Elysium Headcount
Loading snapshot…
Elysium testnet

— holders on paper. — look like people.

Holder count is the easiest number to fake on a launchpad. This register takes every Elysium testnet token with 100+ holders, removes wallets that move in lockstep, and shows what is left.

— look organic — are in balance clusters or repeat wallets

— contract holders, such as pools, are left out of both sides.

—tokens where under 10% of holders look organic
—tokens held by the same wallet set
—highest organic share

Wallet families

Tokens whose holder lists overlap so much that they are effectively held by one group of wallets. Two tokens join a family when the wallets they share make up at least half of both lists combined.

Loading families…

Every token

Select a row to see the signals behind its score.

TokenPatternReportedLook organicOrganic shareFamily24h volume
Loading tokens…

How it is measured

Everything comes from the public Elysium testnet explorer API. Nothing is inferred from names, socials or off-chain data.

Balance clusters clone
Wallets holding almost the same amount, within a few percent of each other, in groups large enough to be a pattern (at least 15 wallets and 3% of holders). Organic buyers rarely land on identical balances.
Repeat wallets cross
Wallets that sit in a balance cluster in three or more different tokens.
Ladder buys ladder
Share of pool buys inside runs of five or more consecutive buys that differ by less than 0.2%. This is what a script spending a fixed amount per buy leaves behind.
Burst burst
Share of pool buys that land in the busiest 6-hour window.
Buy only buyonly
Share of wallets that received the token from a contract, a pool or a router, and never sent it on within the analysed transfers.
Top wallet
Share of the supply held by the single largest wallet, contracts excluded. Tokens where one wallet holds 20% or more get a warning tag, whatever their organic share.
Task-farm wallets farm
From a sample of 15 clustered wallets, the share with 300 or more transactions. High values point to wallets that run long scripted task lists across many contracts.
Wallet family family
Tokens whose holder sets have a Jaccard overlap of 0.5 or more, with at least 50 shared wallets.
Look organic
Wallet holders that are neither in a balance cluster nor a repeat wallet. Organic share is this number divided by all wallet holders. Contracts, such as pools, are left out of both.
Pattern label
Names the dominant pattern. 50%+ organic is organic-leaning, 15 to 50% is mixed. Below 15%, the label says why: a shared wallet set, task-farm wallets, repeat wallets or uniform balances.

What this does not claim

  • Flags describe on-chain patterns. They do not identify people or prove intent, and one operator can run wallets that pass every check.
  • Testnet price, volume and market cap fields are sparse and can be manipulated.
  • Transfer-based signals use up to the 2,000 most recent transfers per token. Holder-based signals use the full holder list.
  • A new snapshot runs every day. Each token's latest numbers and daily history are public at /data/tokens/<address>.json, with a list of all tokens at /data/index.json.