How blocks are found

A block is a batch of transactions plus a header. The header names the previous block, so the batches form one chain. Nodes accept a new block only if it follows the consensus rules and its proof of work is valid.

Proof of work

Miners build a candidate block from transactions they choose, then hash the header with SHA-256, twice. They change a nonce and try again. The block is allowed only when the hash is numerically below a target. There is no shortcut: each guess is independent.

That work is what makes a rewrite expensive. To replace a buried block, a miner has to redo its work and the work of every block after it, faster than the rest of the network is extending the chain everyone else is on.

Ten minutes is an average

The target is set so that a block is found about every 600 seconds across the whole network. The gap between two blocks is a Poisson process. Some arrive in a minute. Some take half an hour. Ten minutes is the mean, not an appointment.

Difficulty

Every 2,016 blocks the network looks at how long those blocks actually took. Two weeks is the aim (2,016 × 10 minutes). If they came faster, the target gets harder. If they came slower, it gets easier. The change is capped at a factor of four in either direction.

Hashrate is not a dial the chain reads. It is an estimate, worked out from the difficulty and the recent gaps between blocks. A node checks blocks. It does not need a hashrate figure to do that.