In any activity with a large random component, short samples are dominated by luck and long samples by skill. Trading sits far toward the luck end over days and weeks, and moves slowly toward skill over hundreds of trades.
The practical consequence is that a good month proves very little and a bad month proves very little. What carries information is whether your executions matched your rules, whether your edges behaved as measured, and whether your worst stretches stayed inside the size you planned for.
A quick test: could the same result have been produced by a random entry with your exit rules? If yes, the result is not evidence about your entries yet. See sample-size and confidence-calibration.
Related: outcome-bias, sample-size, self-attribution-bias, confidence-calibration