Start here · Lesson 10 of 10

Reading finfluencer picks

When someone online shares a 'pick,' ask one question: is it a scored record? A scored record states, in advance and with a date, exactly what it claims and exactly when that claim will be judged — then judges it automatically, keeping the misses.

Most picks are not that. An undated 'level to watch' can never be evaluated, because there is no moment it comes due and no defined outcome. A wall of green account screenshots is not evidence either — it shows selected wins, not the full record of every call, and it can be staged.

The honest test is boring and public: every forecast dated and logged, scored against what happened, calibration checked, the whole ledger verifiable. That is the bar we hold ourselves to, and a fair one to hold anyone else to.

See also: calibrationWhether our probabilities match reality — of all the times we said an outcome was, say, three-in-five likely, did about three in five actually happen?, resolutionThe scheduled day a forecast is graded against what actually happened — automatically, with no edits., point-in-timeAs it was known on that date — no hindsight; membership and data reflect what was true when the forecast was made.