Short journal · Original editorial · Observation before explanation
Three similar events tempt someone to announce a rule.
An explanation should not erase the observation that prompted it. Record what happened, under which conditions, and what you still do not know. A plausible interpretation can remain a hypothesis while you decide what would help distinguish it from alternatives.
Describe the small sample and look for a counterexample.
Leave an easy return path where possible. Changing a method should not require pretending the earlier attempt never existed.
Try a question
What would you notice after trying this in one small, real situation?