It really is. You can spin it in any way to your favor. Unless you're paying attention or can actually get the data, you can completely fall into the trap. The example of the risk 'doubling' is a perfect example of that. It makes it sound much more of a concern than it really was...it went from a 0.00014 chance to a 0.00028 chance...it's really an insignificant difference.
You are viewing a single comment's thread from:
This recalls me something. Someone was fighting to reduce a source of uncertainties on a prediction from 10% to 5%. It is nice, in principle, except when the dominant source of uncertainties corresponds to a 50% error.