Well, widening / loosening the “margin” or distribution of suitability criteria is indeed one of the valid approaches, but one this is still only part of the equation for the problem AFAICT.
Yes, currently, to my model, that P() really is in that ballpark. I’m currently hitting (with P>.98) way off my current “sweet spot in personspace”, with few hits ever getting closer to it and forming a cloud around a completely different area, so my best WAG pretty much give those numbers when trying to project how many I’d have to meet to expect at least one statistical outlier to hit the margin. Making said sweet spot larger is something that would indeed help a lot, but doing so without reducing the total expected payoff of this whole calculation is also non-trivial, for reasons I hope are obvious.
I strongly suspect that my current noise is in no small part due to my current approaches / general behaviors. There’s at bare flat minimum 1 in 50 people (assuming IQ stats are any indication) with sufficient reasoning ability for me to find them very interesting, of those at least 1 in 3 is using that ability in a way that I probably wouldn’t perceive as noise (so I’d probably notice quickly enough), my preferences / personspace “sweet spot” check would eliminate around (WAG: intuitions from personspace stuff) 80-95% of those remaining.
Which means that, by those numbers and assumptions, around 1 in 750 to 1 in 3000 would be a valid match if I were meeting persons according to a uniform personspace probability distribution and breaking the sufficient-common-knowledge barrier in a proportionally uniform manner over persons met. The clear difference indicates that I’m probably doing something wrong, so the most efficient way I know of solving the problem is to find what I’m doing wrong and fix it first, not just meeting more people.
IMO, 1 in 750 is not a particularly constraining margin, especially if you consider that under ideal circumstances you should do the reverse of what I’m doing and actually be concentrating your hits around your sweet spot, not some other place far away from it.
Also, I dislike the term “lowering your standards”. The imagery puts person on a scale basically equivalent to transforming personspace into a Me.perceivedValue(X) function that outputs the scalar distance between Me.perceivedPSLoc(X) and Me.sweetSpotCenter. It gives exactly zero information about the other components of the equation. It also gives very little information on the measurement unit of the scalar.
Unfortunately, even living in a very student-dense city and deliberately targeting locales near universities doesn’t seem to have quite the effect I was hoping for. Things are not helped by the fact that French, the main language of 2⁄3 of the population here, distinctly lacks key words and concepts that seem necessary for bayesianism. The word “evidence”, for example, has no French equivalents to my knowledge—even the French wikipedia page on Bayes’ Theorem struggles with this.
As I’ve said, I’m most likely doing a lot of things wrong, because even going to places near university campus(es) (which I’d go to anyway, since they’re otherwise still the places I’d prefer going to) gives these results. I’m also going with the assumption that the actual odds for people I am meeting there are much higher than 1⁄50 for the intelligence criterion, but calculating flat minimum ratios for an IQ level I’m certain is high enough seemed like a more appropriate conservative figure.
It’s the way it looks and feels from here too—I seem to be a rare exception in considering reason, logic and knowledge to have any value (besides the obvious monetary value of “knowledge” of things related to a business) among native French speakers here.
Campaigns to “preserve language and culture” and keep forcing children to go to only French schools and study only in French make me cringe constantly.
That’s kinda spurious reasoning. By that standard, people who speak languages where evidentiality is considered so relevant it’s marked grammatically (like Turkish, or Apache, or Yukaghir) should on average be much more rational than people who don’t. Appeal to the Sapir-Whorf Hypothesis is usually a quick ticket to confusion.
I did not mean to imply an appeal to Sapir-Whorf concepts. Disregarding every other factor, I do seem to be a rare exception among native French speakers in this culture. Whether I would also be a rare exception, a less rare exception, or an even more rare exception in some other language or some other culture, is a different matter, which is itself worth examining in its own separate right for its own reasons.
Other than that, I agree that what you’ve said does follow, and to the best of my knowledge isn’t currently supported by any public research and only has sporadic anecdotal evidence.
My objection to teaching only French is that it’s a well-known fact that knowing multiple languages helps immensely with various aspects of cognition and intelligence, and learning multiple languages during childhood has been shown to be an overwhelming net positive. It follows that forcing children to learn only one language has a net negative impact. This fact is perceived, agreed, and then waived by appeal to consequence: “If children learn English, they will only speak English [because all regional neighbors do], so less and less people will speak French, so our culture will die!”
Well, you’ve got to consider complicating factors, which makes this hard to measure. Those other countries aren’t very affluent compared to the USA and their educational system is probably worse, plus they don’t have access to the institutional infrastructure of knowledge like we do. Also, measuring rationality seems hard, etc. there’s tons of problems that always pop up when we try to evaluate things like this.
I mean, I think you’d probably be mostly right and that there’s not much difference in rationality between different language users, but for other reasons than the apparent average rationality of certain language-users.
The word “evidence”, for example, has no French equivalents to my knowledge
It has no terribly good Italian equivalent either, but this doesn’t make it hard to talk in a Bayesian way: you just say stuff like “it’s likely that X, given that Y”. (In particular, ISTM that—among the kind of people usually I hang with at least—“folk probability” resembles Bayesianism much more than frequentism, and most people who use frequentist statistics only give lip service to it without being actually convinced it makes all that sense.)
There’s at bare flat minimum 1 in 50 people (assuming IQ stats are any indication) with sufficient reasoning ability for me to find them very interesting
Maybe part of the problem is you go around saying things like that.
Or, to put it in this particular context: I don’t go around saying things like that. This is a discussion about relationships and attraction, and the things I say here are (or so I perceive them to be) very relevant to the subject at hand. You’ve seen me once say something like that, in a place where saying things like that is both appropriate and productive, and you deduce that I always go around saying things like that to random people I’ve just met before I even know them? I’d be very afraid if I were a suspect in a murder investigation led by you.
You also seem to have misinterpreted just what it is that “that” was saying. To put it in other words that might be less prone to “pompous elitist” pattern-matching, I’m basically saying that there’s a statistical guarantee that I’d be very interested in maintaining an intellectual discourse (and hopefully long-term relationship of some kind, even as acquaintances) with, given enough time to talk with and get to know them, at least one out of every fifty people out there. Even more than that in practice, since there will be many people who are interesting despite not being Mensa material. That sentence just puts a lower boundary on the amount of people I could find very interesting.
I apologize. Even if my comment had had a small probability of being helpful, I should have stated it differently, and I did jump to more conclusions than was warranted.
I didn’t mean to imply a misinterpretation, though. If you did go around saying things like that, the pattern-matching would be the whole problem. If you actually believed something to the effect of “people with IQ less than X are not worth knowing”, that might also be an obstacle, but at a later stage of relationship-forming. In any case, that appears to be irrelevant.
Well, widening / loosening the “margin” or distribution of suitability criteria is indeed one of the valid approaches, but one this is still only part of the equation for the problem AFAICT.
Yes, currently, to my model, that P() really is in that ballpark. I’m currently hitting (with P>.98) way off my current “sweet spot in personspace”, with few hits ever getting closer to it and forming a cloud around a completely different area, so my best WAG pretty much give those numbers when trying to project how many I’d have to meet to expect at least one statistical outlier to hit the margin. Making said sweet spot larger is something that would indeed help a lot, but doing so without reducing the total expected payoff of this whole calculation is also non-trivial, for reasons I hope are obvious.
I strongly suspect that my current noise is in no small part due to my current approaches / general behaviors. There’s at bare flat minimum 1 in 50 people (assuming IQ stats are any indication) with sufficient reasoning ability for me to find them very interesting, of those at least 1 in 3 is using that ability in a way that I probably wouldn’t perceive as noise (so I’d probably notice quickly enough), my preferences / personspace “sweet spot” check would eliminate around (WAG: intuitions from personspace stuff) 80-95% of those remaining.
Which means that, by those numbers and assumptions, around 1 in 750 to 1 in 3000 would be a valid match if I were meeting persons according to a uniform personspace probability distribution and breaking the sufficient-common-knowledge barrier in a proportionally uniform manner over persons met. The clear difference indicates that I’m probably doing something wrong, so the most efficient way I know of solving the problem is to find what I’m doing wrong and fix it first, not just meeting more people.
IMO, 1 in 750 is not a particularly constraining margin, especially if you consider that under ideal circumstances you should do the reverse of what I’m doing and actually be concentrating your hits around your sweet spot, not some other place far away from it.
Also, I dislike the term “lowering your standards”. The imagery puts person on a scale basically equivalent to transforming personspace into a Me.perceivedValue(X) function that outputs the scalar distance between Me.perceivedPSLoc(X) and Me.sweetSpotCenter. It gives exactly zero information about the other components of the equation. It also gives very little information on the measurement unit of the scalar.
1 in 50 people among the whole population has IQ >= 131; in places such as university towns that fraction is likely to be substantially higher.
Yes, it would seem so.
Unfortunately, even living in a very student-dense city and deliberately targeting locales near universities doesn’t seem to have quite the effect I was hoping for. Things are not helped by the fact that French, the main language of 2⁄3 of the population here, distinctly lacks key words and concepts that seem necessary for bayesianism. The word “evidence”, for example, has no French equivalents to my knowledge—even the French wikipedia page on Bayes’ Theorem struggles with this.
As I’ve said, I’m most likely doing a lot of things wrong, because even going to places near university campus(es) (which I’d go to anyway, since they’re otherwise still the places I’d prefer going to) gives these results. I’m also going with the assumption that the actual odds for people I am meeting there are much higher than 1⁄50 for the intelligence criterion, but calculating flat minimum ratios for an IQ level I’m certain is high enough seemed like a more appropriate conservative figure.
Oh, that explains why Quebecois seem to think and behave in such silly ways :). At least it’s the way it looks from the other end of the country.
It’s the way it looks and feels from here too—I seem to be a rare exception in considering reason, logic and knowledge to have any value (besides the obvious monetary value of “knowledge” of things related to a business) among native French speakers here.
Campaigns to “preserve language and culture” and keep forcing children to go to only French schools and study only in French make me cringe constantly.
That’s kinda spurious reasoning. By that standard, people who speak languages where evidentiality is considered so relevant it’s marked grammatically (like Turkish, or Apache, or Yukaghir) should on average be much more rational than people who don’t. Appeal to the Sapir-Whorf Hypothesis is usually a quick ticket to confusion.
I did not mean to imply an appeal to Sapir-Whorf concepts. Disregarding every other factor, I do seem to be a rare exception among native French speakers in this culture. Whether I would also be a rare exception, a less rare exception, or an even more rare exception in some other language or some other culture, is a different matter, which is itself worth examining in its own separate right for its own reasons.
Other than that, I agree that what you’ve said does follow, and to the best of my knowledge isn’t currently supported by any public research and only has sporadic anecdotal evidence.
My objection to teaching only French is that it’s a well-known fact that knowing multiple languages helps immensely with various aspects of cognition and intelligence, and learning multiple languages during childhood has been shown to be an overwhelming net positive. It follows that forcing children to learn only one language has a net negative impact. This fact is perceived, agreed, and then waived by appeal to consequence: “If children learn English, they will only speak English [because all regional neighbors do], so less and less people will speak French, so our culture will die!”
And that’s what really makes me cringe.
Well, you’ve got to consider complicating factors, which makes this hard to measure. Those other countries aren’t very affluent compared to the USA and their educational system is probably worse, plus they don’t have access to the institutional infrastructure of knowledge like we do. Also, measuring rationality seems hard, etc. there’s tons of problems that always pop up when we try to evaluate things like this.
I mean, I think you’d probably be mostly right and that there’s not much difference in rationality between different language users, but for other reasons than the apparent average rationality of certain language-users.
It has no terribly good Italian equivalent either, but this doesn’t make it hard to talk in a Bayesian way: you just say stuff like “it’s likely that X, given that Y”. (In particular, ISTM that—among the kind of people usually I hang with at least—“folk probability” resembles Bayesianism much more than frequentism, and most people who use frequentist statistics only give lip service to it without being actually convinced it makes all that sense.)
Maybe part of the problem is you go around saying things like that.
Maybe I should refer you to this other comment I made on this topic.
Or, to put it in this particular context: I don’t go around saying things like that. This is a discussion about relationships and attraction, and the things I say here are (or so I perceive them to be) very relevant to the subject at hand. You’ve seen me once say something like that, in a place where saying things like that is both appropriate and productive, and you deduce that I always go around saying things like that to random people I’ve just met before I even know them? I’d be very afraid if I were a suspect in a murder investigation led by you.
You also seem to have misinterpreted just what it is that “that” was saying. To put it in other words that might be less prone to “pompous elitist” pattern-matching, I’m basically saying that there’s a statistical guarantee that I’d be very interested in maintaining an intellectual discourse (and hopefully long-term relationship of some kind, even as acquaintances) with, given enough time to talk with and get to know them, at least one out of every fifty people out there. Even more than that in practice, since there will be many people who are interesting despite not being Mensa material. That sentence just puts a lower boundary on the amount of people I could find very interesting.
I apologize. Even if my comment had had a small probability of being helpful, I should have stated it differently, and I did jump to more conclusions than was warranted.
I didn’t mean to imply a misinterpretation, though. If you did go around saying things like that, the pattern-matching would be the whole problem. If you actually believed something to the effect of “people with IQ less than X are not worth knowing”, that might also be an obstacle, but at a later stage of relationship-forming. In any case, that appears to be irrelevant.
You meant to say “despite not being”?
Yes, thanks for catching that. Fixed.