Of course, when you start defining "safe" in a probabilistic way, you end up very quickly estimating the degree of safety instead of binary safe/not safe.
Correct. I'm sure its has been done, but the general public does not think about it that way.
Nice subtle joke(?) with that link to a Moody's March 2008 report on rating bonds given that rating agencies were probably the closest single thing you could point to as the fraud that caused the mortgage bubble and 2008 collapse. Yes, sometimes a feeling can be closer to the truth than a lot of complicated math done with wrong assumptions.
> general public does not think about it that way.
I agree that general public is utterly incapable of understanding basic concepts of probability. (See: Clinto had probability 52% of winning. Trump won -> Forecast was wrong. How stupid is that?) I just commented when you told you have not seen safe even defined in a probabilistic way.
> Yes, sometimes a feeling can be closer to the truth than a lot of complicated math done with wrong assumptions.
To me that is like a stopped clock is correct twice a day compared to a working clock that is just having wrong time all the time.
At least with the complicated math you are forced to spell out the assumptions and there is a chance you are able to recognize them before it is too late.
https://www.moodys.com/sites/products/DefaultResearch/200710...
Of course, when you start defining "safe" in a probabilistic way, you end up very quickly estimating the degree of safety instead of binary safe/not safe.