Taking risks is good in testing/design, but not a great idea for a production rocket. You want to find all the failure points during the design phase instead of blowing up expensive payloads.