A shared ledger, not a single database
Most records you deal with every day live in one place. A bank keeps its own ledger, a shop keeps its own stock list, and if you want to know what the ledger says, you ask whoever owns it. A blockchain turns that arrangement around. The same record is copied across thousands of independent computers, called nodes, and each one checks new entries against a common set of rules before accepting them.
Because no single party holds the master copy, nobody can quietly rewrite history on their own. To change the record, you would have to convince the rest of the network to accept your version, and the rules are designed to make that extremely costly.
What goes inside a block
Transactions are not added one at a time. They are gathered into batches called blocks. A typical block contains:
- a list of transactions, such as "address A sends 0.5 coins to address B";
- a timestamp showing roughly when the block was produced;
- a reference to the block that came before it;
- extra data the network uses to prove the block was produced according to the rules.
Each new block points back to the previous one, so the blocks form a single line stretching back to the very first block. That line is the "chain".
Hashes: the glue between blocks
The link between blocks is made with a cryptographic hash. A hash function takes any input and produces a short, fixed-length fingerprint. Change even one character of the input and the fingerprint changes completely and unpredictably.
Every block stores the hash of the block before it. If someone tried to edit an old transaction, the hash of that block would change, which would break the link stored in the next block, which would break the one after that, and so on. Tampering is therefore easy to spot: honest nodes simply reject a chain whose fingerprints do not line up.
Consensus: agreeing on the next block
If everyone can propose blocks, the network needs a way to agree which one comes next. This process is called consensus. The two most widely used approaches are:
| Method | How a block producer is chosen | What makes cheating expensive |
|---|---|---|
| Proof of work | Miners race to solve a computational puzzle; the first valid answer wins. | Rewriting history needs more computing power than the honest majority. |
| Proof of stake | Validators lock up coins as collateral and are selected to propose and confirm blocks. | Misbehaving validators can lose part of their locked coins. |
Both methods aim for the same result: a single, agreed history that is costly to attack.
Confirmations and finality
When your transaction is included in a block, it has one confirmation. Each further block added on top counts as another. The deeper a transaction sits, the harder it becomes to reverse, because an attacker would have to rebuild every block after it. This is why exchanges and wallets often wait for a set number of confirmations before crediting a deposit. Some networks also offer a formal point of finality, after which a block is treated as permanent by protocol rules.
Public addresses and private keys
On a blockchain you are identified by an address, a string derived from a public key. To move coins from that address you sign the transaction with the matching private key. The network checks the signature without ever seeing the key itself. Lose the key and nobody, including the network, can move those coins for you. Our guide to wallet security basics covers how to protect it.
What a blockchain does not do
A blockchain makes sure that recorded data follows the network's rules. It does not make an asset valuable, stop prices from swinging, or protect you from sending funds to the wrong address. Transactions are usually irreversible, so accuracy matters. When you are ready to see how trading sits on top of all this, read how a spot order works or explore prices on the markets page.
Crypto assets are highly volatile and you may lose all the capital you invest.
Practise without risk to real funds
The practice account trades live prices with virtual USDT.