I’d rather you shared things that were of broader interest. Autism spectrum folks are only at a concentration of about 5% around here (source), and recent college graduates only about 15% (source, though it depends on how you define “recent”), so we’re below 1% even before considering factors like college major, unemployment, and willingness to move to DC.
This may be an example of something sounding “typical,” even though it really has low probability.
A quick estimate (20% unemployment, 20% economics/statistics/CS, 10% would move to DC, readership of 1000), puts the audience at a few percent of a person. Audience decays exponentially with the specificity of the information.
For economics/statistics/CS, I think you may be using the representativeness heuristic—there’s a lot of variation even among people with slightly different brains. I’m reminded of a radio interview I listened to with a woman who had a photographic memory, who had a job as an antiques dealer or something similarly boring. I thought “man, I would totally use that power for good, rather than.. antiques dealing.” But no, many people will grow up and be extremely… ordinary, even if their brains are slightly different. Anyhow, I think you’ll find that 20% is quite a large increase over the base rate—I was being conservative.
As for unemployed people being willing to move to DC… 10% is a total guess, but it’s vaguely based off of unemployed recent college graduates I know. I’m not sure what normal labor mobility actually looks like, though.
Autism spectrum folks are only at a concentration of about 5% around here (source), and recent college graduates only about 15% (source, though it depends on how you define “recent”), so we’re below 1% even before considering factors like college major, unemployment, and willingness to move to DC.
I’m not sure assuming independence of those variables is justified.
True, but in the absence of any strong information about what changes, treating the correlation as weak is a good assumption. And even if one of the numbers is a factor of 2 off, it doesn’t change my point—it was already a back-of-the-napkin kind of thing.
I’d rather you shared things that were of broader interest. Autism spectrum folks are only at a concentration of about 5% around here (source), and recent college graduates only about 15% (source, though it depends on how you define “recent”), so we’re below 1% even before considering factors like college major, unemployment, and willingness to move to DC.
This may be an example of something sounding “typical,” even though it really has low probability.
The discussion section is overdue for further partitioning.
This was a great thing to share, as would be anything similar that hit a non-negligible percentage of LW users.
A quick estimate (20% unemployment, 20% economics/statistics/CS, 10% would move to DC, readership of 1000), puts the audience at a few percent of a person. Audience decays exponentially with the specificity of the information.
I’d say four upvotes count as evidence that your calculation is wrong. :)
People will upvote pretty much anything :P
Upvoted
I’d guess P(economics/statistics/CS) > 20, and that people’s willingness to move to DC isn’t independent of their unemployment.
For economics/statistics/CS, I think you may be using the representativeness heuristic—there’s a lot of variation even among people with slightly different brains. I’m reminded of a radio interview I listened to with a woman who had a photographic memory, who had a job as an antiques dealer or something similarly boring. I thought “man, I would totally use that power for good, rather than.. antiques dealing.” But no, many people will grow up and be extremely… ordinary, even if their brains are slightly different. Anyhow, I think you’ll find that 20% is quite a large increase over the base rate—I was being conservative.
As for unemployed people being willing to move to DC… 10% is a total guess, but it’s vaguely based off of unemployed recent college graduates I know. I’m not sure what normal labor mobility actually looks like, though.
Isn’t that four people?
I think he’s including 5% on the autism spectrum and 15% having recently graduated college in that estimate.
I’m not sure assuming independence of those variables is justified.
True, but in the absence of any strong information about what changes, treating the correlation as weak is a good assumption. And even if one of the numbers is a factor of 2 off, it doesn’t change my point—it was already a back-of-the-napkin kind of thing.