Security & Resilience

What could go wrong, and what stops it

Bitcoin is often described as secure without saying what that means. This hub collects the pages that say it precisely: what a miner can do with the block it finds and what it cannot, how a double spend is attempted and why confirmation depth is the defence, what a majority of hash rate would actually buy an attacker, how a single node can be isolated from the network, and what happens when the network is split or the hash rate falls. The recurring theme is that the protocol's protection is a cost imposed on the attacker rather than a wall that cannot be climbed.

Security writing about Bitcoin tends to run to one of two extremes. Either the network is unbreakable, or it is one clever attack away from collapse. Neither is useful. The honest position is that Bitcoin replaces a trusted operator with a set of rules and an economic cost, and that both can be described exactly. A miner that produces an invalid block loses the reward it spent electricity to earn. An attacker that wants to rewrite a confirmed transaction has to outrun the entire honest hash rate for as long as the rewrite takes. A node that is isolated from its peers can be deceived, which is why the number of confirmations a merchant waits for is a risk decision rather than a protocol constant.

Each page in this section names the attack, states the assumption the attacker needs, and says what the defence costs the defender. Where a figure is an estimate rather than a measurement — the cost of mounting a majority attack, the size of the security budget in future epochs — the page says so rather than presenting it as a settled number. Where a calculation would genuinely help, the writing points to a tool that does that work properly rather than embedding a calculator in the reference material.

What a Miner Can Do

Can the people who find blocks break the rules?

A miner is the only participant who gets to choose what goes into a block, so the natural question is what that power is worth. The answer is narrower than it first appears. A miner can order the transactions it includes, it can leave transactions out, and it can pay the block reward to an address of its choosing — but it cannot create coins, spend someone else's output, or make an invalid block acceptable. Every other node checks the block against the same consensus rules and discards it if it fails. This strand works through that boundary: what the block template actually controls, why a double spend is a race against confirmation rather than a rewrite of history, and what a majority of hash rate would and would not let an attacker do.

Start with Can Miners Cheat?

Attacks on the Node

What happens when a node cannot see the honest network?

Not every attack targets the chain. Some target the individual node, and the most direct of them is to surround it with peers the attacker controls so that it sees only the attacker's view of the network. An eclipsed node can be fed a stale or fabricated chain, and a merchant relying on that node alone can be shown a payment that the rest of the network never accepted. This strand covers the isolation attack and the peer-management defences that make it harder, then turns to the two ordinary disruptions that are not attacks at all: a node going offline, and a network that splits into two groups which cannot see each other. In both cases the protocol's recovery is the same — the chain with the most accumulated work wins, and the node that was behind catches up.

Start with Eclipse Attacks

Paying for Security

What funds the hash rate that makes rewriting history expensive?

Bitcoin's resistance to a rewrite is not a property of the software; it is a cost imposed on anyone who tries. That cost is set by the hash rate, and the hash rate is paid for out of the block reward — the subsidy plus the fees collected in each block. This strand follows the money. It sets out how the two components divide the reward today, why the split matters as the subsidy halves on its fixed schedule, and what a fall in hash rate actually does to the network: block times lengthen, the difficulty adjustment follows within a retarget period, and the system returns to its ten-minute target without anyone deciding that it should. The pages here are careful to separate what is measured from what is projected.

Start with The Security Budget

Withholding and Racing

Can a miner profit by breaking the honest publishing rule?

The honest strategy is to publish a block the moment it is found. A family of attacks asks what happens when a miner does not. A selfish miner withholds the blocks it finds and releases them strategically, hoping to orphan honest work and raise its share of the reward above its share of the hash rate. A Finney attacker pre-mines a block and spends the same coins twice before publishing it. A race attacker broadcasts a conflicting payment to outrun a merchant's node. Each of these is a real result about the incentive structure, and each depends on assumptions — about propagation, about how other miners behave, about how many confirmations a merchant waits for — that the pages state rather than bury. The strand also covers the Sybil attack, and why creating many identities does not buy influence over the chain.

Start with Selfish Mining

The Mining Record

What does the measured record say about the miners themselves?

The security argument rests on the hash rate, and the hash rate is a measured quantity with a history. This strand holds the record: how difficulty has moved and the retarget rule behind each adjustment, how efficient the hardware has become and how that efficiency is measured, what is known about the energy sources behind the machines, and how the hash rate is distributed across pools and regions. Every figure on these pages is attributed to a named primary source with a date, and each page states the vintage of the data it shows. Where the published record is an estimate rather than a measurement — the energy mix is the clearest case — the page says so rather than presenting it as settled.

Start with Mining Difficulty History

Every page in this section

The strands above are the way in. This is the full contents of the section, grouped by whether the page describes an attack on the chain or the network's response to being stressed.

Attacks on the Chain

What a participant with resources can attempt against the ledger itself, and the specific rule or cost that stops each attempt from succeeding.

Resilience Under Stress

The disruptions that are not attacks — a node offline, a network split, a falling hash rate — and the recovery the protocol performs without anyone in charge of it.

The Mining Record

The measured record behind the security argument: how difficulty has moved, how efficient the hardware has become, where the energy comes from, and how the hash rate is distributed. Each page names its source and the vintage of its figures.

The reference pages behind the strands

The strands above explain the attacks. These pages hold the underlying record and the methodology, for when you want to check a figure or follow a source.

Where the security argument meets the record

The cost of attacking Bitcoin is set by the reward paid to miners. These pages carry the measured record of that reward and the supply schedule behind it.