Another recommendation in the same vein is How To Measure Anything by Douglas Hubbard. And Superforecasting by Tetlock. Estimation/forecasting (same thing different emphasis) is a problem we know quite a lot about, and we know how to train people to issue calibrated estimations.

The first step to learning anything is figuring out how to get feedback on your performance. For estimation, probabilistic estimations are the way: https://two-wrongs.com/verifiable-software-development-estim...