2 ms·
It is not a matter of statistics. You can smugly measure water for 10 years and learn nothing about ice. No amount of data on Swiss shootings has any bearing on
by Daniel_Newby 14y ago
It is not a matter of statistics. You can smugly measure water for 10 years and learn nothing about ice. No amount of data on Swiss shootings has any bearing on whether Joe Gun-Controlled will back a gasoline tanker up to a high school basketball stadium.
To predict unknown unknowns, you have to do RAND Corporation-style operations research and hypothetical projections. This has been done, and it turns out that mass killing is easy. Really easy. As in, the 9-11 attackers were a bunch of pikers.
That's why after 9-11 security for prominent targets shifted to a strange combination of bollards and guard patrols. Metal detectors and gun screening do not affect safety much. Intercepting truck size weapons does.
I'm flabbergasted that my comment was voted down and justified by "what matters is the past, not your ridiculous hypotheticals". It reads like the Microsoft buffer overflow manual circa 1998.
- Joeri 14y agoWhat matters is what is likely, and what is likely is largely what is observed. People don't blow up schools with tanker trucks. They could, and if they had the intent to kill they should, but they don't. But that's not my point. My point is that the best source for wise decisions is hands-on experience. Many countries have a better handle on the gun situation than the U.S. What's wrong with learning from their experience?
- MaysonL 14y agoOklahoma City bombings: how many copycats?