Regardless, if your goals are genuinely instrumental you very much want to figure out what parts of the effect are due to placebo effects and what parts are due to real effects, so you can maximise your beneficial outcomes with a minimum of effort.
There is a word for the problem that results from this way of thinking about instrumental advice. It’s called “akrasia”. ;-)
Again, if you could get people to do things without taking into consideration the various quirks and design flaws of the human brain (from our perspective), then self-help books would be little more than to-do lists.
In general, when I see somebody worrying about placebo effects in instrumental fields affected by motivation, I tend to assume that they are either:
Inhumanly successful and akrasia-free at all their chosen goals, (not bloody likely),
Not actually interested in the goal being discussed, having already solved it to their satisfaction (ala skinny people accusing fat people of lacking willpower), or
Very interested in the goal, but not actually doing anything about it, and thus very much in need of a reason to discount their lack of action by pointing to the lack of “scientifically” validated advice as their excuse for why they’re not doing that much.
I’d prefer not to discuss this at the ad hominem level. You can assume for the sake of argument whichever of those three assumptions you prefer is correct, if it suits you. I’m indifferent to your choice—it makes no difference to my utility. I make no assumptions about why you hold the views you do.
My view is that the rationalist approach is to take it apart to see how it works, and then maybe afterwards put the bits that actually work back together with a dollop of motivating placebo effect on top.
The best way to approach research into helping overweight people lose weight is to study human biochemistry and motivation, and see what combinations of each work best. Not to leave the two areas thoroughly entangled and dismiss those interested in disentangling them as having the wrong motivations. I think the same goes for forming and maintaining romantic relationships.
I’d prefer not to discuss this at the ad hominem level.
Me either. I was asking you for a fourth alternative on the presumption that you might have one.
FWIW, I don’t consider any of those alternatives somehow bad, nor is my intention to use the classification to score some sort of points. People who fall into category 3 are of particular interest to me, however, because they’re people who can potentially be helped by understanding what it is they’re doing.
To put it another way, it wasn’t a rhetorical question, but one of information. If you fall in category 1 or 2, we have little further to discuss, but that’s okay. If you fall in category 3, I’d like to help you out of it. If you fall in an as-yet-to-be-seen category 4, then I get to learn something.
So, win, win, win, win, in all four cases.
The best way to approach research into helping overweight people lose weight is to study human biochemistry and motivation, and see what combinations of each work best.
This is conflating things a bit: my reference to weight loss was pointing out that “universal” weight-loss advice doesn’t really exist, so a rationalist seeking to lose weight must personally test alternatives, if he or she cannot afford to wait for science to figure out the One True Theory of Weight Loss.
My view is that the rationalist approach is to take it apart to see how it works
This presupposes that you already have something that works, which you will not have unless you first test something. Even if you are only testing scientifically-validated principles, you must still find which are applicable to your individual situation and goals!
Heck, medical science uses different treatments for different kinds of cancer, and occasionally different treatments for the same kind of cancer, depending on the situation or the actual results on an individual - does this mean that medical science is irrational? If not, then pointing a finger at the variety of situation-specific PUA advice is just rhetoric, masquerading as reasoning.
I imagine you’d put me in category #2 as I’m currently in a happy long-term relationship. However my self-model says that three years ago when I was single and looking for a partner that I would still want to know what the actual facts about the universe were, so I’d put myself in category #4, the category of people for whom it’s reflexive to ask what the suitably blinded, suitably controlled evidence says whether or not they personally have a problem at that point in their lives with achieving relevant goals.
I think we should worry about placebo effects everywhere they get in the way of finding out how the universe actually works, whether they happen to be in instrumental fields affected by motivation or somewhere else entirely.
That didn’t mean that I chose celibacy until the peer-reviewed literature could show me an optimised mate-finding strategy, of course, but it does mean that I don’t pretend that guesswork based on my experience is a substitute for proper science.
The difference between your PUA example and medicine is that medicine usually has relevant evidence for every single one of those medical decisions. (Evidence-based medicine has not yet driven the folklore out of the hospital by a long chalk but the remaining pockets of irrationality are a Very Bad Thing). Engineers use different materials for different jobs, and photographers use different lenses for different shots too. I don’t see how the fact that these people do situation-specific things gets you to the conclusion that because PUAs are doing situation-specific things too they must be right.
I don’t see how the fact that these people do situation-specific things gets you to the conclusion that because PUAs are doing situation-specific things too they must be right.
It doesn’t. It just refutes your earlier rhetorical conflation of PUA with alternative medicine on the same grounds.
At this point, I’m rather tired of you continually reframing my positions to stronger positions, which you can then show are fallacies.
I’m not saying you’re doing it on purpose (you could just be misunderstanding me, after all), but you’ve been doing it a lot, and it’s really lowering the signal-to-noise ratio. Also, you appear to disagree with some of LW’s premises about what “rationality” is. So, I don’t think continued discussion along these lines is likely to be very productive.
It doesn’t. It just refutes your earlier rhetorical conflation of PUA with alternative medicine on the same grounds.
My intent was to show that in the absence of hard evidence PUA has the same epistemic claim on us as any other genre of folklore or folk-psychology, which is to say not much.
At this point, I’m rather tired of you continually reframing my positions to stronger positions, which you can then show are fallacies.
I admit I’m struggling to understand what your positions actually are, since you are asking me questions about my motivations and accusing me of “rhetoric, not reasoning” but not telling me what you believe to be true and why you believe it to be true. Or to put it another way, I don’t believe you have given me much actual signal to work with, and hence there is a very distinct limit to how much relevant signal I can send back to you.
Maybe we should reboot this conversation and start with you telling me what you believe about PUA and why you believe it?
Maybe we should reboot this conversation and start with you telling me what you believe about PUA and why you believe it?
Ok. I’ll hang in here for a bit, since you seem sincere.
Here’s one belief: PUA literature contains a fairly large number of useful, verifiable, observational predictions about the nonverbal aspects of interactions occurring between men and women while they are becoming acquainted and/or attracted.
Why do I believe this? Because their observational predictions match personal experiences I had prior to encountering the PUA literature. This suggests to me that when it comes to concrete behavioral observations, PUAs are reasonably well-calibrated.
For that reason, I view such PUA literature—where and only where it focuses on such concrete behavioral observations—as being relatively high quality sources of raw observational data.
In this, I find PUA literature to be actually better than the majority of general self-help and personal development material, as there is often nowhere near enough in the way of raw data or experiential-level observation in self-help books.
Of course, the limitation on my statements is the precise definition of “PUA literature”, as there’s definitely a selection effect going on. I tend to ignore PUA material that is excessively misogynistic on its face, simply because extracting the underlying raw data is too… tedious, let’s say. ;-) I also tend to ignore stuff that doesn’t seem to have any connection to concrete observations.
So, my definition of “PUA literature” is thus somewhat circular: I believe good stuff is good, having carefully selected which bits to label “good”. ;-)
Another aspect of my possible selection bias is that I don’t actually read PUA literature in order to do PUA!
I read PUA literature because of its relevance to topics such as confidence, fear, perceptions of self-worth, and other more common “self-help” topics that are of interest to me or to my customers. By comparison, PUA literature (again using my self-selected subset) contains much better raw data than traditional self-help books, because it comes from people who’ve relentlessly calibrated their observations against a harder goal than just, say, “feeling confident”.
Here’s one belief: PUA literature contains a fairly large number of useful, verifiable, observational predictions about the nonverbal aspects of interactions occurring between men and women while they are becoming acquainted and/or attracted.
Why do I believe this? Because their observational predictions match personal experiences I had prior to encountering the PUA literature. This suggests to me that when it comes to concrete behavioral observations, PUAs are reasonably well-calibrated.
The problem with this line of reasoning is that there are people who believe they have relentlessly calibrated their observations against reality using high quality sources of raw observational data and that as a result they have a system that lets them win at Roulette. (Barring high-tech means to track the ball’s vector or identifying an unbalanced wheel).
Roulette seems to be an apt comparison because based on the figures someone else quoted or linked to earlier about a celebrated PUAist hitting on 10 000 women and getting 300 of them into bed, the odds of a celebrated PUAist getting laid on a single approach even according to their own claims is not far off the odds of correctly predicting exactly which hole a Roulette ball will land in.
So when these people say “I tried a new approach where I flip flopped, be-bopped, body rocked, negged, nigged, nugged and nogged, then went for the Dutch Rudder and I believe this worked well” unless they tried this on a really large number of women so that they could detect changes in a base rate of 3% success I really don’t think they have any meaningful evidence. Did their success rate go up from 3% to 4% or what, and what are their error bars?
What’s the base rate for people not using PUA techniques anyway? People other than PUAs are presumably getting laid, so it’s got to be non-zero. The closer it is to 3% the less effect PUA techniques are likely to have.
I’ve already heard the response “Look, we don’t get just one bit of data as feedback. We PUAs get all sorts of nuanced feedback about what works and does not”. If that’s so and this feedback is doing some good this should be reflected in your hit rate for getting laid. If picking up women and getting them in to bed is an unfair metric for PUA effectiveness I really think it should be called something other than PUA.
My thinking is that you don’t have enough data to distinguish whether you are in a world where PUA training has a measurable effect, from a world where PUA have an unfalsifiable mythology that allows them to explain their hits and misses to themselves, and a collection of superstitions about what works and does not, but no actual knowledge that separates them in terms of success rate from those who simply scrub up, dress up and ask a bunch of women out.
I want to see that null hypothesis satisfactorily falsified before I allow that there is an elephant in the room.
Notice that nowhere in my post did I say pickup artists get laid, let alone that they get laid more often!
Nowhere did I state anything about their predictions of what behavior works to get laid!
I even explicitly pointed out that the information I’m most interested in obtaining from PUA literature, has notthing to do with getting laid!
So just by talking about the subject of getting laid, you demonstrate a complete failure to address what I actually wrote, vs. what you appear to have imagined I wrote.
So, please re-read what I actually wrote and respond only to what I actually wrote, if you’d like me to continue to engage in this discussion.
Okay. What observable outcomes do you think you can obtain at better-than-base-rate frequencies employing these supposed insights, and why do you think you can obtain them?
As I said earlier I think that if PUA insights cannot be cashed out in a demonstrable improvement in the one statistic which you would think would matter most to them, rate of getting laid, then there is grounds to question whether these supposed insights are of any use to anyone.
But if you would prefer to use some other metric I’m willing to look at the evidence.
That didn’t mean that I chose celibacy until the peer-reviewed literature could show me an optimised mate-finding strategy, of course, but it does mean that I don’t pretend that guesswork based on my experience is a substitute for proper science.
Guesswork based on your experience isn’t supposed to be a substitute for science. It’s the part of science that you do when choosing which phenomena you want to test, well before you get to the blinding and peer review.
The flip side is that proper science isn’t a substitute for either instrumental rationality or epistemic rationality. Limiting your understanding of the world entirely to what is already published in journals gives you a model of the world that is subjectively objectively wrong.
I don’t disagree but a potentially interesting research area isn’t an elephant in the room that demands attention in a literature review, and limiting yourself to proper science is no sin in a literature review either. Only when the lessons we can learn from proper science are exhausted should we start casting about in the folklore for interesting research areas, and we certainly shouldn’t put much weight on anecdotes from this folklore. In Bayesian terms such anecdotes should shift our prior probability very, very slightly if at all.
There is a word for the problem that results from this way of thinking about instrumental advice. It’s called “akrasia”. ;-)
Again, if you could get people to do things without taking into consideration the various quirks and design flaws of the human brain (from our perspective), then self-help books would be little more than to-do lists.
In general, when I see somebody worrying about placebo effects in instrumental fields affected by motivation, I tend to assume that they are either:
Inhumanly successful and akrasia-free at all their chosen goals, (not bloody likely),
Not actually interested in the goal being discussed, having already solved it to their satisfaction (ala skinny people accusing fat people of lacking willpower), or
Very interested in the goal, but not actually doing anything about it, and thus very much in need of a reason to discount their lack of action by pointing to the lack of “scientifically” validated advice as their excuse for why they’re not doing that much.
Perhaps you can suggest a fourth alternative? ;-)
I’d prefer not to discuss this at the ad hominem level. You can assume for the sake of argument whichever of those three assumptions you prefer is correct, if it suits you. I’m indifferent to your choice—it makes no difference to my utility. I make no assumptions about why you hold the views you do.
My view is that the rationalist approach is to take it apart to see how it works, and then maybe afterwards put the bits that actually work back together with a dollop of motivating placebo effect on top.
The best way to approach research into helping overweight people lose weight is to study human biochemistry and motivation, and see what combinations of each work best. Not to leave the two areas thoroughly entangled and dismiss those interested in disentangling them as having the wrong motivations. I think the same goes for forming and maintaining romantic relationships.
Me either. I was asking you for a fourth alternative on the presumption that you might have one.
FWIW, I don’t consider any of those alternatives somehow bad, nor is my intention to use the classification to score some sort of points. People who fall into category 3 are of particular interest to me, however, because they’re people who can potentially be helped by understanding what it is they’re doing.
To put it another way, it wasn’t a rhetorical question, but one of information. If you fall in category 1 or 2, we have little further to discuss, but that’s okay. If you fall in category 3, I’d like to help you out of it. If you fall in an as-yet-to-be-seen category 4, then I get to learn something.
So, win, win, win, win, in all four cases.
This is conflating things a bit: my reference to weight loss was pointing out that “universal” weight-loss advice doesn’t really exist, so a rationalist seeking to lose weight must personally test alternatives, if he or she cannot afford to wait for science to figure out the One True Theory of Weight Loss.
This presupposes that you already have something that works, which you will not have unless you first test something. Even if you are only testing scientifically-validated principles, you must still find which are applicable to your individual situation and goals!
Heck, medical science uses different treatments for different kinds of cancer, and occasionally different treatments for the same kind of cancer, depending on the situation or the actual results on an individual - does this mean that medical science is irrational? If not, then pointing a finger at the variety of situation-specific PUA advice is just rhetoric, masquerading as reasoning.
I imagine you’d put me in category #2 as I’m currently in a happy long-term relationship. However my self-model says that three years ago when I was single and looking for a partner that I would still want to know what the actual facts about the universe were, so I’d put myself in category #4, the category of people for whom it’s reflexive to ask what the suitably blinded, suitably controlled evidence says whether or not they personally have a problem at that point in their lives with achieving relevant goals.
I think we should worry about placebo effects everywhere they get in the way of finding out how the universe actually works, whether they happen to be in instrumental fields affected by motivation or somewhere else entirely.
That didn’t mean that I chose celibacy until the peer-reviewed literature could show me an optimised mate-finding strategy, of course, but it does mean that I don’t pretend that guesswork based on my experience is a substitute for proper science.
The difference between your PUA example and medicine is that medicine usually has relevant evidence for every single one of those medical decisions. (Evidence-based medicine has not yet driven the folklore out of the hospital by a long chalk but the remaining pockets of irrationality are a Very Bad Thing). Engineers use different materials for different jobs, and photographers use different lenses for different shots too. I don’t see how the fact that these people do situation-specific things gets you to the conclusion that because PUAs are doing situation-specific things too they must be right.
It doesn’t. It just refutes your earlier rhetorical conflation of PUA with alternative medicine on the same grounds.
At this point, I’m rather tired of you continually reframing my positions to stronger positions, which you can then show are fallacies.
I’m not saying you’re doing it on purpose (you could just be misunderstanding me, after all), but you’ve been doing it a lot, and it’s really lowering the signal-to-noise ratio. Also, you appear to disagree with some of LW’s premises about what “rationality” is. So, I don’t think continued discussion along these lines is likely to be very productive.
My intent was to show that in the absence of hard evidence PUA has the same epistemic claim on us as any other genre of folklore or folk-psychology, which is to say not much.
I admit I’m struggling to understand what your positions actually are, since you are asking me questions about my motivations and accusing me of “rhetoric, not reasoning” but not telling me what you believe to be true and why you believe it to be true. Or to put it another way, I don’t believe you have given me much actual signal to work with, and hence there is a very distinct limit to how much relevant signal I can send back to you.
Maybe we should reboot this conversation and start with you telling me what you believe about PUA and why you believe it?
Ok. I’ll hang in here for a bit, since you seem sincere.
Here’s one belief: PUA literature contains a fairly large number of useful, verifiable, observational predictions about the nonverbal aspects of interactions occurring between men and women while they are becoming acquainted and/or attracted.
Why do I believe this? Because their observational predictions match personal experiences I had prior to encountering the PUA literature. This suggests to me that when it comes to concrete behavioral observations, PUAs are reasonably well-calibrated.
For that reason, I view such PUA literature—where and only where it focuses on such concrete behavioral observations—as being relatively high quality sources of raw observational data.
In this, I find PUA literature to be actually better than the majority of general self-help and personal development material, as there is often nowhere near enough in the way of raw data or experiential-level observation in self-help books.
Of course, the limitation on my statements is the precise definition of “PUA literature”, as there’s definitely a selection effect going on. I tend to ignore PUA material that is excessively misogynistic on its face, simply because extracting the underlying raw data is too… tedious, let’s say. ;-) I also tend to ignore stuff that doesn’t seem to have any connection to concrete observations.
So, my definition of “PUA literature” is thus somewhat circular: I believe good stuff is good, having carefully selected which bits to label “good”. ;-)
Another aspect of my possible selection bias is that I don’t actually read PUA literature in order to do PUA!
I read PUA literature because of its relevance to topics such as confidence, fear, perceptions of self-worth, and other more common “self-help” topics that are of interest to me or to my customers. By comparison, PUA literature (again using my self-selected subset) contains much better raw data than traditional self-help books, because it comes from people who’ve relentlessly calibrated their observations against a harder goal than just, say, “feeling confident”.
The problem with this line of reasoning is that there are people who believe they have relentlessly calibrated their observations against reality using high quality sources of raw observational data and that as a result they have a system that lets them win at Roulette. (Barring high-tech means to track the ball’s vector or identifying an unbalanced wheel).
Roulette seems to be an apt comparison because based on the figures someone else quoted or linked to earlier about a celebrated PUAist hitting on 10 000 women and getting 300 of them into bed, the odds of a celebrated PUAist getting laid on a single approach even according to their own claims is not far off the odds of correctly predicting exactly which hole a Roulette ball will land in.
So when these people say “I tried a new approach where I flip flopped, be-bopped, body rocked, negged, nigged, nugged and nogged, then went for the Dutch Rudder and I believe this worked well” unless they tried this on a really large number of women so that they could detect changes in a base rate of 3% success I really don’t think they have any meaningful evidence. Did their success rate go up from 3% to 4% or what, and what are their error bars?
What’s the base rate for people not using PUA techniques anyway? People other than PUAs are presumably getting laid, so it’s got to be non-zero. The closer it is to 3% the less effect PUA techniques are likely to have.
I’ve already heard the response “Look, we don’t get just one bit of data as feedback. We PUAs get all sorts of nuanced feedback about what works and does not”. If that’s so and this feedback is doing some good this should be reflected in your hit rate for getting laid. If picking up women and getting them in to bed is an unfair metric for PUA effectiveness I really think it should be called something other than PUA.
My thinking is that you don’t have enough data to distinguish whether you are in a world where PUA training has a measurable effect, from a world where PUA have an unfalsifiable mythology that allows them to explain their hits and misses to themselves, and a collection of superstitions about what works and does not, but no actual knowledge that separates them in terms of success rate from those who simply scrub up, dress up and ask a bunch of women out.
I want to see that null hypothesis satisfactorily falsified before I allow that there is an elephant in the room.
Once again, you are misstating my claims.
Notice that nowhere in my post did I say pickup artists get laid, let alone that they get laid more often!
Nowhere did I state anything about their predictions of what behavior works to get laid!
I even explicitly pointed out that the information I’m most interested in obtaining from PUA literature, has notthing to do with getting laid!
So just by talking about the subject of getting laid, you demonstrate a complete failure to address what I actually wrote, vs. what you appear to have imagined I wrote.
So, please re-read what I actually wrote and respond only to what I actually wrote, if you’d like me to continue to engage in this discussion.
Okay. What observable outcomes do you think you can obtain at better-than-base-rate frequencies employing these supposed insights, and why do you think you can obtain them?
As I said earlier I think that if PUA insights cannot be cashed out in a demonstrable improvement in the one statistic which you would think would matter most to them, rate of getting laid, then there is grounds to question whether these supposed insights are of any use to anyone.
But if you would prefer to use some other metric I’m willing to look at the evidence.
Guesswork based on your experience isn’t supposed to be a substitute for science. It’s the part of science that you do when choosing which phenomena you want to test, well before you get to the blinding and peer review.
The flip side is that proper science isn’t a substitute for either instrumental rationality or epistemic rationality. Limiting your understanding of the world entirely to what is already published in journals gives you a model of the world that is subjectively objectively wrong.
I don’t disagree but a potentially interesting research area isn’t an elephant in the room that demands attention in a literature review, and limiting yourself to proper science is no sin in a literature review either. Only when the lessons we can learn from proper science are exhausted should we start casting about in the folklore for interesting research areas, and we certainly shouldn’t put much weight on anecdotes from this folklore. In Bayesian terms such anecdotes should shift our prior probability very, very slightly if at all.
No ad hominem fallacy present in grandparent.