Treat the opening structure as a starting hypothesis, not a contract. If new evidence changes the problem, say so and update the structure.

A visible pivot can actually demonstrate judgment because you are showing that evidence changes your priorities rather than mechanically following a framework.

Briefly state what changed, which branch is now less relevant, and what new branch or question becomes more important.

Keep the revision short and move back into analysis. You do not need to rebuild every bucket from scratch unless the case has fundamentally changed.

I used to overthink this too. What helped was deciding what I needed to prove next instead of trying to make the whole answer perfect.