the unsafe tail.
• Do not treat a six-block rule from one deployment as the theorem.

What this page does not treat as proven

• This does not price bitcoin.
• It does not cover a stake chain.
• An implementation with a different delay has a different bound.

Why a venture studio still reads it

A confirmation number copied from a wallet is not this analysis. Ask what delay it assumes.