There is steady pressure in risk management to put a number on everything — a probability, an expected loss, a score. Quantification has real value. It also has limits that are easy to forget once the number is on the page.
False precision
A single point estimate placed on a deeply uncertain risk carries an air of confidence that the underlying knowledge does not support. The figure gets quoted, rolled up, and acted on as though it were measured, when it was really a guess dressed for a meeting.
Structure before number
Often the more useful question is not “how likely, how bad” but “what depends on this, and what breaks when it breaks?” You can frequently know the structure — the dependencies, the paths, the single points of failure — with far more confidence than you can know the probabilities. And the structure is what tells you where to look.
Numbers in their place
None of this is an argument against measurement. Quantify where you can defend the number, and lean on it there. For everything else, model the structure honestly and reason about it — rather than reaching for a precision you do not have.