Case context: A bank team is building a model to flag potentially fraudulent card transactions. They want the alerts to be highly accurate, but they also want to catch as many true fraud cases as possible.

Question: What should the team understand about these two goals?

Sample answer: They should understand that improving one of these goals can make the other harder to achieve. In practice, precision and recall often pull in opposite directions, so the team should evaluate them together rather than assume they can both be pushed up independently.

Key points:

• Precision rewards making few false alerts.
• Recall rewards finding most of the real fraud cases.
• These aims often conflict during model tuning.
• The conflict is common, not guaranteed in every situation.

Rubric: A strong response identifies the usual tension between precision and recall, ties that tension to the team's objective, and avoids saying the tradeoff always appears in every model.