What the whitepaper actually says
Nine pages, published in October 2008. It is more readable than its reputation, and several things people attribute to it are not in it.
On 31 October 2008, a paper titled Bitcoin: A Peer-to-Peer Electronic Cash System was posted to a cryptography mailing list under the name Satoshi Nakamoto. The first block of the resulting network was produced on 3 January 2009.
The problem it states
The opening paragraphs are about payments, not about money supply or inflation. The stated problem is that internet commerce depends on financial institutions to process payments, that this makes small transactions costly, and that the possibility of reversal requires merchants to collect more information about customers than they otherwise would.
The words most associated with Bitcoin today — digital gold, store of value, inflation hedge — do not appear in it. That is not an argument against those ideas; it is a statement about what the document contains. The store-of-value case developed later, in response to how the system was actually used. See Bitcoin as an asset.
The mechanism, in its own order
- A coin is defined as a chain of digital signatures, each owner signing the transfer to the next.
- To prevent double-spending, transactions are publicly announced and participants must agree on a single history of the order they arrived in.
- That history is built as a chain of blocks, each containing a proof of computational work and a reference to the block before it.
- Nodes always treat the longest chain — the one with the greatest accumulated work — as correct.
- An attacker wanting to change a past transaction must redo the work of that block and every block after it, while competing with the rest of the network.
The honest section
Section 11 contains an explicit calculation of the probability that an attacker with a given share of the network's computing power catches up from a given number of blocks behind. The paper's own conclusion is probabilistic: the chance of a successful rewrite falls exponentially with the number of confirmations, without ever reaching zero.
It does not claim transactions are irreversible. It claims they become computationally impractical to reverse, and it shows the arithmetic. It also assumes honest nodes collectively control more computing power than any attacker, and says so plainly as an assumption rather than a guarantee.
What it does not address
- Any mechanism for setting or changing the rules later. Governance is absent entirely, which is why later disputes had no defined process. See the scaling debate.
- What happens when the block subsidy runs down and fees must pay for security. Acknowledged in passing, not solved.
- Mining becoming an industrial activity concentrated in large operations, which is what happened. See mining and difficulty.
- Price, market capitalisation, or investment merit, in any form.
Why it is worth reading yourself
It is nine pages, written in plain technical English, and it is free. Almost everything written about Bitcoin since is either an elaboration of it or a departure from it, and you cannot tell which is which without knowing what it says.