Every figure here comes straight from my own PulseChain node. No aggregators, no third-party feeds. No logs, no cookies, no accounts.
Consensus-layer income is what actually landed. Withdrawal events, swept to each validator's withdrawal address, read off the chain. Exited validators are excluded. Their withdrawals are returned stake, not income. Rewards earned but not yet swept are estimated at lookup time as balance above the 32,000,000 PLS effective stake. Sweeps run roughly every 8 hours per validator, so single-validator figures step up. They don't glide.
Execution-layer income is priority fees paid to the block proposer. Base fees are burned by the protocol. They are never counted. Sets larger than ~400 validators estimate the 24h CL APR from a sample. Sets over 1000 validators extrapolate all figures from a capped sample. The page says so when it applies.
Sometimes a block pays a priority fee of ≥ 100,000 PLS. That income is real and stays in the reward totals. But one whale payment annualised looks absurd, so they're excluded from APR. The whale block count is shown so you can see when a figure is affected.
28D totals include every priority fee in the window, including whale blocks. APR excludes them, so a 28D income figure can be far larger than its APR implies. A whale payment early in the window can dominate the whole figure.
24h means trailing 24 hours. 28d means trailing 28 days. The 7-day sparkline samples 21 points at 8-hour intervals, matching sweep cadence. Each point is trailing-24-hour income annualised. No smoothing, no interpolation. Combined APR is CL plus EL, shown only when both cover the same validators.
Some addresses get blocks from far more validators than are mapped to them. That's a pool sharing one fee recipient. Income and APR would then describe different sets of validators, so APR is suppressed and a badge says why. The income figures stand.
Seats held in the current committee, out of 512. Small sets holding a couple of seats is normal. Occupancy moves around between periods.
State queries are cached for seconds to minutes to keep the node alive under load. Figures can sit slightly behind chain head. Blocks and withdrawals are indexed continuously and may trail by a short interval.
← New lookup · No logs. No cookies. No accounts.