What a study is

One venue, one window. The study rebuilds every wallet's completed trades from the venue's own trade tape, tests which wallets did better than chance would explain, and describes what the survivors had in common at the moment they bought. A paper-only shadow then follows that behaviour forward, so the description is checked against what happened next and not only against the past.

The method on an invented case

Take 4,000 wallets screened and one with 25 closed trades and a handsome average. Out of 4,000 tries, some wallets look skilled by luck alone, so that average has to clear a bar set for 4,000 tries, not for one. If the same wallet created the token or bought in the launch's own block, its trips are set aside before anything is inferred about it. What remains is described by what anybody could see at entry: the token's age, how many wallets had bought, and how long such trades were held.

An illustration of the method. It is not a result and no wallet is meant by it.

What you can do now

• Create an account, so the full wallet table of a study is open to you once one is published and your access allows it. A signed-out visitor sees a short sample.
• Ask for an invite if you do not have a code yet.
• Watch the live market, the tape the studies are read from, open to anybody.