Why counting wallets overstates how distributed a token is

A project advertised fifty thousand holders as proof of decentralisation. A clustering analysis grouped the addresses by shared funding sources and coordinated movement and found that a few hundred real entities controlled the supply, with one cluster of two thousand wallets tracing back to a single deployer. The holder count was technically accurate and analytically meaningless. Creating a wallet costs nothing, so the number counted addresses, not people.

What everyone sees

A buyer reads holder count as a headcount of independent owners and reasons that fifty thousand holders cannot be manipulated the way ten can. A large number signals a broad, resilient base. The metric feels like a democratic tally — one address, one owner — and the token inherits the safety of a crowd, because a crowd is hard to coordinate and hard to dump all at once.

What is actually happening

On-chain address counts do not map to distinct entities. Chainalysis and academic clustering work show that a single actor routinely splits holdings across thousands of addresses to fabricate the appearance of distribution — a Sybil pattern. The true measure requires entity resolution: grouping addresses by common control. Once clustered, a token boasting fifty thousand holders often shows Gini coefficients near those of the most concentrated assets, with a handful of entities able to move the market alone.

Why it stays hidden

The hidden mechanism is the substitution of a costless proxy for a costly truth. Because generating addresses is free, the holder count measures the willingness to create wallets, not the breadth of ownership. The metric is chosen precisely because it is easy to inflate and hard for a casual reader to deflate, so the concentration hides behind a large, reassuring, meaningless number.