[DonkeyKong]

It may technically be a pure math problem but doing the math involves estimates and when you use estimates, there is error. If the math is compelling, you have to go with the math. It is when the math is close that you have to use non-math skills.

In the stated problem, what hands others will call you with cannot be calculated exactly. This automatically makes for error in your estimate.

This is similar to finance. An asset is worth the present value of future cash flows. This sounds great but it is highly sensitive to your assumptions in your discounted cash flow model. A real-estate project or a stock price might be worth X or it might be worth 1.5x because of simple assumption differences (discount rate, revenue forecast, expenditure forecast, tax rate forecast etc...).

This is NOT a pure math problem but like finance, it is certainly based on math. Knowing the math is crucial fundamentals so David is definitely correct to pound on that but even he should insert the proper caveats and not come off like this is ONLY math. Surely he understands the challenges to an all-math approach.