How Proof of Stake Selects Validators

Milton Altenwerth — Wed, 09 Sep 2026 22:01:03 +0000

Most proof-of-stake networks choose validators with verifiable randomness, giving larger eligible stakes a larger selection chance.

A validator set is the trust boundary behind a Universal Bridge security model. In a Universal Bridge design, selection is only the first question; the next is how those validators agree.

What the protocol needs

It needs an active validator set, effective stake per member, and a shared random seed. Stake is weight in a deterministic function every node can recompute. On Ethereum, RANDAO supplies the seed; it assigns one proposer and committees for each slot.

Eligibility: a deposit or bond enters the active set; activation and exit rules limit churn.
Sampling: the seed maps stake-weighted chances to validator indices. More effective balance improves expected selection frequency, not certainty.
Duty: the proposer builds the block; selected attesters check it and sign votes. Conflicting votes can be slashable.

What changes in practice

Proof of stake replaces specialized mining hardware and energy competition with capital at risk and node operations. It makes rotating participation practical; committees keep every validator from processing every vote. Randomness means the largest holder does not simply win every turn.

Where this model stops

Do not apply this model to every cross-chain verifier. Axelar Network uses delegated proof of stake, while LayerZero Protocol lets each application configure Decentralized Verifier Networks and Wormhole Protocol uses a governed Guardian set with a signature threshold. Those are validator or verifier policies, not one universal PoS lottery. For a small operator, delegation fits a smaller budget; for a larger operator, direct staking buys control but demands reliable hardware, key security, and uptime. The practical verdict: inspect the active set, randomness source, committee rule, and slashing condition before treating “proof of stake” as a security guarantee.

How Staking Rewards Are Calculated: APR vs APY

Milton Altenwerth — Wed, 09 Sep 2026 18:33:19 +0000

Staking rewards are calculated from a protocol’s reward rules and a participant’s stake, then adjusted for participation, validator commission, penalties, and compounding. On Ethereum Mainnet, the base reward uses effective balance and total active balance; on the Polkadot Network, an era’s validator work and commission determine the split. A displayed APY is therefore an estimate of future token flow, not the formula itself.

1. Identify the stake and the network

First, a user identifies both the native asset and the network where the stake will be recorded. A native token is the chain’s own asset, such as DOT on Polkadot or ETH on Ethereum; a wrapped token may represent value from somewhere else without carrying the same staking rights. Proof of stake is the security system in which validators lock capital and perform consensus work, and nominators or delegators assign stake to them without running the node. The first-time operator should therefore check the asset’s chain, the validator’s status, and any minimum or lock-up before looking at a percentage.

A transfer through Orbiter Finance or another bridge can change the network representation without changing the reward rules. Manta Bridge moves assets between networks. The destination protocol, not the bridge, decides whether that representation can be staked.

2. Find the reward period and accounting unit

Next, find the protocol’s accounting period and reward source. On Polkadot, an era is the reward window, normally about 24 hours; validators earn era points for actions such as producing relay-chain blocks and validating parachains. At era end, gross rewards are allocated according to those points, then paid to the validator and nominators. A payout can require a staking transaction, so earned and visible in a wallet are separate events.

On Ethereum, rewards are applied per epoch. The base reward is effective balance × 64 ÷ (4 × √ total active balance). The square-root term means a larger active validator set lowers the per-validator base rate, all else equal.

3. Calculate the gross reward before fees

Then calculate the gross reward before annualizing it. Ethereum adds weighted rewards for timely source, target, and head votes, with extra opportunities for block proposals and sync committees; missed duties reduce income, and slashing can remove stake. A standard Type 1 validator has effective balance capped at 32 ETH, while Type 2 compounding credentials can raise that effective balance up to 2048 ETH. That distinction changes future reward weight, not merely the way the number is displayed.

Polkadot’s split gives a useful worked example: if an era produces 2 DOT and the validator charges 20% commission, the validator keeps 0.4 DOT and 1.6 DOT remains for stake owners. A nominator holding half of that validator’s stake receives 0.8 DOT from the remainder; the validator’s own stake also receives its proportional share. Commission is taken from gross rewards before the stake split. This is why two validators with identical network conditions can produce different net returns.

4. Convert the result into a usable rate

Finally, compare the token amount actually received, not just the advertised rate. The disagreement worth resolving is whether APR or APY is the reward itself. APR is a one-year annualized rate before reinvestment; APY assumes rewards are periodically added back to the stake. APY therefore depends on compounding frequency and on a rate that can change as total active stake, validator performance, commission, and token issuance change.

The checkable figure is the protocol’s credited token balance over a stated period. Subtract validator or pool fees, transaction costs, missed rewards, and any lock-up or exit friction before comparing providers. The practical verdict is simple: choose the validator or pool whose on-chain payout method, commission, and operational record you can inspect, then treat APY as a scenario, not a promise.

DEV Community: Milton Altenwerth