Consensus: Proof-of-Work vs Proof-of-Stake
Thousands of strangers, no boss — how do they agree on one truth?
Builder · 9 min · Updated 2026-11-08
💡 Two kinds of lottery
In Proof-of-Work, you buy lottery tickets with electricity — the more computing power you burn, the more tickets you hold. In Proof-of-Stake, your tickets are a security deposit you lock up; cheat and the deposit is confiscated ("slashed"). Either way, cheating is designed to cost more than it could earn.
Why consensus is needed
In a network with no central server, nodes must agree on which transactions happened and in what order. A consensus mechanism decides who proposes the next block and how everyone agrees it is final.
Side-by-side
| Proof-of-Work (Bitcoin) | Proof-of-Stake (Ethereum since Sept 2022) | |
|---|---|---|
| Who proposes blocks? | Miner who first finds a valid hash | Validator chosen pseudo-randomly, weighted by stake |
| Cost of participation | Hardware + electricity | Locked capital (32 ETH per validator minimum) |
| Punishment for cheating | Wasted electricity, orphaned blocks | Slashing — part of the stake is destroyed |
| Energy use | High | Very low (Ethereum cut energy use by ~99.95% at the Merge) |
| Finality | Probabilistic — more blocks = more certain | Economic finality after checkpoints (~13 min on Ethereum) |
| Main criticism | Energy consumption, mining centralisation | Wealth concentration, complexity, staking-provider centralisation |
The 51% attack
If one party controls a majority of hash power (PoW) or stake (PoS), it could censor transactions or reorganise recent blocks. On large networks this is extremely expensive; on small chains it has happened repeatedly. This is one reason small, new chains carry extra risk.
🧮 Practical example: cost of attack thinking
Key takeaways
- Consensus decides who writes the next block.
- PoW spends energy; PoS locks capital and slashes cheaters.
- 51% attacks are a bigger risk on small chains.
- Both have trade-offs — neither is "perfect".
Education only — not financial advice.