4 ms·
>The point of noting that 90% figure is not so you can come up with some sort of Bayesian probability of your getting assaulted based on population level statis
by Jiro 5y ago
>The point of noting that 90% figure is not so you can come up with some sort of Bayesian probability of your getting assaulted based on population level statistics.
It's quite common that people have a point when they are presenting statistics, but they have made this point in error. Statistics aren't something you're only supposed to use for the one conclusion someone wants you to draw from them.
>Again, it's not just about the frequency of encounters, it's about the nature of them. On any given day I encounter more strangers than I do friends and family.
You have fewer man-hours of encounters with strangers than you do with people you know, even if the absolute number of the strangers is higher. (And you shouldn't be invoking "friends and family", since that isn't the same as "people you know".)
- ModernMech 5y ago> It's quite common that people have a point when they are presenting statistics, but they have made this point in error. I'm sorry, I'm not sure what you're getting at. > You have fewer man-hours of encounters with strangers than you do with people you know You can only speak for yourself, this is not true for everybody.
- nsxwolf 5y agoI’m trying to imagine a building or office manager using these stats as a defense against his or her failure to protect employees from a trespasser. “That assault happened but it was really unlikely so I should be in the clear” You do not allow a situation like this to occur. If it does, there is a serious problem with your security. Period.