This is the raw log, not a summary. Each line is one withdrawal from an account funded out of my own wallet, with the transaction ID published so you can check it against the chain yourself.
Nothing is removed once it is entered — including the tests that went badly, and including the figures that later turned out to be wrong. Those carry a correction instead of a quiet edit.
This dataset holds 1 test. That is enough to show what happened in each case, and not enough to say what a bookmaker usually does.
So there is no median below 5 measurements and no distribution below 12. Those parts of this page appear on their own once the log is long enough to carry them.
A distribution needs something to distribute. With 1 of the 12 measurements this chart needs, it would be five bars describing single cases — so it is not drawn. The individual results are in the log below, which is the honest version of the same information.
1 measured so far — no median below 5. One measurement is a measurement, not an average.
A normal account and my own coins. No press account, no pre-loaded balance, no contact with the operator.
An account that deposits and immediately withdraws measures the fraud system, not the payout.
No support ticket first. The clock starts at the click and stops when the coins are spendable.
Every figure carries where it came from — read from the chain, or entered by hand. What was not measured says so.
The whole ledger sits behind one fixed address as JSON, regenerated whenever a test is entered. Use it, quote it, check the arithmetic. Attribution appreciated, permission not required.
null means not recorded — never zero. Times are UTC. Corrections travel with the entry they belong to. Bank references are held back; everything on a public chain is there in full.