What are you trying to figure out?

Ask a real selection question.

The more specific the context — scale, region, stack, constraints — the sharper the answer.