Educational content only · We don't sell crypto or give financial advice

Lesson 3.3 · Module 3: How a blockchain works

Reaching consensus: proof of work and proof of stake

12 min read · Level: Beginner+ · Written by Daniel Reed · Reviewed 20 October 2026

In this lesson you will learn to:

  • Explain why a blockchain needs a consensus method.
  • Compare proof of work and proof of stake.
  • Describe what a fork is and why some are temporary.
Proof of work compared with proof of stake.Proof of workProof of stakeMiners try nonce after nonceuntil hash < targetSecurity comes fromthe cost of computationFollow the chain with most workValidators lock a depositand are chosen to proposeSecurity comes fromdeposits that can be cutBreaking rules → slashing
Proof of work compared with proof of stake.

Who writes the next page?

Thousands of nodes hold the same waiting transactions. If all of them wrote their own next block, there would be thousands of different histories. The network needs a fair way to decide who adds the next block, and a rule for choosing between competing versions. That is what a consensus method does.

Proof of work

In proof of work, used by Bitcoin, participants called miners compete to find a block header whose hash is below a target number. The only way to find one is to try again and again, changing a field called the nonce each time. It typically takes an enormous number of attempts across the whole network.

Finding a valid hash is hard, but checking it is instant: anyone can hash the header once and see that it meets the target. The winner adds the block, and the protocol creates a set number of new units for them as an incentive, plus the transaction fees.

The rule for competing versions is simple: nodes follow the chain with the most accumulated work. To rewrite history, an attacker would need to outpace the combined computing work of everyone else — the origin of the phrase "51% attack". The cost of that work is what protects the record, and it is also why proof of work uses a lot of electricity.

Proof of stake

In proof of stake, participants called validators lock up tokens as a security deposit. The protocol selects validators to propose blocks and others to attest that blocks are valid. Ethereum switched from proof of work to proof of stake in September 2022, an event known as "the Merge", and its energy use fell by more than 99.9%.

Instead of burning electricity, validators put their deposit at risk. If a validator breaks the rules — for example, by signing two conflicting blocks — part of the deposit can be destroyed automatically. This penalty is called slashing. Honest behaviour is encouraged by the prospect of losing the deposit rather than by the cost of computation.

Forks

Sometimes two valid blocks are found at nearly the same moment, and parts of the network briefly follow different versions. This is a fork. It resolves itself as soon as one branch grows longer (or, in proof of stake, receives more attestations); nodes switch to it and the other block is discarded. This is why some services wait for several further blocks before treating a transaction as settled.

Forks can also be deliberate. When participants adopt new rules that old software rejects, the chain can split permanently into two networks. This happened in 2016, when a disagreement after a smart-contract exploit split Ethereum and Ethereum Classic.

Finality

In proof of work, a transaction becomes more certain with every block added after it, but never absolutely final. Many proof-of-stake systems add explicit finality: after a set point, reversing a block would require destroying a large share of all deposits. Different designs make different promises here, and it is worth knowing which kind a system offers.

Key takeaways

  • A consensus method decides who adds the next block and which history counts.
  • Proof of work makes adding blocks costly in computation; nodes follow the chain with the most work.
  • Proof of stake uses locked deposits that can be cut for breaking the rules.
  • Temporary forks resolve themselves; permanent forks happen when rules diverge.

Check yourself

In proof of work, why is checking a block easy but creating one hard?

Finding a header hash below the target takes huge numbers of attempts, but anyone can verify the result by hashing once.

What is slashing?

The automatic destruction of part of a validator's deposit when it breaks the rules.

Why do some services wait for several blocks?

Because short temporary forks can occur; each additional block makes reversal less likely.

Sources

This lesson is for general education about how the technology works. It is not a recommendation to acquire, sell or hold any crypto-asset and is not financial advice.