Brilliant jerks, crazy hotties, and other artifacts of range restriction (2019)
towardsdatascience.com
towardsdatascience.com
Berkson's Paradox - https://news.ycombinator.com/item?id=26566810 - March 2021 (39 comments)
Berkson's Paradox - https://news.ycombinator.com/item?id=18667423 - Dec 2018 (21 comments)
Berkson's Paradox - https://news.ycombinator.com/item?id=8264252 - Sept 2014 (20 comments)
"You can't be hot and a security specialist!"
Just because the average women takes more care of herself than the average man.
There are a fair few outliers to this rule! Usually as a result of car-crash TV, though not always.
Survivor bias arises when you study what you think is a full population, while in fact you're looking at a selected sample that did survive the very thing you're studying.
A classical example in finance happens when you want to study, say, the performance of tech companies over the last 10 years. If you just pick current tech companies and look at their past performance, you will be biased towards companies that did make it. You have to include companies that don't exist anymore but would have been included were you to perform the study say 5 years ago.
"Survivorship" is the collider being conditioned on.
Especially in the modern NBA, where players are pushed very hard to get out from the post and shoot. Being unable to shoot puts a limit on a players career, even at 7+ feet. There’s a reason why in the off-season every big in the league posts insta videos of themselves draining 3s in an empty gym.
The level of competition in the NBA is fierce, as is the incentive to learn to shoot. If many bigs don’t, the only reasonable explanation is that it’s physically more challenging.
They only need to be better than a few dozen or hundreds of others at the required skills. The average 7 footer in the NBA is a way better shooter than the average player in the general population and probably better than any average college player. But shooting skill is inversely proportional to height in the NBA for the exact reason the GP stated.
To be something like 6'3'' or shorter in the NBA means you need to be an insane outlier in other skills, better than literally millions of others at your height and hundreds of thousands with your general athleticism.
That said, it still seems reasonable that if you're going to be paid millions of dollars a year to throw a ball in a hole, it wouldn't be crazy to be expected to practice and get good enough not to miss if literally no one is trying to block you (e.g. free throws). I guess that if you're good enough at the other skills, it still helps your team win enough to sign you even if you don't get good at that, but it still feels like fairly low hanging fruit...
I know that sounds facetious, but like, there’s a reason no one jumps on a free throw. It’s easier if you don’t. But you have to jump to get your shot over a defender.
If you are sufficiently tall, you could just… jump less, so that your release height is the same.
I actually think the bigger reason is simply that there are more important skills for tall players to work on. If your main job is rebounding , it is better to work on rebounding than shooting.
Of course, the changes to how NBA teams play over the last decade or so have changed this, which is presumably why you now see more players who are both tall _and_ good shooters.
Same with the dating situation. Must be more comforting to think "I'm not the most attractive person ever, but at least I'm sane and many of the more attractive people aren't" rather than "There are probably people out there who are both more attractive and with a nicer personality".
I also suspect people tend to think everything in life must have tradeoffs, D&D character points style. Like, you have X points to go around in total and if a percentage of that goes to intelligence/skill, there must be very little left over to put into charisma.
You can see that a lot in films and TV shows, where being popular or attractive is often treated as if it's incompatible with being intelligent or a nice person. But that's not reality, and some intelligent/attractive people are also perfectly nice, sane people.
The truth is probably somewhere in between though. The time/energy constraint is what's most apparent to me since working full time.
As an adult you have to cram cooking, cleaning, hobbies, relationship building, etc all into your small limited evening windows and weekends.
If you want to be good at something you need to put time and energy into it. This means that you can't be good at everything.
So if you like to spend your time reading, maybe selecting nicer clothes (that often need ironing or handwashing, i.e. time) will maybe fall by the wayside, because you prefer reading books. Meaning that you look worse than you could, because you prefer your hobby.
Similarly somebody who spends a lot of time on their outward appearance might not have the time left to be as interesting as they could be.
In the end of course you may have talent or natural good looks. But as soon as you start needing to think about where to allocate time you will face tradeoffs.
These people have resumés that look like they rolled a 20 for every one of their stats.
"America's Finest".
On the other hand, meeting more attractive persons in real life proved much less biased. YMMV
All that said, very attractive persons can go a little crazy from the stares and clumsy approaches that follow them everywhere they go. To the point that they avoid public places like coffee shops or feel uneasy about going to the grocery store or even to work. So don’t do those things its not nice
Advice like this makes the situation worse, because nice people will take your advice and not approach extremely attractive women, because they understand it is the nice thing to do, but assholes will approach extremely attractive women, because they don't care about being nice. So these women end up only being approached by bad actors.
If you approach someone, be honest with your intentions while also caring for them. What does that mean? Both attractive and un-attractive women like to date people and get attention. So it’s okay to approach an attractive women. However, when you do it, be honest with yourself that the reason you want to be around her is whatever (a good relationship) and then if she says no, just leave it at that.
Approaching? Ok. Approaching to be a friend? Ok. Approaching to be a friend because you want to have sexy time later? Not ok, this isn’t honest. Not honest = not ok = being a jerk. Approaching because you share an interest in Venus fly trap plants and you want to talk about it? Perfect. She says she doesn’t want to talk? That’s ok. Leave it at that.
>Approaching to be a friend because you want to have sexy time later
Some would argue that first almost always implies second
cant read this article without "upgrading". no i dont want to "upgrade". can we downrank Medium until they stop doing bullshit like this? (yes towardsdatascience is a well known data publication, would love to push them gently into moving off medium)
The specific scenario they're modeling here, where a bunch of samples with correlated properties x and y are clustered by x + y, essentially takes a strip of datapoints that are distributed over a diagonally-oriented rectangle, and then cuts it into a series of shorter strips, by cutting at thresholds (lines where x+y=n) that run along the opposite diagonal. If the resulting short strips are narrow enough (if the upper and lower thresholds are closer together than that degree to which x and y typically differ), they will be rectangles oriented the other way - in other words they will represent groups with strong negative correlation.
Reminds me of someone I've seen in the news a lot lately.
The self-deception assumption is charitable. Of course it's also possible to just be a pathological liar; the author is being polite.
Not great examples. Source #1 (https://www.nytimes.com/2018/02/21/well/move/how-exercise-ma...) says "And of course, mice are not people, and it is impossible to know if the same changes occur in our synapses when we exercise." and source #2 (https://www.project-syndicate.org/commentary/why-do-smart-pe...) says that smart people tend to live longer and they list several reasons for that including a lower accident and homicide rates. One contributing factor was that smart people are less likely to have certain health problems like high blood pressure and cholesterol that can kill people later in life. It's a big leap from that to smart people are healthier.
The population of everyone without Down syndrome is both smarter and healthier than the total population. That extends to things we don’t think of as lowering intelligence like high blood pressure: https://www.nia.nih.gov/news/high-blood-pressure-linked-cogn...
The classical example is taking the same random test twice and seeing a lower result the second time around. In the post there's a correlation between the tests, so the second test will not have a low value in absolute terms, but will be low relative to the first one (negative correlation). Taking test B before or after the filtering shouldn't matter.
My intuition is that this sort of selection picks some samples that got "lucky" on test A, and will score closer to the mean on test B, bringing the distribution down and obscuring or reverting any potential correlations.
[1]: https://en.wikipedia.org/wiki/Regression_toward_the_mean
They can both be true. The selection process is enough to explain a negative correlation regardless of whether the true correlation is negative, zero, or strongly positive.
Separately, it's very possible for a girl who naturally has 40 attractiveness points and 15 personality points, but who's happy dating men who only demand 30 points from their partners, to lower her expressed personality from +15 to -10. This would be a causative influence of high attractiveness (interacting with low ambition) on personality.
I do think there is a big correlation between crazyness and “professional hot people” (sex workers, intimacy workers, many service people, models, actors, etc), being that the mental game of attraction is inherently irrational and it helps to operate on vibes vs. numbers.
Sounds like a certain entrepreneur who bought a social media site he didn’t actually want anymore, and now is asking everyone to work extra hardcore.
> "...and mutational load affects both facial symmetry and mental illness."
There's little solid evidence for inherited mental illness, not excluding severe neuro-developmental abnormalities (which may also have environmental causes such as teratogen exposure), in fact there aren't even any solid independent physical-chemical tests for DSM disorders.
A quick web search seems to suggest that ADHD, depression, and autism are all known to run in families?
> (8) Family genetic studies. The phrase “family genetic studies” is commonly used in psychiatry to refer to designs in which investigators examine the familial aggregation of one or more disorders, such as panic disorder or major depression, within intact (i.e., non-adoptive) families (e.g., Weissman, 1993). Given that the familial aggregation of one or more disorders within intact families could be due to shared environment rather than—or in addition to—shared genes (Smoller and Finn, 2003), the phrase “family genetic study” is misleading. This term implies erroneously that familial clustering of a disorder is necessarily more likely to be genetic than environmental. It may also imply incorrectly (Kendler and Neale, 2009) that studies of intact families permit investigators to disentangle the effects of shared genes from shared environment. Twin or adoption studies are necessary to accomplish this goal.
> (9) Genetically determined Few if any psychological capacities are genetically “determined”; at most, they are genetically influenced. Even schizophrenia, which is among the most heritable of all mental disorders, appears to have a heritability of between 70 and 90% as estimated by twin designs (Mulle, 2012), leaving room for still undetermined environmental influences. Moreover, data strongly suggest that schizophrenia and most other major mental disorders are highly polygenic. In addition, the heritability of most adult personality traits, such as neuroticism and extraversion, appears to be between 30 and 60% (Kandler, 2012). This finding again points to a potent role for environmental influences.
> Even schizophrenia, which is among the most heritable of all mental disorders, appears to have a heritability of between 70 and 90% as estimated by twin designs (Mulle, 2012), leaving room for still undetermined environmental influences.
You have set the bar high.
They basically look at fraternal and identical twins and assume (because their environments are otherwise probably close) that any differences in outcome are heritable. It’s a huge leap of faith.
You’re making it sound like the researchers are idiots who have never thought about how their tools might not be measuring what they hope they are.
As well as identical/fraternal twins you can also look at other sorts of relatedness, e.g. half siblings are on average as related as first cousins and on average you’re as closely related to each parent as each full sibling. Psychological traits are inherited like height is, not like language is[1]. Mental illness is highly heritable though the expression varies[2].
As a parallel line of evidence to confirm classical twin studies and those based on degrees of shared ancestry cheap genetic testing has allowed testing how closely people are related. So you don’t have to assume a sibling shares 50% of their genes. You can see if they’re 73% similar or 41%. Or you can take people who have no known shared recent ancestry and see how similar they are on the trait of interest.
To the best of my knowledge behaviour genetics is holding up. For psychological traits genetics is more powerful than developmental noise.
[1]> Genetic Influence on Human Psychological Traits
> There is now a large body of evidence that supports the conclusion that individual differences in most, if not all, reliably measured psychological traits, normal and abnormal, are substantively influenced by genetic factors. This fact has important implications for research and theory building in psychology, as evidence of genetic influence unleashes a cascade of questions regarding the sources of variance in such traits. A brief list of those questions is provided, and representative findings regarding genetic and environmental influences are presented for the domains of personality, intelligence, psycho- logical interests, psychiatric illnesses, and social attitudes. These findings are consistent with those reported for the traits of other species and for many human physical traits, suggesting that they may represent a general biological phenomenon.
https://utahpolicy.com/wp-content/uploads/2013/12/bouchard.0...
[2]> The p factor: genetic analyses support a general dimension of psychopathology in childhood and adolescence
> Diverse forms of psychopathology generally load on a common p factor, which is highly heritable. There are substantial genetic influences on the stability of p across childhood. Our analyses indicate genetic overlap between general risk for psychiatric disorders in adulthood and p in childhood, even as young as age 7.
https://acamh.onlinelibrary.wiley.com/doi/pdfdirect/10.1111/...
You’re basically saying “some limits of cognition are heritable” which is so obvious as to be inarguable.
The huge qualifier in what you quoted is “reliably measured psychological traits.” They also have to be defined within psychological language and linked to genetics through behavior or biological indicators. There are exceedingly few of these, and schizophrenia, autism, etc, are not them.
Psychology as a whole without evolutionary considerations to articulate on is free-floating.
There is a long history of psychologists justifying their approach, but the reality is that reproducibility in psychology is still at something like 35%.
If I understand this right, GWAS are capturing everything from the genome.
This includes any genetic determination of appearance, food preferences, etc.
Then the sample is only healthy twins from narrow genetic/cultural stock: England/Wales born over two years in the 90s.
This means that in addition to sweeping up genetic factors associated with brain development, you are capturing all visible/inherent-behavioral inputs to a multi-stage developmental process that includes interactions with technology, other people, school systems, etc.
Of course you can explain a majority of the outcome. It won’t replicate though, because it is dependent on that narrow time slice of culture, nutrition, parenting practices, technology, etc.
If this study replicates in Asia, South America, or Africa, with similar genetic components in the PCA, I’ll admit I’m wrong.
Heck if it even replicates in England with twins from the 1960s I’ll admit I’m wrong.
We have known for millennia that original sin is carried and transmitted by everyone in the human race. John Bradshaw posited the thesis that shame is carried in families, across generations. We all know that cycles of abuse happen again and again, passed father to son, mother to daughter, uncle to niece.
Psychiatrists flail about and disclaim any knowledge about the genesis of "mental illness". But we know the causes. The causes are preventable, but sometimes the cost of prevention is everything we hold dear.