Similar to its close cousin, geology; no one can exactly predict the exact behavior of soil especially in chaotic geological condition. Unless we are talking about monolithic engineering fill, geotechnic requires raw information to be interpreted into a shrewd manner when executing analyses.

For example, when designing for slope stability.

Most of the time engineer will just go with data acquired from site investigation especially water table, cohesive and friction angle to analyse the stability of a slope. This is inefficient and insufficient! The right way of doing slope stability analysis requires the more than a single stage of possibilities. Conditions and astute understanding of predictive precursor(s) play important roles in making good judgment or prediction.

In this scenario, conditions of a slope stability should be undermined with the in-depth understanding of geology when it comes to composites and movement including creep, and soil taxonomy for superficial disposition. There is more to it than just parametric derived or sampled from the site. Soil taxonomy tell a story about the soil historical formation and composition beyond parametric and the possible behavior under certain stresses or thresholds. By understanding these non-parametric properties, one can make good if not close assumption about possible behavior of the slope under duress when soil is at rest, passive or active.

These are the things that most academicians have failed to instill in the mind of undergraduates. It is never about fancy Greek alphabets or a set of fucked up formula to calculate; it is all about making good prediction with your imagination.