The really astounding thing about this is that the scientific community found out about this because he was posting about his lab on his blog and wrote a post about how one student had the grit and determination to get the p-values below .05 and the other didn't. The guy truly seems to not have understood what he was doing wrong.
This is how I assumed 95% of the nutritional science that gets in the newsfeeds works. Wansink can't be responsible for all or it, or even most.
It's so utterly aggravating to me that a person like this gets handed so much in the way of resources and free passes. So many people must have known that he is the type of person that cuts corners and thinks in lazy patterns, and yet he kept getting rewarded.
This is how I assumed 95% of the nutritional science that gets in the newsfeeds works.
If you try hard, I bet you can assume that about 97% of it.
I don't think I've ever done an interesting study where the data 'came out' the first time I looked at it," he told her over email.
FFS.
I could understand not having all your best hypotheses a priori, but if you can't at a minimum partition your data and validate your new hypotheses against a test set--yet alone actually going out and getting new data--you aren't being serious.
This seems so inefficient. He actually conducted all these studies! I guess it wasn't enough to fool everyone else, he had to fool himself too.
I recently competed for some internal funds, and my co-applicant and I were uniquely identified as the only project (out of 20ish) not deserving of funds.
The money is peanuts and we were able to find alternate funding, but the situation is so egregious that I keep being so angry about it. Basically it was a call for undergraduate summer research projects. Ours is in the vein of these gerrymandering/redistricting projects. Others that got funding aren't even research, and still others aren't original research - walking a student through a known result.
The reason we've been told that we were denied funding is that everyone else described how they'd promote their students' results next year, and we didn't. I went meticulously through the application and addressed every point, and there was nothing close to that in the application. Of course I'd pack them along to conferences to present and have them write it up if it's publishable, etc.
I suppose I'm just venting. Whenever I tell the story, others respond that the "promote results" is an ad hoc explanation to cover up that it must have been political. The admin person who I usually respect a lot has sworn it wasn't partisanship, and my colleague and I don't stir up strong negative emotions in anyone on campus, I'm pretty sure.
Ah well, this Wasink guy sure did get a lot of freebies!
7: You can do post hoc tests without subsetting. We do at times, because data points cost a fuckton of money. You just need to be honest about what you are doing (e.g. report the null finding on the a prior tests and detail the number and kinds of all post hoc tests you do) so that reviewers can decide whether it is worth publishing or not.
data points cost a fuckton of money
I know what this means, but I love this way of putting it. "If we save up our allowances for a year, we'll be able to buy a datapoint each!"
Has anyone done a study on what proportion of Gladwell's work has survived the replication crisis?
You know what would be really interesting? For some stats-critic type (Andrew Gelman is the name I know, but there must be others) to identify click-baity type studies where there's enough information known about them to be sure that they're not garbage. You couldn't be sure the results were the truth, but 'not garbage' would be interesting.
I should have said "degrees of freedom," not "data points."
I suppose that's what peer review is supposed to be, kind of, but it doesn't seem to be at all practical to make it rigorous enough on a regular basis.
10: Huh. I would have thought that the smaller n would be more than balanced by the lack of multiple comparisons. But I guess as long as you're being honest, the system can correct, with the understanding that your data might be slightly lower quality, and that your study is a candidate for retesting (if it said anything meaningful).
A smaller N doesn't mean there are fewer comparisons somebody wants to run.
I meant lack of multiple comparisons on the test set.
Oh. Yes, it would, if there were sufficient power to run the comparisons after subsetting. There usually is not.
This seems the right place to link to this list, in case it hasn't been mentioned here before.
The solution is to apply the time-honoured tactic of circumlocution to disguise the non-significant result as something more interesting. The following list is culled from peer-reviewed journal articles in which (a) the authors set themselves the threshold of 0.05 for significance, (b) failed to achieve that threshold value for p and (c) described it in such a way as to make it seem more interesting.
That annoys me. Yes, if you establish a cut point for significance, you should report based on that cut point. But it's not like .05 is an objective line between true and falsehood.
THE HELL YOU SAY!
It really depends on what you're doing, but I'm guessing that even the FDA isn't going to say "trial medication X was not associated with nuts-falling-off syndrome" and not require a warning at p = .054.
They will, for good reason, hold you to the planned comparison at p = .05 (or higher) for approval.
13. Andrew Gelman has written about or linked to others on Wansink extensively. He is quite perturbed about it.
Right, that's why I thought of him. But it'd be really interesting to see him doing the reverse -- 'this clickbaity research appears to have been done respectably, with high professional standards.' That's probably harder, though.
The click-baity stuff is supposed to say something clear and surprising about some situation that's noisy and complicated. Which is to say, even if you do the research completely respectably, if you get a strong result that doesn't mean the conclusion is correct it just means that you got a false conclusion honestly due to bad (good?) luck. Gelman has lots of posts about this, here's the most recent one. He does have posts on research he likes, but you can usually tell the difference from the click-baity ones just by seeing the claims and kind of data produced, and ignoring the details of how the work was done poorly
Local mom has this one weird trick for research with large numbers of statistical test, but p-hackers don't want you to know about her Bonferroni correction.
"If your theory is weak and your data are noisy, all the preregistration in the world won't save you."
link
Right. If somebody just preregisters 20,000 tests, it's still going to be 1,000 false positives at .05.
All you've done is made their graduate student miserable for having to type out the 20,000 tests. Which, is a win, but only a small win.
Only 66.67% of those commas were needed.
On the plus side, the graduate student might become a case study for the next paper on grit.
I, for one, think it's just fine to write a clickbaity research paper every once in a while.
"Local scientist discovers boson with the key to weight loss."
Nestled in her boson, the scientist found a key.
Isn't this what happens when a university sets up an incentive structure that demands a high publication rate, with - ideally - generation of positive media exposure for the university? Livings have to be made whether the secrets of the universe and its inhabitants are there to be revealed or not.
25: Wansink has also been really shitty about it.
Also, 33, here's a great link about grit!
"This paper is used as a foundation for the entire literature on grit that made Angela Duckworth famous but there is simply no evidence that it is appropriate to combine perseverance and passion to form a construct called grit."
38: Generations of white middle infielders cry out in protest. (Or at least their sportswriters and commentators do.)
These kinds of research problems aren't just about nutrition or even social science. It's about a general problem in an academic pursuit that doesn't really answer to anyone and yet somehow continues to be fawned over by the public who's dying to spend $70,000/year to send their kids to the "best" schools.
At what point does the establishment have to actually answer for what they do? At what point will the public question the relevance of the academy? I wish I could be optimistic that it will improve.
I think the public already questions the relevance of most of the academy, with the EXCEPTION of people like Wansink who are doing "applied" and "relevant" research on consumer products and how the average Joe can learn tips to better learn to navigate the modern world.
That's how he got away with it, by fulfilling Cornell's need for publicity by being engaging and writing books and giving TED talks and partnering with corporate entities to turn his research into real-world projects, not by getting big research grants like most scientists. He wasn't very answerable to the scientific community.
A) Is "answerable" a word? Is that what it means? No matter.
B) Sorry for the part about learning to learn things. My brain has been broken by the editable or deletable comments everywhere else on the internet.
THE BLOG ISN'T WRITTEN IN PENCIL, NED. IT'S WRITTEN IN INK.
41: See also the Scott Walker plan for U of Wisconsin. Building a Better Future for a Foxconn-dominated Serfdom.
He's finally decided to hold elections to fill vacant seats at least.
Because a bad black man* forced him to waste the hard-earned money of Wisconsites in his telling.
*Eric Holder.
It does make me wonder as to how much p-hacking, intentionally or not, goes on in the worlds of 'data science' and 'machine learning'.
I never took statistics and am not inclined right now to spend the time or money on taking a class. Are there any "Teach yourself Statistics 101 for Dummies" books that go at your own pace.?
I'd probably be willing to devote about 45 minutes every other week to teaching myself.
Even cooler is if the include bits about the history of how a concept was developed.
50: For data science, certainly. But for machine learning, the metric for success isn't publishable units* but how well the software does what it's intended to do; this can be measured against unseen data points (a test set, as I was saying above) or whether it creates interesting output (in a generative-adversarial network) or whatever.
* For ML in practice, that is; in an academic setting ML is more mathematical than statistical, so it's going to usually be proofs of mathematical properties of models or whatnot.
I don't really remember how I learned statistics. The main thing I work on now is how to say non-committal things in meetings so I can go look up the answer later when I have time to think.
51: It's now been years since I've read it, but I used to recommend How to Lie with Statistics.
I see that _Statistics Done Wrong_ is available online: https://www.statisticsdonewrong.com I remember liking it.
I don't actually know what "data science" is. This is not a request for information.
Anyway, the U.K. has just discovered super gonorrhea, so maybe science will get funding for research into antibiotics and less for buzz-worthy nutrition studies.
Maybe it's "Super Gonorrhea". I don't have my Chicago Manual handy.
"But for machine learning, the metric for success isn't publishable units* but how well the software does what it's intended to do"
Maybe - but that all depends on whether actually solving the problem accurately is ultimately 'what it's intended to do' - and whether there is constant re-testing/re-training post the initial training period.
There'll be plenty of cases where to the operator of the software, having something somewhat accurate that - combined with no reasonable chance of appeal - saves money means that no one looks at it particularly closely.
Think ML software to assess people's fitness for work, or presence on a watchlist etc etc.
Someone tried to trace down the origin of the "tanks in a forest" story, and while long, it is quite interesting reading. You can skip to near the end for the payoff, if you wish.
Ed Fredkin (an AI researcher, if you aren't familar with the field), was at a meeting in LA and the original "tank recognizer" paper was presented. This was in about 1964. Ed reported (at LessWrong):
"At the end of the talk I stood up and made the comment that it was obvious that the picture with the tanks was made on a sunny day while the other picture (of the same field without the tanks) was made on a cloudy day. I suggested that the 'neural net' had merely trained itself to recognize the difference between a bright picture and a dim picture."
I remember hearing the story very early in my career and told by someone who almost certainly heard it directly from Ed.
As an aside, Ed also suggested the idea of hair-cutting robots who would live in your hair like little mechanical lice and keep each strand at exactly the length you wanted.
60: True. I'd consider that a qualitatively different problem from p-hacking, and no different from what might occur under human judgment, beyond the unearned respect some people give to automated judgments.
||
Does anyone have recommendations for cordless electric lawnmowers? I don't really work well with internal combustion engines.
||>
63: How much lawn? I'm pretty happy with a non-powered reel mower (one of these) for around 1000 square feet of lawn.
If you really want to go electric, I'd look at the Ryobi 40V stuff. I have their string trimmer from the same product line, I like it, and it's nice that they can all share the (expensive!) battery.
It's not a huge lawn -- I think the lot is about 45' x 125', incl. A bunch of driveway, so I don't need a veritable monster, I just want something durable.
62.
I think they shade into the same thing - mining a set of past data to find correlations with a certain probability weighting in a vaguely scientific way, to give an imprimatur of objectivity to the end result.
64: Yeah, I actually have a push mower, but the yard is pretty bumpy, and it never really worked that well. Too dull, I think. There's only one hardware store in town that still sharpens them.
(The diminution of neighborhood [and even regionally-famous] hardware stores in this city is quite advanced. The one I went to as a kid is still hanging on, mostly because it's in an ever-wealthening neighborhood and they keep the mark-ups high. But a lot of the ones that were known as places to find extremely rare/unusual tools & hardware have gone the way of the dodo.)
I don't actually know what "data science" is. This is not a request for information.
I don't know what it is either, but one of my colleagues downloaded an off-the-shelf neural network package and had some undergrads write a very quick paper and then the university gave him an extra $50k in funding because suddenly he was a data scientist.
The people a couple floors down with the experiment that generates terabytes of data a year that they have to analyze? Not data scientists. They don't even use Python.
You can do a lot with an IBM punch card machine and a dream.
I have a push mower with the spinning reel thing. I think that what it takes to be successful with it is a realization that a lawn mowed with a reel mower is a fundamentally different product than a lawn mowed with a power mower.
You aren't going to get the same straight lines and even heights in your mowed grass. On the other hand, you also don't have to walk around in a pattern covering everything. You just pass the mower over the higher parts of the grass for ten minutes every few days, and that gets you a perfectly serviceable lawn.
ll
the dairy queen household is on the move! it is a class SF move, so we'll be paying more $$ for less space, hence some serious purging of possessions.* i write because we have a small single bed + rollaway bed combo that we purchased for a reasonable price when our lad was quite small and we were quite skint. he used it for years - he went from crib to the bed with the rollaway positioned strategically to break falls, and then the rollaway was super useful for sleepovers, all in all it is a swell device. surely this group knows of a sf bay area household with one or more small children who would find this set up extremely useful? we would hate for it not to be used, it has years of usefulness in it. anyone interested???
l>
*have i mentioned how much i love the library? i so love the library. because they keep all the books. so so so many books. i don't feel this way about the lps, oddly, even though they are bulky and very very heavy. also much cogitation and strategic planning with respect to the ancestral stove. i am very attached to its btus. it would be so hard to go back to normal fire power, like living in black and white. but are we really going to be the people moving in their 25+ year old viking???
Wow. Your current digs are so charming! The new place must clear a high bar.
||
Unbroken processions [of laborers] extended one after another for a thousand li. Among the laborers drafted for the eastern capital, four or five out of every ten fell prostrate and died on account of being under such great pressure. Every month in carrying the dead bodies east to Chengkao and north to Heyang [along the Yellow River] the carts were always in view of each other on the road.|>
I just figured out where the deer in my neighborhood live. On the hill behind my house there are six of them. It's full daylight and they're not going anywhere. They don't look thin and they seem to have plenty of energy to chase each other around. I'm going to Lyme disease next time I cook out.
Yes! High on Edwardian charm & great neighbourhood (much less stuffy), but much less storage space, and Yes! A trundle bed!
I'm sure you'll be fine if you grill them thoroughly.
I have their string trimmer
The first time I heard the term "string trimmer," I was at a vineyard, and the winemaker was telling me they'd just finished up a bunch of work with string trimmers. It took me like five more minutes of context to realize that "string trimmer" = "weed whacker."
One would expect oenophiles to have some weird name for scissors.
It's not a pruning shear thing. It's a machine that spins a piece of string (or plastic cord) around in a circle with sufficient force to cut grass.
I know what a weed-whacker is.
I've only just realized that "strimmer," which is what they're called here, comes from "string trimmer." Duh.
I thought Americans had lawns in a way that others didn't understand.
77: And the park. And more visibly the cemetery north of the park--every time I take a walk there I'll see multiple groups.
But why should Moby go all the way to the park when the deer and the barbecue are right there in his own yard?
True, for Lyme disease convenience it can't be beat.
Am I really the only person having trouble not reading his name as "Professor Wanksink"??
OT ATM: I just need to check in about something. I am trying to make something happen with a shy nerd and bewilderingly (if he liked me) he easily had the opportunity to be the last person at my house last night when we'd all been drinking and then when it was clear that was going to happen hurriedly announced he would go. I was disappointed and sad. I had cleaned my room just in case he wanted to be in it! But then after a few moments of disappointment I thought, does this mean he doesn't like me? Not necessarily. It may have been last minute panic. So I wrote him this a few minutes after he left:
"I am going to take advantage of my lowered inhibitions right now to say: I'm sorry you had to go. Ever since we sat next to each other during that Tidal Cycles jam I thought, this is nice and I'd like to keep doing it. If you want to spend time together at some point in the future, let me know. "
And he wrote back: "Sure!"
I guess I would have liked, given the effort I went to up there, to get a few more words back, but I what I would really like to check in about is: I unambiguously asked him out, right? No one could possibly misunderstand that? So I can follow up with making plans for this week without being weird and assume he also thinks he is on a date? I guess I'm just worried because I feel like given some of my recent non-verbal communication, someone who left my house last night would have to be one of: pretty confused, not interested, or skittish. So I just want to know that I've communicated clearly enough at this point that 1 and 2 are not what's happening.
This guy is such a shy skittish one that he's been absolutely careful not to tip his hand either way. Which is to say that he could be interested or not interested, but he's anxious enough not to be willing to communicate whatever's underneath. (That "sure!" is particularly annoying to me. "Sure, I'd like to spend time with you!" or "Sure, I will let you know if that happens!"?)
However, you have been perfectly clear and he's not misunderstanding you. He's just having a big intense wave of anxiety and either isn't clear what he wants, or is too shy to say so yet. If you proceed with making plans for next week, it will be clear to him that it's a date. He might continue to muddy the waters though, because he finds that a safe place to sit.
I unambiguously asked him out, right?
Nope, you ambiguously asked him out. Unambiguously wouldn't have started with a disclaimer about being drunk that in his mind means the whole thing can be explained away as "oh, she probably didn't really mean that."
"Sure!" was the defensive response in case you didn't really mean that.
But yeah, go ahead an ask him out for an evening of skee-ball and drinks, or whatever it is the kids are into these days.
People don't actually ask each other out if they aren't drunk.
That's why cultures without binge drinking have traditions of arranged marriage.
Except Mormons, who are just straight-up horny.
94 seems pretty correct. It may also be that he doesn't like to take the initiative, for fear of rejection. (Not quite the same thing as being shy.) Does he talk during these get-togethers? Does he start conversations or just sort of join them once they are going?
IME shy nerds tend to find the whole rigmarole about asking/being asked out intimidating to the point of physical nausea. Depending on how much effort Tia is willing to put into it this can be got round by the other person being completely up front about what they'd like to do- "Shall we have coffee after work tomorrow?"- until the relationship has reached the level they want. Don't expect much in the way of initiative from the shy nerd until they're confident that the relationship has stabilised at some level they understand, after which they can reciprocate happily. Shy nerds over-analyse everything, it's who they are: be forewarned.
After all that trouble, you should steal a kidney.
I just assumed that organ theft was what she was trying to make happen.
He talks but he'll never initiate a conversation. He also has his phone out a lot. He's doing pretty well at touching me much more than is necessary (which I am also doing to him). I mean, honestly, given that on Thursday night I gratuitously leaned on him for like 20 minutes and he didn't move away I wish he could have braved five minutes alone with me. Anyway, I'll try to get him to go out this week. Sheesh.
Sigh. When I did that, I got a reply that, even if somehow it wasn't *supposed* to totally shut me down, would so predictably feel that way that I probably don't want to date someone who can't be a little more giving in communication than this. I don't really understand. He has definitely been touching me a lot, too, although I guess only when drinking. Oh well.
Best of luck. If you're already on with the prolonged physical contact, I suggest you go for coffee, invite him... no tell him he's coming to dinner at your place next Friday, have a few drinks and fuck him. As a gazetted shy nerd, I know this works: it was the technique used by my partner of 38 years.
Good luck, Tia, you're not doing anything wrong.
I can count several instances in which shy nerd panic thwarted me in situations which, considered in retrospect, I totally could have gotten laid. There is a weird incongruity that happens when a hot girl is into you after so many cases of hot girls not even seeing you.
106 Same here. I'm sure all the instances will come flooding back to me on my deathbed, "damn my eyes, I really could have got with Jennifer."
93 is right. Unless you're prepared to do all the work for the indefinite future, while he ventures nothing because he has to protect his fragile ego at all times, forget about it.
Don't people just text naked pics back and forth these days anyway?
106, 107 -- You can console yourself that even if you're not better off now than you would have been in the alternative timeline where you'd understood the possibilities better, she's probably better off in this timeline.
"For the greater good."
Hold on wait. Barry isn't a wizard?
I think dead-guy Barry in "The Casual Vacancy" was s wizard.The rest of the village, especially the men, are such shits that there must be some special gift he had to have been decent.
106,7- in college a girl who always sat next to me in an early morning class would frequently lean her head on my shoulder at the start of class and I just sat there not moving, figuring it was a morning "haha look how tired I am" thing and never said anything about it. I did have a girlfriend at the time (moderately long distance) and I think she knew that so it wasn't until years later I thought Wait did she like me?
Jesus Christ. America is going to die out. I thoroughly recommend Chris Y's suggestion, though possibly with contemporary mores you need to make the whole thing explicit in your invitation: "Hi, I was thinking you could come round on Friday and have supper and a few drinks before I jump on your bones and ravish you all night. 6:30 for 7:00
RSVP"
America is going to die out
We were only trying to help.
I was once at home with someone alone, who was clearly interested in me, and at some late hour she remarked "Oh, the buses have stopped running. Guess you'll have to stay here!" My extremely suave reaction: "Oh, don't worry, I can walk home." [six miles].
(It did work out, eventually.)
That's right up there with replying to "Can I interest you in a night cap?" with "No thanks, I never wear one."
Thank you for making that explicit.
Of course, a lot of things worked for him.
"Oh, don't worry, I can walk home." [six miles].
I've done that (in High School, and the person I was talking with was married and not, I believe, interested in me).
So many hideously familiar situations in the last 15 or so comments. Tia, you have about six brains, you can do better than we did.
Y'all. (By which I mean, anyone suggesting I take an aggressively explicit approach right now.)
Please imagine yourselves speaking to a woman that way in this situation. Imagine all the reasons that would be kind of horrifying. Now imagine I feel the same way! Srsly, I:
1) am anxious about rejection, just like you would be
2) don't want to be presumptuous about what he wants or say anything that indicates that I don't care about his desires, anxieties, ambivalence, or enthusiastic consent
3) don't know for sure that he doesn't have some partner I know about, or even that he is attracted to women, or is generally in the category of available at all
4) don't actually want to announce that I want to have sex, because I'm very far from knowing that's what I want -- I wanted to be alone with him, and physically close to him, and to see how that felt
This would all be true if I hadn't just gotten a very rejecting followup communication, but I did. To be explicit, I said:
":) sometime this week? if so, times I have are: tonight, Tuesday later afternoon and night, Thursday I am planning on [doing thing], so could meet after that, and then next Sunday night."
he said:
"probably not for a few weeks because I am going out of town and then starting a new job immediately after I get back"
then I said to have a good trip and he told me where he was going and I did not respond.
while MAYBE it is not unreasonable to be worried about being distracted at the beginning of a new job, choosing the interpretation as charitable to the dwindling possibility that he is interested in me as I possibly can, you would have to really generously couch that in an affirmative expression of enthusiastic interest for it to not sound a whole lot like no. The propagation of the species is just going to have to not depend on me right now.
then I said to have a good trip and he told me where he was going and I did not respond.
That seems like a good way to have handled it, unless you really do need a kidney from him. If he's interested but temporarily distracted, he will text back at some point.
I wouldn't invest much more time in him at this point, unless he either actually picks a time to see you, or you see him in other social situations. Even I would think you've been unambiguous about things. I would guess that he's either not interested, or interested but for some reason* doesn't want to pursue that.
*I wouldn't underestimate people's ability to set priorities for themselves that don't seem to make sense to people other than themselves. Maybe he really does see being out of town and starting a new job as reasons to not date (or "date" or whatever) in the immediate future.
I wouldn't invest much more time in him at this point
That's my inclination as well. If he's 19 then this is tolerable behavior but if this is a 30 year old acting this way then this is ridiculous. If he's attached he should tell you or if not then man up and fuck you already.
Dating 19-year-olds sounds like it would solve a lot of problems.
I wouldn't underestimate people's ability to set priorities for themselves that don't seem to make sense to people other than themselves.
I feel like the last three words there imply a serious overestimate of human self-awareness.
I don't think people underestimate people's ability to set priorities that don't seem to make sense.
I assume everybody but me makes sense.
That's not right. I assume that everybody but me can and should be held responsible when they don't make sense. I'm obviously doing my best and thus blameless.
124 explains why he doesn't seem to realise what you want, viz. because you haven't worked out what you want yet yourself.
134: Nobody goes into these things knowing exactly what they want. That's why we call it "dating" rather than "arranged marriage".
Sometimes we change our minds, even.
I never changed my mind. All I wanted was a hot chick who would give me a son.
Rumpelstiltskin was the same way, kind of.
while MAYBE it is not unreasonable to be worried about being distracted at the beginning of a new job, choosing the interpretation as charitable to the dwindling possibility that he is interested in me as I possibly can, you would have to really generously couch that in an affirmative expression of enthusiastic interest for it to not sound a whole lot like no. The propagation of the species is just going to have to not depend on me right now.
Yeah - like everyone else apparently, I have had multiple instances of realizing years later that a woman was interested in me, instances that outnumber the times when I realized at the time - but I never got a missive confirming it as in 91, let alone the later communication. Sorry.
Y'all. (By which I mean, anyone suggesting I take an aggressively explicit approach right now.) Please imagine yourselves speaking to a woman that way in this situation. Imagine all the reasons that would be kind of horrifying. Now imagine I feel the same way! Srsly, I:
I think what brought this about was this part:
bewilderingly (if he liked me) he easily had the opportunity to be the last person at my house last night when we'd all been drinking and then when it was clear that was going to happen hurriedly announced he would go. I was disappointed and sad. I had cleaned my room just in case he wanted to be in it!
This is not explicit enough interest such that you should be disappointed. If he hasn't started the evening envisioning staying over with you as something within the realm of possible scenarios, he won't realize it's a possibility during that same evening. But you showed interest immediately afterwards.
Dissapointing!
One night the other week, I was in a bar. I was eating a late dinner but engaging in a bit of conversation with a couple of people. A woman I'd never seen before who appeared to be about my age joined in, as one does. Then, after I'd finished eating and the other people I was talking with moved on to conversing among themselves, she and I kept talking. The gist of this conversation was that her husband was older and had some debilitating condition and that she needed to stay with him to take care of him but was, because her husband understood that she still had needs, free to date other people.
I pretended I didn't notice the obvious implications of what she was saying for the usual reason (small, but non-trivial odds of her being a serial killer) and continued talking with her about other things until it was time for me to go home.
Of course a serial killer wouldn't care if you were married.
There's also Betty to think about, but possibly being killed is a bigger deal.
The whole thing with shy nerds -- "you need to work on them!" -- but seriously are they worth the effort? I know I wasn't.
eennnnneewaze, I think we're now overanalyzing a situation in which I asked someone out and he said no. The only thing interesting/confusing/ambiguous about the situation for a while was his touch communication and one "sure!" But everything a little bit open or flirtatious that happened happened when he was drunk, and he didn't continue that behavior sober. It's sort of annoying that he couldn't muster a clear yes or no but otoh, given that not finding it within yourself to say yes or no when someone asks you out is not very considerate of their feelings, maybe I don't have to stress that some precious gem is inaccessible to me.
In general shy people are similar to normal people. I don't know if it's "needing to work on them", but they just move slower.
One thing that makes shy people shy is lack of spontaneity. If I'm out somewhere with friends and the plans change because the big fun guy who's organizing the outing decides we should now go to his place, and play video games! I have an intense dichotomy between "I like these people, I like video games" and "This wasn't the plan, I was planning to be home by midnight". This goes double when dating, when you don't want to indicate that you don't want to spend as much time as possible with the other person, because you do! But you start panicking about sudden changes of plans. And "changes of plans" can definitely mean "having sex on the first date".
"Sometimes when we touch, the honesty's too much. I have to close my eyes and hide."
No, my earworm this morning is Lyin' Eyes.
148: I've probably told you all before that this song is actually about my late-great hamster, Dmitri, who had a beautiful mane, and it goes, "You can't hide your lion eyes."
|| I have long since learned to switch stations when Shankar Vedantam comes on Morning Edition to tell us about a breakthrough discovery of human nature based on a single study done on 5 white American college kids who volunteered to get extra credit, but I should just turn off anything related to social or behavioral science, like today's Do the Patterns in Your Past Predict Your Future?
He got a massive trove of data on 5,000 kids who had been followed from the day they were born, then made that information available to data geeks and researchers across the globe. Four hundred teams were given incredibly detailed information about the kids from birth until age nine, then told to predict their grades -- and a handful of other outcomes -- at age 15.
***
the predictions were a lot closer to 0 percent accurate than 100 percent.
Isn't it glaringly obvious that prediction would be awfully difficult without data on students who are already 15? All the algorithms that are ruining our lives are based on data of previous users: if many people who click on link A then click on link B, you have a piece of data to help to predict future behavior of people who click link A.
The researcher says he used Netflix as a model, but they were looking for patterns that were already there.
when you're talking about individual outcomes, there's a lot of randomness
1. Duh.
2. But this stupid study didn't test that hypothesis; it just proved that prediction isn't the same as identifying patterns.
Thank you for letting me rant and get off my lawn.
|>
145
eennnnneewaze, I think we're now overanalyzing a situation in which I asked someone out and he said no.
That's fair. People said upthread there were lots of times they could have got laid and only realized it in hindsight, but personally, there is only one incident I'm sure of. It was our first or second "real" date, there was heavy snow, she invited me to spend the night, I said I thought it would be safe to go home anyway, and I didn't see much of her for a week or two after that. I can be sure that that was a real chance for something more because it's Cassandane. (To be pedantic I don't assume that spending the night == sex, but "something more" doesn't necessarily equal sex either but still would have been nice.) I can think of a few more incidents that might have been missed opportunities, or women something might have developed with even if I can't point to specific incidents, but no time I missed out on a sure thing.
150: Agreed. I thought that was a new pseudonym I had missed. It should be.
If they really did this exercise the way you're reading it, it is indeed monumentally, perplexingly stupid, but I actually have trouble imagining that they did, and I don't think the article makes it clear. For it to have anything in common with Netflix Prize type data analysis, there would have to be training, validation, and test sets of data for kids at 15. I don't think anything the article says clearly states that there weren't such sets; it's just not a very detailed article. Performance would be measured on the test set. I think it's pretty notable that the computer models couldn't do well; I would have predicted otherwise.
Yeah, 144 speaks for me also.
eennnnneewaze, I think we're now overanalyzing a situation in which I asked someone out and he said no.
And yup, that's it. We can't imagine what pathology on his part would lead to this situation, but there it is.
||
NMM2 Efraín Ríos Montt. At long fucking last!
|>
One thing that makes shy people shy is lack of spontaneity
I'm not shy, but I'm likely to say, "Goddamnit, I had all these rigid plans fixed in my head and now you want me to think these things through from scratch? These spontaneous adventures better have a real pay-off or I'm giving the whole evening the side-eye."
151, 153
Here are publications from the project.
https://crcw-new.princeton.edu/publications
The specific result in the interview, a bunch of machine learning models run against this data, doesn't seem to be there anywhere. In other words, the interview, rather than describing what has been published and is useful (of which there's a lot), describes a single in-process effort that hasn't worked yet, and isn't described anywhere.
Science journalism that emphasizes process often seems to me pretty incoherent.
Machine learning in particular is known for producing sensitive correlation detection systems that are weird black boxes, do not translate readily at all into identifying comprehensible facets of data that lead to particular outcomes. Moreover, the number of possible black boxes that can be built is huge, with only hunches to guide what kind to build. It's worth trying this, but I dont see why success is an outcome to particularly expect, much less comprehensible success.
||
This from the Justice Sotomayor this morning:
The majority today exacerbates that troubling asymmetry. Its decision is not just wrong on the law; it also sends an alarming signal to law enforcement officers and the public. It tells officers that they can shoot first and think later, and it tells the public that palpably unreasonable conduct will go unpunished. Because there is nothing right or just under the law about this, I respectfully dissent.
|>
160: Good to have someone on the Supreme Court saying this. Unfortunate that she is the only one.
RBG joined her opinion.
RBG joined the dissent.
The fact pattern in Kisella was bad, and I think if anything Justice Sotomayor understates the harm that will come from this being incorporated into officer training.
||
Doctors: "This drug is super-expensive, has risky side effects, and whether it has any benefit at all is not clearly established."
NYT: "WHY ARE DOCTORS WITHHOLDING VITAL CARE FROM PATIENTS, DO THEY JUST NOT GET IT"
OT: I've been reading the news about "Rep. Esty" for a week or so now and I've only just noticed it isn't "Rep. Etsy." Anyway, she's dropping out and not named after an internet craft store.
|| So, the other evening, my wife was out of town, so I flipped around the dial just to see what there was on local TV. (We have rabbit ears only). I watched some of a feature on the local news about Sokolow v. PLO, sticking because this is an area of law I know something about. I kept waiting for the local tie-in -- surely a victim lived in Montana or something -- but it never came. I looked it up, and saw that cert was pending: ok, this must have been one of those national Sinclair dealios, lobbying the Supreme Court via local TV. Which seems like a dumb idea, although if the point is that our courts elites like the terrorists better than they like "everyday Americans" I suppose it almost makes sense.
Anyway, today, at the top of the cert denied list: Sokolow v. PLO.
(The question is whether US courts can exercise personal jurisdiction over the PLO and the PA when they're accused of complicity in a terrorist attack in Israel, targeted at Israelis but incidentally killing Americans. This turns in part on whether they're sovereigns, which don't have due process rights, or persons, which do.)
|>
Of course the local news story wasn't about personhood or due process or international shoes. It was about how much the relatives of those killed have suffered, how long they've waited for justice, how the jury awarded over $200 million (I don't remember hearing that that got trebled by the district judge), but somehow that got thrown out.
168: "It means he gets results, you stupid chief!"
It's Deep Sate sabotage!
https://pjmedia.com/trending/sokolows-16-year-fight-justice-comes-12-days-short/
No reason mention that the State Department actually would have a position on whether the US has recognized the PA as a state.
PA is a state, just like OH and MT.
1) PA is a commonwealth.
2) I don't recognize it.
It is a commonwealth. I always forget that.
4-slide presentation on why all blockchain stuff is bullshit. This was helpful to me in boiling down the key points about inherent exploitability. I'm going to remember the summary of "smart contracts" as "million dollar bug bounties".
On topic since blockchain is marketed as a weird trick.
164: Thanks for that link. I read the original with interest but didn't have the time to do the homework on the original research.
174: It is bullshit, but you still can buy weed online with Visa.
What about paypal? Like maybe if you knew the guy and were a regular customer.
I don't think the Cannabis Growers and Merchants Cooperative offers that option.
175: Glad it was useful.
176: I feel like it's probably more anonymous, on balance, to go to an ATM and then a dispensary, even considering they check your ID. But maybe you're somewhere they don't have those, I forget.
No dispensaries yet. I have to hold out until July when they open up across the border in MA.
||
When speaking the abbreviation "CE"do you say CE or do you say "Common Era" out loud?
|>
||
When speaking the abbreviation "CE"do you say CE or do you say "Common Era" out loud?
|>
Not that I say it often, but when I do that's how I say it.
Thanks. I usually do the same but this is for a voice-over I'm doing and I'm suddenly self-conscious of that.
When Barry does voice-overs he doesn't like saying "IN A YEAR...." without specifying rigorously exactly which year he's talking about.
I'm with Teo usually, but in a situation like that I might spell it out, because it's still unfamiliar to some people.
"Common Era" seems so dated. I prefer "Basic Era".
188 That was my inclination too. It's for a general audience.
That means you can't call it the "Fucking Common Era."
"In the year of our Lord, Darth Vader"
In sort of a funny coda, I just logged into OkCupid with seriousness of purpose for the first time in a while now that I'm out of this environment that was a fertile ground for meatspace crushes and discovered, looking through my old correspondence, that this guy actually wrote me eight months ago, and I wrote back, and then he eventually just disappeared from the conversation. I guess he's attracted to women.
Static electricity or heterosexuality?
Dead Common Era