Analytics

Pool discovery scope

The scan covers 0 pools from 53,587 pairs the factory has created, walking the most recently created window first. It is a window rather than a complete ranking, because the endpoint throttles bulk reads. Pairs holding less than 0.0001 ETH are left out as untradeable.

Historical coverage

Volume and liquidity are stored as hourly buckets over exact clock hours, 168 for 7D and 720 for 30D. A window missing even one hour has its total withheld rather than shown smaller — an hour with no trades and an hour that was never ingested produce the same sum, and only one of them is a fact. The volume series is still bootstrapping. The 24-hour figures measure a rolling block range instead, which is a different window and is never added to the hourly totals.

ETH denomination

Volume, fees and liquidity are measured on the ETH side of each pool. That is a measurement rather than a valuation, and it is why token-to-token pools carry -- in those columns: pricing them would need a rate for a third asset Poolix cannot verify. Such pools are still listed and still counted for transactions.

USD pricing

Current TVL prices scanned ETH liquidity at ETH / USD, read fresh each refresh and rejected if the round is stale or malformed. There is no historical USD series: the feed publishes no usable history on this chain, so historical liquidity stays ETH-denominated rather than being priced at today’s rate, which would misstate every past hour.

Token universe

10 of up to 10 unique non-WETH tokens from the scanned pools are tracked. The sweep is still widening, so this is the set found so far rather than the deepest tokens on the chain. It is not every token on Robinhood Chain.

Holders

Every ERC-20 Transfer of the tracked tokens is replayed to find each address that could hold one, then each token contract is asked for balanceOf and only the addresses it reports a positive balance for are counted. The contract decides, not the event log — some tokens emit Transfers that move no state, and counting events alone reported them as holders. Currently 0 of 0 tokens are confirmed, so the count reads -- until every one is. Contracts count as holders; the zero address does not. An address that holds a token without ever appearing in one of its Transfer events cannot be found this way, so this is a lower bound.

24-hour activity

Transactions counts distinct transaction hashes that emitted a Swap, Mint or Burn, so one transaction touching several pairs counts once. Active addresses counts the accounts that sent them. Both cover every pair including token-to-token, because counting a transaction does not require valuing it — and both describe the contracts, not Poolix.

Per-pool figures

The pool table is built from 0 pools with their own indexed history, summed from each pool’s own Swap events and reconstructed from its own Sync events, so no pool borrows another’s activity. Pool APR uses the same formula and the same 0.30% rate as the chain-wide figure above.

Unattributed swaps

A swap is only counted once Poolix knows which side of its pair holds ETH. A pair first seen part-way through an hour leaves that hour’s swaps recorded as unattributed rather than counted, so the 24-hour volume is a floor, not a ceiling. No swaps are in that state right now. They are not discarded and not guessed at; the count is published beside the total so the gap is visible rather than absorbed into it.

No invented comparisons

No card shows a percentage change. Poolix keeps no previous-period snapshot to compare against, so a delta would be fabricated — and a fabricated trend is read as fact more readily than a fabricated level.