Random signals and parameterization

During my internship at IBM, I was working on a process of verification of hardware implementations by injecting signals into the design under test. Here the verification engineer has to be smart enough to guess the potential breakdown points of the hardware and inject those signals. To further test for robustness, random signals are injected. For another level of robustness, partialy random signals - biased towards the "guessed" potential breakdown points.

The whole idea when applied to life, seems to make so much sense that I was startled when I first thought about it. To test other persons' reaction, dont we drop hints suggesting things that we are not sure of the reaction from the other person?

This same thing, in a different setting sounds so much like a pre-product launch market survey. "Have a look at the need to be catered to(hardware) - Guess the ways which the market may respond in the way we want it to(guess the potential breakdown signals) - Give the market some products that hover around the previous guesses, see how the market reacts(Random signals biased towards the guessed potential breakdown points) - All this is done before the final product (chip) is launched into the market"

Fascinating, how concepts learnt in one sphere of life can be used in other spheres as well.