Proof of Stake (PoS) is the energy-efficient alternative to proof-of-work. Instead of miners competing with computing power, validators put up their own money as collateral. If you haven’t read blockchain-basics yet, start there.
The Idea in One Line
Put your money where your mouth is. If you validate honestly, you earn rewards. If you cheat, you lose your deposit.
The Friend Group Version
Your friend group maintains the shared expense spreadsheet, but instead of solving Sudoku puzzles (that was PoW), you use a deposit system.
Anyone who wants the right to add entries puts ₹1000 into a shared fund. The group randomly picks one depositor to add the next entry. If they add it honestly, they earn ₹50 as a fee. If they try to write fake entries, they lose their entire ₹1000 deposit.
Think about it — why would you risk losing ₹1000 to fake a ₹200 expense? The math doesn’t work in the cheater’s favor.
flowchart TD A["Want to be a validator?<br/>Deposit ₹1000 as stake 💰"] --> B["Network randomly picks<br/>one validator"] B --> C["You're selected!<br/>Validate the next block"] C --> D{"Did you validate<br/>honestly?"} D -->|"Yes ✓"| E["Earn fees 💸<br/>Keep your deposit"] D -->|"No ✗"| F["Lose part of your<br/>deposit 🔥 (slashing)"]
How It Actually Works
Step 1: Become a validator
To participate, you lock up (stake) a minimum amount of the cryptocurrency. For Ethereum, that’s 32 ETH (~$100k+ depending on price). This stake is your collateral — your skin in the game.
You also need to run a validator node (a computer that stays online and connected to the network).
Step 2: Get selected
The network selects a validator to propose the next block. Selection is generally weighted by stake size — if you staked more, you have a higher chance of being picked. But there’s randomness involved too, so it’s not purely “richest wins.”
Different PoS systems handle selection differently:
- Ethereum — committee-based, validators are randomly assigned to committees that attest to blocks
- Cosmos — weighted random based on stake
- Some systems use coin-age (how long you’ve been staking) as a factor
Step 3: Propose and attest
The selected validator proposes a block. Other validators then attest (vote) that the block is valid. Once enough attestations come in, the block is finalized.
Think of it like: Ravi proposes the spreadsheet entry, and then Priya, Aman, and Sneha each sign off on it. Once a majority agrees, it’s locked in.
sequenceDiagram participant Net as Network participant V1 as Validator (Ravi) participant V2 as Attester (Priya) participant V3 as Attester (Aman) participant V4 as Attester (Sneha) participant BC as Blockchain Net->>V1: You're selected to propose! V1->>V2: Here's my proposed block V1->>V3: Here's my proposed block V1->>V4: Here's my proposed block V2->>BC: I attest: looks valid ✓ V3->>BC: I attest: looks valid ✓ V4->>BC: I attest: looks valid ✓ Note over BC: 3/3 attestations received BC->>BC: Block finalized ✅ BC->>V1: Reward: transaction fees 💰 BC->>V2: Reward: attestation fee BC->>V3: Reward: attestation fee BC->>V4: Reward: attestation fee
Step 4: Rewards and penalties
Rewards:
- Block proposer earns the biggest share (transaction fees + protocol rewards)
- Attesters earn smaller rewards for participating
- All rewards come from transaction fees and newly issued coins
Penalties (slashing):
- Propose an invalid block? Lose some of your stake
- Validate conflicting blocks (trying to double-spend)? Lose a LOT of your stake
- Go offline for too long? Gradual small penalties (your node needs to be reliable)
Why Is It Secure?
The security comes from economic incentives. Let’s do the math:
Say you stake 32 ETH (~4,000/year** in rewards.
If you try to cheat, you get slashed. Depending on the violation, you could lose anywhere from 1/32 of your stake to your entire 32 ETH. So you’d risk $100,000 to maybe gain… what exactly? The reward for cheating is almost always less than what you’d lose.
And if many validators collude (coordinated attack), the slashing penalty increases dramatically. Ethereum’s slashing mechanism is designed so that if 1/3 of validators act maliciously at the same time, they ALL lose their entire stake.
flowchart LR subgraph honest["Play Honest ✅"] H1["Stake: $100k"] --> H2["Earn ~$4k/year"] end subgraph cheat["Try to Cheat ❌"] C1["Stake: $100k"] --> C2["Lose $3k - $100k<br/>+ get kicked out"] end
The Centralization Problem
This is the biggest criticism of PoS. Here’s how it plays out:
- Rich person stakes 1000 ETH
- They get selected more often (bigger stake = higher chance)
- They earn more rewards
- They stake those rewards too
- Now they have 1050 ETH and get selected even more often
- Repeat forever
The rich get richer. In PoW, at least you need to keep buying hardware and paying for electricity — there’s ongoing cost. In PoS, your money just… makes more money.
Mitigations that exist:
- Minimum stake thresholds — anyone with 32 ETH can participate (not just whales)
- Liquid staking (like Lido, Rocket Pool) — lets people pool smaller amounts together to reach the 32 ETH minimum
- Randomized selection — so it’s not purely “biggest stake wins”
- Quadratic voting in some systems — diminishing returns for larger stakes
These help, but don’t fully solve the problem. It’s an active area of research.
PoS vs PoW — The Tradeoffs
| PoW | PoS | |
|---|---|---|
| Energy | Massive (mining rigs 24/7) | Minimal (just a regular computer) |
| Entry cost | ASICs (15k) + electricity | 32 ETH (~$100k for Ethereum) |
| Attack cost | Need 51% of hashpower | Need 51% of staked coins |
| Rich get richer? | Somewhat (mining farms) | More directly (stake compounds) |
| Battle tested | 15+ years (Bitcoin) | ~3 years at scale (Ethereum) |
| Speed | ~10 min blocks (Bitcoin) | ~12 sec blocks (Ethereum) |
The Merge (Ethereum’s Switch)
Ethereum ran on PoW from 2015 to September 2022. “The Merge” switched it to PoS, and it was a massive deal:
- Energy consumption dropped ~99.95%
- Block times went from ~13 seconds to a consistent ~12 seconds
- No mining hardware needed anymore
- The network went from ~900 validators at launch to 900,000+ today
It was one of the largest infrastructure upgrades in crypto history, done without any downtime.
Staking as a User
You don’t need to run a validator to benefit from staking. There are three levels:
1. Solo staking (run your own validator)
- Need 32 ETH + a dedicated computer
- Full rewards, full control
- You’re responsible for uptime
2. Staking services (Lido, Rocket Pool, etc.)
- Stake any amount (even 0.01 ETH)
- They run the validators for you
- Take a ~10% commission on rewards
- You get a liquid token back (stETH, rETH) that you can trade or use in DeFi
3. Exchange staking (Coinbase, Binance, etc.)
- Easiest option, just click “stake”
- Higher commissions (~25%)
- Not really decentralized (the exchange controls the validators)
flowchart TD A["Want to stake?"] --> B{"How much ETH?"} B -->|"32+ ETH"| C["Solo staking 🖥️<br/>Full rewards, full control<br/>Run your own node"] B -->|"Any amount"| D{"Want to stay<br/>decentralized?"} D -->|"Yes"| E["Liquid staking 🌊<br/>(Lido, Rocket Pool)<br/>~10% commission"] D -->|"Don't care"| F["Exchange staking 🏦<br/>(Coinbase, Binance)<br/>~25% commission, easy"]
My Take
PoS solves the energy problem, and I think that alone makes it worth it. The centralization concern is real though — I’d love to see more innovation in this space. Maybe there are consensus mechanisms that give better guarantees against wealth concentration. The space is still young.
For now, if you’re interested in staking, liquid staking is probably the best balance of accessibility, rewards, and decentralization for most people.