3 ms·
Nitpick: > If you write, read, bookmark, share or search for something like this [0], does your conditional probability of killing someone go up? I would gues
by philh 15y ago
Nitpick:
> If you write, read, bookmark, share or search for something like this [0], does your conditional probability of killing someone go up?
I would guess yes - it seems likely that the fraction of killers who look up things like that is greater than the fraction of not-killers. Even taking into account that some killers will look up stuff like that and hide their tracks.
(Remember that killers, too, may be morbidly curious. Assume they're just as likely as anyone else to be. Then we're really comparing "how many killers looked stuff up specifically to learn to hide their crimes?" versus "how many people who subsequently became killers refrained from looking stuff up out of morbid curiousity, because they didn't want it used as evidence against them in future?")
But not so much higher that it's strong evidence, so your point stands.
- firebones 15y agoIn fact, it is still so low that it is likely not useful evidence at all. Let's say there are 50 serial killers in operation in the US. And the MeFi thread cited had 100,000 uniques. Even if all 50 killers read the thread you would still get a huge false positive rate if you are counting on that thread in the browser history alone telling you anything.