3 ms·

Well, either something like that happened... or she won a lottery with a 10 billion to 1 chance. My money is on the former. I've read enough stories about peopl

by jquery 4y ago

Well, either something like that happened... or she won a lottery with a 10 billion to 1 chance. My money is on the former. I've read enough stories about people assuming false identities to understand there's a number of ways to pull something like that off.

amluto 4y ago
Whoa, watch out for a very common statistical error. For any particular property with a 10 billion to 1 chance, the probability that it applies to someone living is quite high. If you have a group of properties, each with an independent 10 billion to 1 chance, the probability that someone living has one of these properties is very high indeed. The fallacy of thinking that winning a lottery is rare is common and has horrible effects like sending innocent people to jail on a regularly basis. For example, if you search Clearview AI for someone who is a 99.99% match for a surveillance picture of someone committing a crime, you should expect tens of thousands of matches. If the AI really did its job, it would return many hits along with a prominent warning that, with very high probability, any given one of these people did not commit a crime! And, of course, someone always wins the lottery, since that’s the whole point.
- lovemenot 4y ago
  Similar example leading to false imprisonment: https://en.m.wikipedia.org/wiki/Sally_Clark https://en.m.wikipedia.org/wiki/Sally_Clark >> Sir Roy Meadow, who testified that the chance of two children from an affluent family suffering SIDS was 1 in 73 million. He had arrived at this figure by squaring his estimate of a chance of 1 in 8500 of an individual SIDS death in similar circumstances. The Royal Statistical Society later issued a statement arguing that there was no statistical basis for Meadow's claim, and expressed concern at the "misuse of statistics in the courts".[3]
- jquery 4y ago
  Why are you talking to me like I'm ignorant about statistics? I'm not the guy sending innocent people to jail on a 99.99% match. Give me a plausible and ordinary way to cheat the lotto and I guarantee you there will be a lot of cheats. But I don't know of any plausible way to cheat the lotto. Also, yes, the probability that at least one person on earth was the 10 billion to 1 chance is high, but that doesn't mean the chances of her being that person is high. There's simply too much "fog of war" in her past, such as the "daughter" dying of tuberculosis at a convenient time.

by jquery 4y ago

amluto 4y ago

Whoa, watch out for a very common statistical error. For any particular property with a 10 billion to 1 chance, the probability that it applies to someone living is quite high. If you have a group of properties, each with an independent 10 billion to 1 chance, the probability that someone living has one of these properties is very high indeed. The fallacy of thinking that winning a lottery is rare is common and has horrible effects like sending innocent people to jail on a regularly basis. For example, if you search Clearview AI for someone who is a 99.99% match for a surveillance picture of someone committing a crime, you should expect tens of thousands of matches. If the AI really did its job, it would return many hits along with a prominent warning that, with very high probability, any given one of these people did not commit a crime! And, of course, someone always wins the lottery, since that’s the whole point.

lovemenot 4y ago
Similar example leading to false imprisonment: https://en.m.wikipedia.org/wiki/Sally_Clark https://en.m.wikipedia.org/wiki/Sally_Clark >> Sir Roy Meadow, who testified that the chance of two children from an affluent family suffering SIDS was 1 in 73 million. He had arrived at this figure by squaring his estimate of a chance of 1 in 8500 of an individual SIDS death in similar circumstances. The Royal Statistical Society later issued a statement arguing that there was no statistical basis for Meadow's claim, and expressed concern at the "misuse of statistics in the courts".[3]
jquery 4y ago
Why are you talking to me like I'm ignorant about statistics? I'm not the guy sending innocent people to jail on a 99.99% match. Give me a plausible and ordinary way to cheat the lotto and I guarantee you there will be a lot of cheats. But I don't know of any plausible way to cheat the lotto. Also, yes, the probability that at least one person on earth was the 10 billion to 1 chance is high, but that doesn't mean the chances of her being that person is high. There's simply too much "fog of war" in her past, such as the "daughter" dying of tuberculosis at a convenient time.