Spotting noise in whale data
Market makers, exchange plumbing, mispriced tokens and late data: how to tell a real pattern from an artifact.
Common sources of noise
Market makers and bots move huge volume back and forth. We remove wallets whose trading is far out of proportion to their holdings, but some slip through for a while.
Exchange plumbing: exchanges shuffle coins between their own wallets. Known exchange addresses are excluded, but a new, unlabeled one can look like a whale.
Pricing quirks: thinly traded tokens can be mispriced. Implausible prices are rejected and single rows over $1 billion are left out of totals, but small tokens still deserve suspicion.
Late and revised data
Moves can arrive late or be revised, and chain reorganizations happen. A total for “the last hour” can change a little after the hour ends.
A quick checklist
Is the number driven by one wallet? Is the token thinly traded? Does the wallet hold anything, or only pass coins through? Does the raw transaction on Etherscan match? If any answer is shaky, say so in your conclusion.
Try it on a live page
- Read the Known limits section and list the three limits you think matter most.
- Go to the live feed and find one move you’d treat with suspicion. Which checklist question flags it?
- Write one sentence describing that move honestly, including the doubt.
Check yourself
1. A wallet trades $500M a day but holds $20K. It most likely is…
2. Why are single rows over $1 billion left out of totals?
3. A one-hour total can change after the hour ends because…
Go deeper
How to Spot Fake Whale Activity — Wash Trading, Bot Swaps, and Misleading Alerts →Crypto Liquidity Explained — Why One Whale Trade Can Move an Entire Market → Glossary: mev →Describes past on-chain activity and how our data is built. Not investment advice. Method details: data methodology.
