Strategies: Do it once (perfectly) vs Do it many times

This strategy continuum exists in biological systems and business systems. I happen to be the overly careful type - I want to gather all of the facts and information, plan ahead. My siblings and other friends I notice seem to be the less bothered type and want to just “do it” without worrying about the analysis part - they just react to whatever they find as the plow forward. In a team or a species we need both strategies, at the right time based upon risk or potential loss, and reward.

Should we spend much time planning and gathering information to get ahead of problems? or should we just move forward out of analysis paralysis and see what happens and what issues we need to fix (with much less planning upfront or iteratively - react to the issues when you see them). It all depends on risk severity, risk probability, and risk tolerance.

The Agile software development methodology attempts to do both. With modifications and data driven thinking, I believe that is possible (but always difficult to balance and entirely dependent upon the project and situation).