Many people believe that an answer with 99% confidence is better than one with only 70% confidence. On the surface, with no additional information, maybe this could make sense. But the world of trade-offs almost always makes the 70% solution better.
Trade-offs occur in every decision. Moving any decision from 70% confidence to 99% requires time and complexity. Time is a non-renewable resource that we can never get back. Delaying a decision to increase confidence can often cost a lot of time and only yield false levels of new confidence. Complexity is similar, the more complexity involved in a decision, the more likely an error exists somewhere in the assumptions.
One of the great lessons I’ve learned in my career is that 70% confidence is often enough to move forward with. Get the next 10/20/30% confidence from real life experience and feedback. If you spend time modeling and trying to get everything perfect for release, key opportunities will pass you by.
The best engineers understand this rule. Poor engineers will strive for 100% and take the time to try and get there. This wastes the time it takes and the additional confidence is often false because of the new complexity. But they now have a thick document to fall back on and defend full of assumptions that they can use to justify any change outside of their expectations.