A duplicate join key doubles the reported total

A fictional analyst joins transaction rows to a customer-classification table. The supplied classification data contains two records for one customer key, even though the analyst intended one current classification per customer. After the join, that customer’s transactions appear twice, and a revenue aggregation rises without any new sale. A colleague proposes accepting the higher result because the query ran without a syntax error. Another constructs a tiny example and checks the expected relationship and row counts. This written case accesses no real records and changes no dataframe. It asks why valid syntax can produce an invalid business total and what key/relationship evidence is needed before choosing a justified correction.