Most traders cannot tell skill from luck, because they are judging a sample far too small to say.
Caliber computes about sixty statistics from your trades and then does the thing no other journal does: it tells you honestly whether your record is long enough for any of them to mean anything yet.
Behind this page: one edge, dealt again 55 times. Same win rate, same average trade, same everything — only the order changed. The bright line is the lucky one. Telling it apart from skill is the whole job.
How many trades you need before a result stops being explainable by luck — corrected for the shape of your returns, not assumed normal.
Metrics stay locked until the record can support them. A Sharpe ratio from twelve trades is not a small number, it is a wrong one.
How big a thing is and how sure we are it is real are two different questions. Everywhere else they arrive as one number.