What we measured
Every claim here comes from our own data and links to the page that shows the working. Several of these went against what we expected before we ran them; we publish those the same way.
Does the clock matter?Tokens graduating 04–08 UTC doubled 60% of the time. At 16–20 UTC: 45%. The ordering holds in both halves of the sample.
26,549 graduations over 95 days Do calls move price?Median move 60 seconds after a call: +0.1%. At two minutes: -0.1%. A coin flip — the audience does not arrive.
10,325 tracked calls Does trading bigger work?Win rate climbs with position size — but 69% of every position above 5 SOL comes from five wallets running industrial volume. Not something you can copy by sizing up.
401,770 closed positions What a copier actually keepsLeader return and copier return are different numbers. At real sample sizes, 2 of 4 graded wallets pay a copier anything. Switch the delay to see how little speed changes it.
101 wallets graded What the bots really chargePhoton’s flat 0.0035 SOL is confirmed across 16 fee accounts with zero variance. The proportional bots publish terms our census could not confirm per size band — and we say so.
6,938 fee payments matched on-chain Ten minutes after the callEvery tracked call priced every ten seconds for ten minutes, per channel, with the cost of trading it at your size.
live tape, rebuilt daily Numbers on this page are pulled from each finding’s own dataset when the site rebuilds, so they never drift from the pages they summarise.