In activities with high outcome variance, short-run results carry very little information about ability. Trading sits at the extreme end of that scale: a year is a small sample, and thousands of participants guarantee that some will post excellent records through chance alone.
Three tests help. Sample size, via standard-error-of-expectancy, establishes whether the result could plausibly be zero. Process compliance asks whether the profitable trades were the planned ones - a winning month of improvised trades is luck even when it pays. And mechanism asks whether there is a reason the edge should exist: who is on the other side, and why do they keep taking it?
Held properly, this is not defeatism. It is the reason to judge yourself on process while results accumulate, and the reason to stay sceptical of anyone whose evidence is a screenshot. See process-over-outcome, and be wary of judging a decision by its result.
Related: sample-size-for-edge, sharpe-inflation, process-over-outcome, track-record-selection