Huh. So humans apparently got some evolutionary advantage by talking more about numbers with boys.
If you kill too many lions, the meat goes bad. But you just collect as many wild turnips as you want, sweetie.
As heebie notes, this bit is impressively clueless:
We don't think this is due to some cultural factors. The majority of mothers were upper-middle-class college graduates. That's a group that would be pretty aware of this type of issue and would like to avoid it.
I thought 6 was some kind of joke I wasn't getting. Apparently not.
Seriously, people my age: STOP DYING UNEXPECTEDLY. YOU'RE FREAKING ME OUT.
Is there any downside to 6? Apart from freaking out apo, who is a big boy and can take it...
7: If Breitbart wrote it, it isn't true. Axiomatically.
9 was *not* me. I thought the exclamation point in my 6 was a bit borderline, gloating-wise.
6: Golly.
OP: It's funny, the cliche is that you can have all of the gender-egalitarian ideals you want, but when you have kids you'll see that little boys and girls really are innately different. My experience has been that it's hard to tell about innate differences, but adults certainly do consistently and aggressively treat them differently.
Is there any downside to 6?
Four fatherless children.
little boys and girls really are innately different
Man, no kidding about that.
13. Yeah, that's a downside, alright. Never occurred to me he'd ever had sex.
9: This is unnecessarily unpleasant. As a public figure, I thought he was worthless, but applauding his death is kind of horrible.
The trouble with American liberals is, they don't hate their enemies enough. When a Brit tells you they've got champagne on ice for when Thatcher dies, they mean it.
I've got champagne on ice for Dick Cheney, but my regard for Breitbart doesn't reach the level of disdain where I'd have wished death upon the guy.
I admit that I cheered aloud and poured a glass of wine when Jesse Helms finally died. I still hold that North Carolina is a better place without his foul, hateful breath stinking up our beautiful pine-scented air and would have preferred to have had his body unceremoniously dumped at sea bin Laden-style.
I'm terrible at actually talking to our toddlers (both girls), but I lead them on recitations of the numbers all the time. Yay for me?
Also, apo's "downside" was 4 fatherless children, which, in point of fact, is a downside. His absence from US politics? A boon to us all.
My son is always trying to get me to count to a million.
I do wonder if there's a story. Sudden natural death for a 42 year old really is odd.
From what I can gather, Breitbart died "unexpectedly", "from natural causes", at UCLA Medical Center. At just 43, I guess that would make aneurysm or heart attack the most likely causes.
Thanks for posting, Heebie! I hope someone is able to access the full journal article at some point.
One thing I'm thinking about is how much parents' attentiveness to their children's interests might create a feedback loop. If you notice that your child pays a lot of attention, or asks more questions, in response to certain topics, you might consciously or unconsciously start bringing those topics up more often.
Why armchair neuropsychologizing is any more odious than armchair astrophysics escapes me. People like trying to make sense of their world and neuropsychology provides some insight into a particularly interesting corner of it. Of course non-experts get important parts of it wrong. Getting stuff wrong is part of learning new things.
that would make aneurysm or heart attack the most likely causes
No lie.
On the second study from the OP, the news is even worse. Not only do folks casually flip back and forth between 'brain' and 'mind', adding superfluous neuro-jargon makes explanations *seem* more compelling, as Deena Skolnick Weisberg demonstrated:
The seductive allure of neuroscience explanations (PDF)
Caveat lector: The presentation of neuroscience information in the popular media (PDF)
The funny thing about all this is that there's now both research claiming that we're 'intuitive dualists' and other work claiming we're 'naive neuro-reductionists'. Knowing us, we're probably both.
25: And of course it's a two-way feedback loop. If the parent lights up when the kid shows interest in something, kids are adult-oleaster and will work to get the same reaction.
When a Brit tells you they've got champagne on ice for when Thatcher dies, they mean it.
The only reason I don't is that my fridge is too small. It's in a cupboard. I reckon I'll have enough warning to chill it before the big day.
Auto correct turns "pleasers" into "oleaster". What does oleaster even mean?
Wild olives, apparently. New word for the day!
My armchair hypothesis is that most on line bullshitters understand at some fundamental level that they actually know jack shit about astrophysics. With neuropsychology, which is, as a matter of cold, hard fact, a less mature science (for now), they don't think this matters.
I read it as "kids are adult-molester" and wondered what was going in the LB household.
Breitbart was a good friend of one of my good friends (not a rightwing monster) so that's another loss.
adding superfluous neuro-jargon makes explanations *seem* more compelling, as Deena Skolnick Weisberg demonstrated
I just realized I was about to make the same joke that heebie makes in the post itself.
37: You should do it anyway. It's good enough to repeat. *looks expectantly at nosflow*
If anyone deserved to die, it was Andrew Breitbart. I hope he rots in hell for all eternity, and that vultures feast on his flesh. If Shirley Sharrod and the employees of ACORN all announce that they are sad that he's dead. Until then, fuck him. He belongs on the ninth circle of Hell, with everyone else who has betrayed humanity's chances for a better future.
26, 35: The other problem with armchair neuropsycholgizing is that it tends to limit human potential -- a fundamental type of pop neuropsych argument is that this or that sort of behavior is unchangeable because it's just how the brain works. This seems pernicious to me on a different level than anything involving lay confusion about astrophysics is likely to be.
41: Boy, is this true. Moreover, the idea that if something is part of how the brain 'is wired' then it must be innate/unchangeable also coexists with the contradictory idea that certain things *can* change the brain, but only in suspect ways. See Nicholas Carr's horrible arguments about how 'the Internet is changing our braiiiiiins'. Brain-changing always has sinister associations. Pity everything does it.
Does Man Suit get a Fruit Basket, or have I not been paying attention?
My experience of being a non-gender-standard kid is that a plurality of the adults will seem to completely not notice, a few will actually react to what the kids in front of them are good at, and a constant peppering will react with dismay or disdain or disgust and try to 'fix' the kid. Unthinking disgust is really hard not to be hurt by.
I have read that if you package the same infant in different-gendering clothes, & hand it to the same adults, most adults will treat the babe & describe its reactions gender-normatively. (Original research, mm, late 80s?)
||
Rain! yay! my wetlands might get wet! Moderately unpleasant field day & I'm going to *hate* driving there. Californian drivers are weak on 'safe following distance' to start with, & react to rain as in my 1st para here.
|>
Welcome Wagon brings fresh fruit, Man Suit.
Also, sounds like a heart attack.
Breitbart was walking near his house in the Brentwood neighborhood shortly after midnight Thursday when he collapsed, his father-in-law, Orson Bean, said. Someone saw him fall and called paramedics, who tried to revive him. They rushed him to the emergency room at Ronald Reagan UCLA Medical Center in Los Angeles, Bean said. Breitbart had suffered heart problems a year earlier, but Bean said he could not pinpoint what happened.
41,42: I really want to argue the other side of this (mostly because it's pitiful when everyone agrees), but there's a point where neurobollocks joins on-the-veldt as simply a justification for whatever the hell our current preferences are. On the other hand, actually understanding how our brains work is empowering if we get it right. That's the big draw for me, at least.
I think if anything is hard-wired, it's the desire that the genders be different. Working hard at finding gender differences seems like a vigorous activity that people of all cultures engage in. I'm even mostly not kidding about this.
I've got champagne on ice for Dick Cheney,
Who is now out there campaigning for gay marriage, or something to that effect.
47: Seriously, this I'd buy. The gender-related behavior I see in kids that looks innate to me (and I really could be wrong about this too) is looking for gender differences to sort by: it looks as though they want 'boys are like this, girls are like that' even where the 'this' and the 'that' are unimportant in themselves.
Let alone how intense adults get about it.
45: Oh, how kind you really shouldn't have. I'm pretty sure I'm allergic to... that fruit. But these little cheeses look delicious.
Breitbart was an excitable fellow. That can come back to bite you right in the heart.
Five year old was at Target to pick out new underwear and chose girls' because he likes littlest pet shop, which is not something that strikes me as terribly gendered but apparently it's marketed as a specific girl thing. My only concern was whether he would think it was uncomfortable, but no complaints so far.
26: It also seems to be the case that when an astrophysicist shows up, people assume s/he knows more about astrophysics than other, non-astrophysicist people. Whereas a lot of people have read a book or two about the brain or, worse yet, think they can intuit their way to certain understandings, and don't grant actual neuropsychologists as much relative expertise.
Well come on Blume, not everyone has, like, a telescope or a galaxy of their own, but everyone has a brain.
54: True. OTOH you see experts (in a range of fields) getting extraordinarily worked up over minor errors when the non-expert has the big picture roughly correct.
I dunno, my kid has always had a distinct preference for the girly despite my strenuous efforts to the contrary; my experience really has been that there's a pretty inherent gender thing going on. Can we just say "kids (and people) are different" and leave it at that.
I wonder if he's going to resign. He should.
52: The liberal conspirators are finally getting their act together. About fuckin' time!
46: Yeah, my brain lights up whenever there's a solid explanation. Doesn't happen all that often though.
To be fair, 40% of Montana judges have masturbated dogs, so we should probably give him the benefit of the doubt on this one.
Also, Obama had Breitbart whacked.
It's funny, I really do respect judges. The story I linked in 58 shocked me -- no reason why it should have, I know people suck, but I was honestly stunned at it coming from a federal judge.
I'm all for refraining from speaking ill of Breibart until his corpse has fully chilled, but perhaps news organizations (such as, say, NPR) should refrain from reporting on his career if they're unwilling to mention the viciousness and lying, expecially if at the same time they're inviting his pals to eulogize him.
47, 49: Sometime in the late 1990s a survey was done of high-energy
physicists asking why they included so few women. Almost everyone agreed
that it was due to innate gender differences. However, Anglo
physicists knew that women are naturally social, group workers, not
the solitary thinkers that HEP requires. Japanese physicists knew
that women are naturally solitary, not the cooperative team thinkers
that HEP requires.
It isn't just gender, of course; Racialicious has a polyphony of
descriptions of what we expect of members of different races (and of
how the expectations differ with gender and sometimes class, and of
the pushback you get for not conforming). But we know those categories
can change: the Irish are
practically normal by now -- I startled one a year ago by pointing out
that ethnic Irish Catholics are not, technically, WASPs, but I think he was
right that on the West Coast for most purposes they are. (Other wild
research: a cradle-tongue that distinguishes between snug and loose
fits develops children who are much more conscious of that distinction
in toys.)
I think that people like categorizing, and that we *really really
like* -- to the point of psychological need and game-theoretic profit -- to categorize
hierarchical distinctions. As soon as there are hierarchical
distinctions, we harden the categories: any trait *must* be one thing
or the other, because otherwise we might observe someone who can't be
categorized (CHEATING! CONFUSING!).
Halford, the last time we had a discussion like this, I remember you
continuing to claim that there weren't any techie women despite my
testimony to the contrary.
47 and to both parts of the OP, I haven't said good things about Lise Eliot's Pink Brain Blue Brain in a while, but it does a very good job of arguing that while there are some (very modest) neural gender differences these are swamped by cultural factors. Also, she's mean about Simon Baron-Cohen's lab.
I am reading 65 in what I imagine to be clew's poetry slam voice. Actually, it just morphed to Laurie Anderson doing "O Superman".
All the news outlets reporting Breitbart's death checked directly with the LA coroner, even though it was on Breitbart's website several hours earlier. If people think he'd scam the world in reporting his own death, that says a lot about how respected he was.
I agree strongly with 65. Having heard many, many folk theories from people of all educational backgrounds who were raised all over the world, both the categorization impulse and the "just-so stories" impulse seem robust across cultures.
What does seem to differ somewhat is the degree to which people feel personally threatened by others whose categories they can't easily pinpoint, or who seem to be straddling categories.
I have a totally seat-of-the-pants theory that acquiring this kind of comfort this is a developmental skill, and that people who grow up in settings where their mentors/elders actively acknowledge the malleability of categories and model nondefensiveness are more likely to develop the skill.
The trouble with American liberals is, they don't hate their enemies enough. When a Brit tells you they've got champagne on ice for when Thatcher dies, they mean it.
I don't actually have any on ice, but I'll try to beat the rush to the shops. Hate, though, definitely.
I thought I'd never hate any politician the way I hate(d) Thatcher, but the current lot are doing a good job of stoking the fires.
65 last: I don't know what the fuck you are talking about. Of course there are women who are "techies"; I am related to some of them, and I think there should be more of them. Could you stop casting around random aspersions please, thanks.
65.last, 72: I don't recall that one either -- clew: any hope of some keywords to look for the conversation you're thinking of?
All the news outlets reporting Breitbart's death checked directly with the LA coroner
The first one I found was Fox News, which checked with his lawyer.
Girls submit psychological guest posts about child-raring, boys submit them about abstract armchair theorizing.
One striking thing about the Judge Cebull news stories is how unsupportive of him people have been in comments. Even in his hometown paper (Billings Gazette, not GF Tribune) there is more 'he must go' than you'd expect. You know, given that newspaper comments are a cesspool, and Billings is plenty red.
The first one I found was Fox News
I believe I specified "news outlets."
Witt (and anyone else interested), if you still don't have the full-text article, feel free to email me at the linked address. Not my field by a long shot, but it doesn't look like there's much to it that wasn't captured in the news piece—they pulled samples from a corpus, counted occurrences of number-speech, and did t-tests. (They don't show the distributions of counts across samples, so we have no way to judge whether the t-test was an appropriate choice.) The end.
76: There's not much defensible to say about "Your mother fucks dogs." What makes the joke racist is the implicit 'if she'd fuck a black man, she'd fuck a dog,' but even someone determined to overlook or approve of the racist aspect is left with a pretty explicitly horrible joke.
(Looking. We talk about gender a *lot*, and Fukushima brings up reactors. )
20: I'm terrible at actually talking to our toddlers (both girls), but I lead them on recitations of the numbers all the time. Yay for me?
Definitely yay for you. One of my warmest early memories of my relationship with my dad was a "speed game", as we called it -- this would have been when was I was in 1st grade maybe? so, older than a toddler -- he'd call out "2+2!" and as fast as I could I'd answer "4!"
3+3!
6!
4+4!
8!
4+3!
7!!
7+5!
12!!!
I'd get so excited. Let's play the speed game, daddy! Do that stuff with your girl kids, folks! Ahem. I got excited there for a minute.
65.1: Perhaps the Japanese men should work with the American women. And the Japanese women should try to focus on the aspects of HEP that the American men are looking at.
|| News story about the death of the stepfather of one of my son's best friends. Hell of a thing.|>
54, then 55: not everyone has, like, a telescope or a galaxy of their own, but everyone has a brain.
Likewise, everyone has a philosophy. Or can read poetry and novels. Numerous fields suffer through undergo this type of armchair weighing-in.
81: Your dad may have done a good job inculcating an affection for figures, but he apparently did a lousy job teaching you factorials.
But seriously, 81 is great and made me smile.
(This being the death thread, I should say that my FIL seems to be about as well off as he could be with a cancer diagnosis. As far as they can tell, it hasn't spread -- surgery very soon, and then either no or minimal chemo.)
This being the death thread
"A theme park in London is set to debut a powerful new winged roller coaster this month, but only after first figuring out how not to dismember its riders. [...] The ride's designers ran some test runs with crash test dummies, leaving many shocked when the dummies returned from the experience missing arms and legs."
82: The beliefs are about men and women in toto, and also the national imagery makes it worse: Japanese men believe American women are *especially* bad at teamwork (pers. comm.). Someone else can weigh in on what American men believe about Japanese women.
Someone else can weigh in on what American men believe about Japanese women.
47 is something I say a lot, and 57 is something I hear a lot, probably because it's true for (something like) the majority of cases in which parents wish to neutralize their kids' gender preferences. (cis-gender preferences?) So yeah, that's often true, but there really *are* little girls and little boys who don't go for their traditional accoutrements, and I assume that some of those kids also appear to go along with their parents' efforts to get them to be more balanced.
And yeah, the Lise Eliiot book seems pretty good, although it's hard to find the time to read about small child development while managing a developing small child.
85, 86: Yeah, in retrospect I think it was brilliant: he was teaching me patterns, but also teaching that I must listen and understand, and not get hung up on the patterns. (As I kid I registered this as: daddy is going to try to fool you soon, stay alert!)
I can't say he taught me html very well, as the answers to 4+3 and then 7+5 were supposed to be preceded by a puzzled {pause} on my part before answering.
89: It sounds like maybe women are bad at teamwork and at working alone, but the men only notice deficiencies relevant to their own working style.
Which is a little surprising to me, since in my experience women are not noticeably worse than men at either of those things.
Alternately, there is some other reason why women tend not to work in HEP, and the men are just confabulating reasons why.
93.last is a possibility.
63: I think that people like categorizing, and that we *really really like* -- to the point of psychological need and game-theoretic profit -- to categorize hierarchical distinctions. As soon as there are hierarchical distinctions, we harden the categories: any trait *must* be one thing
or the other, because otherwise we might observe someone who can't be categorized (CHEATING! CONFUSING!).
I'm not sure why the distinctions are necessarily hierarchical, but otherwise, sure.
That is, the distinctions we're most familiar with (gender and race, with sexual preference being a disappeared class) have almost always sorted out to a hierarchical structure, but I'm not sure they started out that way. There is a narrative involved, after all, according to which these ones are privileged and those ones are not; I'm not sure it's built into the distinction in the first place.
I'm tempted to blame capitalism, though that is glib; nonetheless, as time has gone on, it's absolutely essential to construct a narrative according to which some categories of persons are better or best suited to what needs to be done than others.
Is there a good basic primer on game theory that would explain things for people like me who have no head for mathematics and like words instead?
95: game theory per se? Or game theory in some context?
95: Not really a "basic primer", but "Strategy of Conflict" by Schelling is a classic.
You can preview it here and decide if it's worth buying or librarying.
88: Good to see Julijonas Urbonas getting work.
99 was to 84. I guess I html'd my reference away somehow. It was there!
I want to a see a link for 65.1. HEP experiment has a reasonably large fraction of women, as physics goes. HEP theory does not, and if those were the people they were polling, I probably know a lot of the people who were polled and am kind of horrified that they would say those things.
On the other hand, my impression of the gender politics of my colleagues has declined a bit in the last couple of weeks....
I'm tempted to blame capitalism, though that is glib
There was a pro-situ meme 40+ years ago that whenever somebody said "That's life", they meant "That's capitalism". Glib, sure, but a strong element of truth.
87: That is good news, or at least, the best news you can hear about someone with cancer.
Since the research project identified in 80 has now lasted several hours, clew, would you now provide a retraction of the claim that I am going around saying that women can't do science or don't have technical aptitude? Asking for apologies on the internet is surely a fool's game, but I really don't like being falsely labeled a rank sexist based on your imagination. Thanks.
Even aside from the gender politics, I can't really imagine American physicists talking about "the solitary thinkers that HEP requires". We are a highly collaborative and gossipy people, after all.
96: Well, context is good, but I'm thinking more "idiot's guide" except not that or "for dummies". Just, you know, popular writing, about game theory, that explains stuff without forcing me to revisit too much heavy duty mathematics.
99, 100 should provide a commenter name.
97: You can preview it here and decide if it's worth buying or librarying.
Consider bookfinder dot com if you choose to buy.
107: That was me. I know, even your kid could comment better than that.
I liked Pink Brain Blue Brain. Which I probably read on the recommendation of a comment here. It seemed like a reasonably skeptical/empirical take on things.
Nat, do you want a two minute explanation, or do you actually want to work through some game theory style problems?
110: Well, I feel like I've already got about 90 seconds of the two minute explanation, not that I'd want to try to write it down in a blue exam book right this second. I guess more of an overview that would talk about what the classical problems are, why they are interesting, how they were developed, etc. etc.
As best I can tell, game theory means:
1. Assume all people are brilliant mathematicians that work out all possible logical implications instantly.
2. Assume people want to maximize their money, and nothing else in life. If someone criticizes that premise, substitute in "utility" for money. Define "utility" to be "whatever people are actually trying to maximize". Then act as though utility=money.
3. Design logic problems, where characters have to obey weird rules in order to get money. The game is to figure out what your characters will do, since they are infinitely brilliant and want to maximize their utility. The outcome to the game ought to be that the characters acted counter-intuitively. Imply that your players had to act this way, through sheer force of logic. Bonus points if your players were forced to act like assholes.
Here are two painters whose work I like
http://fuckyeahbeksinski.tumblr.com/
http://www.ericpenington.net/art.html
I like my kid's art, but he can't do that.
It also seems to be the case that when an astrophysicist shows up, people assume s/he knows more about astrophysics than other, non-astrophysicist people. Whereas a lot of people have read a book or two about the brain or, worse yet, think they can intuit their way to certain understandings, and don't grant actual neuropsychologists as much relative expertise.
But people *can* intuit their way to lots and lots of understandings about the brain, because people have a lifetime of very personal experience with brains. The issue is the appropriate bounds of the neuropsychologists expertise. A tremendous amount depends on the specific question being asked. What are the appropriate bounds of scientific authority when dealing with a vastly complex system in which only particular small elements are really understood? Where there is an enormous amount of practical common experience with that system that has been gathered outside of the scientific process?
I suspect 'neurobollocks' is more about an illegitimate claim to scientific authority (used to shut down the conversation) than an illegitimate claim to expertise about the brain, broadly construed. In that sense the inappropriate granting of scientific authority when the trappings of science are inserted into the conversation (neurons are changing! brain scans show!) is the problem than the failure to privilege neuropsychologists.
Here's a starter example:
There are 100 Hershey kisses on a table, and two players. The players will take turns.
On a turn, a player has two choices:
1. Take one Hershey's kiss.
2. Take two Hershey's kisses.
At any point, if a player takes 2, the game is over, and the remaining kisses are thrown in the trash can.
What is the first player's move?
(I'm happy to provide a solution, and was about to type it up, but a student just showed up. It will come unless you want me to hold off.)
I need a neuropsychologist to explain all my screwed-up sentences. Last sentence should stop after 'problem'
Heh, there actually is The Complete Idiot's Guide to Game Theory. And nothing beyond the title that makes it look like it would necessarily be no good.
112: That sort of jibes with how I've thought about stuff, although it even somewhat more dismissive than I would have been.
Part of what's prompting this is an insight I had the other day, based on the copies of Psychology Today that hang out in my psychologist's shared waiting room. Now, you know and I know that at least 90% of what's written in that magazine is bullshit. But the thing is, it's sort of the bullshit of the zeitgeist, if you will. What if, know that it was all bullshit, you still read every issue of PT cover-to-cover and followed all of their inane little bromides? Wouldn't you probably be a much healthier, happier, more successful person? Sorta like with Dale Carnegie or whatever. The world runs on bullshit, so if you're constantly seeking truth and meaning, you're always going to be disappointed and end up a shriveled old curmudgeon, fight with imaginary people on the internet.
116: Are these the kisses with the caramel inside?
120: It depends. What's the marginal utility associated with a caramel center?
112 sounds about right, which is why reading Kahneman's new book is so fun.
If not, take two. Noöne should have to eat the regular kind.
119: Hmm, I seem to have forgotten how to type correctly in English. Perhaps I have a brain cloud.
But people *can* intuit their way to lots and lots of understandings about the brain, because people have a lifetime of very personal experience with brains
Whereas most people grew up in an entirely different universe, and thus are unable to talk meaningfully about astrophysics in this one.
I like the insight that to solve most game theory problems you just have to make the choice that maximizes your assholishness.
Also, I read this post as Cognitive Muppet Post.
So, the Hershey's kiss problem is a "caterpillar" problem, meaning assume the players are normal, non-game theory humans, and then watch how it unravels backwards to an asinine conclusion.
So: If normal people were playing, they'd each take 1, so that the game would keep going. In this case, suppose you're about to take the 99th hershey's kiss. Might as well take the 100th, too, right? Doesn't matter if you take 2, if the game is ending anyway.
Therefore: the other guy, on the 98th turn, will think "I know that asshole is planning on taking two. So I'm not going to get my last piece. Might as well take two now, and let the last one get thrown out."
Therefore: On the 97th turn, you think "I know that asshole is over there, being smart. He's going to end the game. I might as well end it now and get at least one extra kiss."
And so on. It unravels all the way to the beginning.
So the rational answer is that the first player takes two, and the game is over. The power of logic.
Natilo, 119 sounds like you might not be thinking clearly.
116: Is the goal to win (get more kisses than the other player) or to get as many kisses as possible? Or is part of the problem determining whether it's better to get very little candy, but have more than the other person, or to get a lot of candy and risk that the other person will have still more?
Is the goal to win (get more kisses than the other player) or to get as many kisses as possible?
The goal is just to maximize your own kisses. You don't care about screwing over the other guy. Yet you're not allowed to discover the power of cooperation. Because that would be illogical, or not the right kind of brilliant, or something.
Anyhow, no, it's not true that people have lots and lots of understandings about the brain. Most people have never even seen one in person.
132: Well, even assuming complete selfishness, you might be more interested in eating a lot of chocolate than winning the game.
112: There are, however, aspects of game theory that do not require that kind of foresight, but can be applied to "mindless" processes in the appropriate context. See Evolutionarily Stable Strategies* for example.
*Big Caution: This way can lead to an even more questionable, sticky morass if you start using it in Ev Psych type ways. Stay away from the peeps and stick to the dumb, brutish animals and plants. (Advice which is the moral equivalent of read the first n chapters of Wilson's Sociobiology, but stop when you get to humans.)
Heebie should definitely read her some Kahneman. You'd love him, H-G!
128: I hate that style of logic problem -- I can follow the argument when explained, but I just don't go there easily. I go straight to "If I take two, the most I could get is two. If I take one, I could get up to fifty, depending on the other guy. Who knows what he's going to do, but clearly it's worth giving up one bird in the hand for forty-nine in the bush."
"It is not enough that I should win. All others must lose."
To model altruism, or cooperation, the game theorist needs to be able to quantify the utility involved. Also, the utility can't be mushy, like "hey, we're getting along!"
Also, the utility can't be mushy, like "hey, we're getting along!"
There are those who rigorously model this.
112 is a little dismissive. It's probably worth your time to work through the Prisoner's Dilemma and Ultimatum games logically and then try to figure out why people don't behave in the ways that simple logic seems to dictate.
Knock over the table, declare that eating Hershey's kisses make you weak, slap your opponent on the head with a blackjack and then burn down his village. Game over.
112 is flippant, true. And Prisoner's Dilemma is really useful for illustrating the general free rider problems in a democracy. I don't know the Ultimatum game.
141 is probably correct. The ways that people deviate from strict rationality in those two games (and maybe the Dictator game) is, like, three quarters of the decisionmaking literature.
There are those who rigorously model this.
It would be stupid not to, I imagine.
It's probably worth your time to work through the Prisoner's Dilemma and Ultimatum games logically and then try to figure out why people don't behave in the ways that simple logic seems to dictate.
Well, a lot of the basic games break down if (a) the parties can communicate and (b) they have the option of trusting each other. Given that most actual interactions with other people involve communication and some degree of trust, figuring out why people don't behave in accordance with very simple game theoretical problems is pretty easy.
Well, it's not necessarily easy to do in a useful way.
Anyhow, no, it's not true that people have lots and lots of understandings about the brain. Most people have never even seen one in person.
How far do you want to push the mind/brain distinction you are making? If most people have no understanding of the brain because they haven't seen one, is it also true that the neuropsychologist has no understanding of the mind or mental processes because he just studies a pile of greyish sludge?
Also, the caterpillar problem has more interesting variations. Suppose the players have the following two moves:
1. Take 1 hershey's kiss
2. Take 1 hershey's kiss and that $50 bill, next to the kisses.
Whoever takes the $50 bill, wins the game.
Now, instead of being unreasonable, everyone would end the game on their first time.
146: but why do interactions with people involve trust? When does that trust break down? Is communication a sufficient condition? You can't really think about those questions without having a basic interaction to work from.
142: The first rule of irritating "logic" problems is to know whether you are involved in a lateral thinking exercise, a game theory exercise, a symbolic logic exercise or taking the LSAT.
146: oh yeah, sure, people can just trust each other, because people are just naturally trustworthy and no one's ever been stabbed in the back ever, except in stupid math puzzles. Pretty trivial stuff.
I haven't read Kahneman's latest, but I enjoyed some of his earlier stuff. People are reliably odd.
148: Yeah, I don't know anything formal about modeling, but my sense is that this is the sort of thing that defies it. I don't know whether there's any point in getting into questions about why we try to model every thing. It seems kind of control-freaky.
I don't really know anything about astrophysics, but it seems like a very messy and difficult subject. Like trying to study a brain only through its indirect effects on other brains. Whereas anyone can see that the ideal way to study a brain is to slam two of them together at high speeds and then inspect the splatter that comes out.
Have you seen, A Beautiful Mind, Natilo? The mathematician-hero uses game theory to figure out how he and his buddies can pick up women.
once you have genuinely repeated games (where it is uncertain what the last game will be), anything goes in game theory, even including non-assholish behavior.
I guess when I think about useful game theory, I'm imagining what Ellsberg & co. were doing at RAND -- creating huge simulations to model the effects of various input options. Like in Hearts & Minds, where the one general or national security apparatchik is talking about the Ho Chi Minh Trail, and how the maximum the US was able to interdict of its traffic was about 20%. So basically, he says, my opponent, the NVA/NLF general, just has to order 125% of whatever he needs, and he'll be fully supplied for his operation.
152: Those are all interesting questions, but I'm not sure that working through no-communication/no-trust game theory problems is a necessary prerequisite to thinking about them. I suppose if you needed to convince yourself that communication and trust were necessary for most normal interactions, the game theory would show you that they must be, because if they weren't present people would behave weirdly.
159: There are two kinds of Russell Crowe movies: The ones I will watch and A Beautiful Mind.
152: 146: but why do interactions with people involve trust? When does that trust break down? Is communication a sufficient condition? You can't really think about those questions without having a basic interaction to work from.
What? There are, and have been for decades if not centuries, numerous areas of endeavor involved in answering these kinds of questions. No one area of inquiry gives a privileged answer: anthropology might consider one set of things, philosophy another, literature another, psychology another, and so on along the line.
Then you have a 50-50 chance of watching A Beautiful Mind.
I'm sounding dismissive here (a change from my normal tone, clearly). I just mean that I don't think there are a whole lot of insights to pick up from working through very basic game theory, unless you're going to go on from there to much less basic game theory. If you're not going to do much serious with it, I think "Yada yada Prisoner's Dilemma" is about all you need.
166: oh I mean I wouldn't recommend spending terribly long on them. But, you know, putting in a couple of hours here or there.
163: Oh. I think A Beautiful Mind is the only Russell Crowe movie I've seen.
Well, then have you seen War Games? That has game theory, too. The computer determines that only way to win a nuclear war is not to play.
Oops! I gave away the ending! : (
Prisoner's dilemma single-play and prisoner's dilemma iterated, considered together, add up to nothing useful, IMO. Sometimes people work together to good effect based on enlightened self-interest, and sometimes everything falls apart - what an insight!
I suppose knowing the types of game other than prisoner's dilemma and what patterns those can result in can be useful (there's the playing-chicken one, for example), but pretty often the game has no stable equilibrium, so any number of outcomes could happen.
Technically, the computer just took two hershey's kisses before the nuclear war ever started.
The only point of caterpillar problems is to illustrate the point that in the complete absence of trustworthy communication, there are plenty of situations in which the utility-maximizing move for either party at any time is to do something that quite obviously dramatically lowers the overall utility that could otherwise be achievable for both parties with a bit of coordination. It doesn't seem like a useless way to do that.
171: Women mathematicians! For them, everything's always about the chocolate!
The jokes on WOPR, though, because like a dog it's allergic to chocolate.
168: Okay, maybe I did mean to be dismissive. I'm thinking of a Law and Economics class I took at NYU, which actually wasn't Law and Economics in the pejorative policy sense at all -- the professor was interested in game theory, and it was all game theory. And we spent a semester doing problems, with math, that got actually kind of complicated and were fun. But spending a semester doing problems never got past "So, why do we get this counterintuitive result?" "Because the assumptions have nothing to do with normal human behavior."
The class on Bayes' rule was useful, but the rest of it was finger exercises -- to get to anything that gave you much insight into human behavior, I think you'd need more than a couple of hours.
173: So, we should use them to teach children and bankers the importance of sharing?
177: Right, the most important result demonstrated by game theory is the vital importance of never letting anyone grow up to be an amoral calculating utility maximizer.
"Homo Economicus" is the formal term. They're evil sons of bitches.
112 is just the right amount dismissive.
The assumption in game theory is that trust, altruism and communication are things that need to be explained against the universal default value of egoism.
But this is the exact reverse of how humans develop. We grow up in a family environment that, no matter what culture you are in, features huge amounts of trust, altruism and communication. You can also develop pretty good theories of how things like distrust and selfishness arise in this environment, but this involves talking about feelings, and is like psychology or being a social worker, not the SCIENCE that they do in game theory.
Luckily, the pure breed isn't very successful. Normal people walk away from Heebie's caterpillar problem with better than forty Hershey's kisses. H.E. gets two or zero.
176: I mean, I think game theory is actually very useful for modelling real world (economic) events within the framework of classical rationality, and a surprisingly large class of real world problems can be seen as "modified prisoner's dilemma" or "modified ultimatum game" or whatever. And on an aggregate level they probably work just about as well analytically as any of the rest of classical economics.
180.last: And we're back to talking about Debt.
The assumption in game theory is that trust, altruism and communication are things that need to be explained against the universal default value of egoism.
Most animal species are dicks, dude. Just a-holes.
One time I read a book that was sort of about game theory, and sort of about economics (in its broadest sense) that had a lot of interesting things about "suppose you have this system with these parameters and this number of people and the people each have these goals. What would the outcome be if everyone acted in a strictly rational way to achieve their goals?"
The example I particularly remember was about a community with X number of People Type 1 and Y number of People Type 2, and they each value both living near people of their own type and people of the other type, but they prefer not to be the minority of immediate neighbors, and each person has 8 neighbors, blah blah blah, how will they all sort themselves out? (the result was a totally segregated neighborhood even though everybody had diversity as an explicit goal).
Plus there was some discussion of real-world traffic studies.
This comment would probably be a lot more interesting if I could remember what the book was. I emailed my dad, I'll keep you all posted as I'm sure you are on the edges of your seats with bated breaths of anticipation to find out...
Game theory has a lot of useful applications, not the least of which is modelling competitive behaviors among oligopolist firms (which can't, or at least legally aren't supposed to, communicate with one another in order to coordinate their actions, because of antitrust laws).
186 cont.: (And which otherwise meet the game theoretical assumptions fairly well.)
modelling competitive behaviors among oligopolist firms assholes.
182: Oh, this is probably right. I guess, when I've run into game theory that struck me as interesting or useful at a level I could comprehend, it didn't require a whole lot of background knowledge: that semester of doing game theory finger exercises didn't make No One Makes You Shop At WalMart easier to understand -- a useful argument is pretty clear even without spending a lot of time on Game Theory 101 first.
Come to think, Natilo, you might pick up a copy of the linked book. I liked it.
If you have a table around which are seated Adidas, Nike, Puma, and Reebok (the corporations themselves, not individual human representatives of the corporations), and you put on the table 100 Hershey's kisses, $50, a flask of bourbon and a handgun...
I still think Natilo should read Kahneman's book. I tell you what Natilo, you pick one but don't tell us, and LB and I will decide independently how much it's worth to us that you picked the book we recommended. Then, whoever's book you picked will give 70% of the amount they decided to you, and the other will give 30% of the amount they decided to you.
There's something to be said for the notion that we're being increasingly driven to behave as though we're members of homo economicus -- we'd be more predictable that way, and modeling would work better.
And then we'll all nuke someplace together.
Puzzlingly, I think 190 was kind of pwned.
I hate Hershey's kisses, so I'm auctioning my turn. Anybody can bid whatever they like - money, booze, sex, decent chocolate, I don't care - but in a sealed envelope. I shall be drunk when deciding the winning bid.
Puzzlingly, I type "190" instead of "191" in 194.
Puzzlingly, I typed "type" instead of "typed" in 196.
185 sounds like Schelling's Micromotives and Macrobehavior,.
192: I've encountered a disturbing number of people who think econ 101 is the font of all morality.
I think it is! The cover displayed on Amazon looks familiar to me anyway. Good detective skills Walt!
when I've run into game theory that struck me as interesting or useful at a level I could comprehend, it didn't require a whole lot of background knowledge: that semester of doing game theory finger exercises didn't make No One Makes You Shop At WalMart easier to understand -- a useful argument is pretty clear even without spending a lot of time on Game Theory 101 first.
This seems like a really weird basis for dismissal. I mean, it's probably largely true, but having exposure to game theory is what allows insights like those in No One Makes You Shop At Walmart in the first place. (I assume--I'm haven't read the book, so I'm trusting you on this. I didn't even know it was game theory.) It's like claiming that formally studying biology is silly because it's not needed to understand E.O. Wilson's books.
I guess that's not really about game theory. I liked reading it though.
See 166. Natilo was asking for a non-mathematical "Game Theory For Dummies". I don't think there's much point to it -- the stuff he'd be exposed to at that level will mostly be subject to Heebie's criticisms in 112. Interesting in terms of finding out what game theory arguments look like, maybe, but not really interesting in terms of content.
There are game-theory sorts of arguments that are interesting and useful, but I don't think they're particularly hard to understand even without a prior exposure to Game Theory 101, so the prior exposure to GT 101 is a waste of time if it's not what you were going to do for fun anyway.
Not dismissing game theory in its entirety, just dismissing 'spending a couple of hours on it' as a useful thing to do other than for amusement because you like logic problems.
Oh well, I guess I'll never be smart then.
Oh, that makes sense. I thought 176 was intended to be more comprehensively dismissive than that.
I was about to post that Mara was playing a counting game in Monkey Preschool Lunchbox (thank you, heebie!) but then she switched to "What Did Snakey Eat?" where you're basically matching the shape of the object inside the snake with a picture of the object. I suppose it's a vocabulary building game too, but I'm not sure how I feel about it for a kid with pica. I mean, it reinforces that she shouldn't eat a hammer, but nothing but vigilance keeps her from eating her raisin boxes and hair that she oulls out and so on. Sciency commenter advice would be welcome there.
Mara came to us with a strong preference for boys' clothes. Now she's super careful about sorting everything into boy vs. girl categories, which is partly something she picked up from Val and Alex and partly something she'd have gotten from her classmates anyway. She insisted that the blue hair clips we bought are for boys and the red for girls because it's all about color, not about how many boys her age she's seen with flowers in their hair. She gets lots of lectures about how it's very rare that something is in fact a boy thing or a girl thing (sorry about that fruit hanging way down there) but I think she basically rolls her eyes at me.
There was someone on Fresh Air yesterday talking about brain plasticity and the idea that what's built by age 3 is the framework you have for the rest of your life, which creeps me out since we got one full with Mara in our custody before she turned three and I know I'm being sort of ridiculous but I really do think a lot about how her early life experiences may impact her long-term and how I can respond to that productively. Keeping her from eating non-food stuff is a start, I guess.
I love mathematical models as much as the next person, but what does a formal model of oligopoly tell you that a verbal description wouldn't?
We did an iterated prisoner's dilemma exercise (with opposing teams) in grad school. The idea was to experience it ourselves before any theoretical foundations or predictions, but I persuaded my team to adopt the tit-for-tat strategy and we did very well.
Nah, 176 was just intended to say that I think the crossover point where it gets productive is past the 'spend a couple of hours working through basic examples' point.
But again, this is largely because where I've seen game-theoretical models that seemed to have some useful content, they also really didn't seem like the sort of thing you needed a lot of background to make sense of. Probably needed a lot of background to come up with them, but not to follow them. I haven't run into useful discussions where I was glad I already understood the difference between a game of chicken, a stag hunt, and the prisoner's dilemma.
204.last is probably fair. I was thinking spending a couple of hours on it so you could read literature in other fields that talked about why it didn't explain individual human behavior very well. But I guess you could just do the latter and pick up the necessary game theory from context.
I mean, it reinforces that she shouldn't eat a hammer, but nothing but vigilance keeps her from eating her raisin boxes and hair that she oulls out and so on. Sciency commenter advice would be welcome there.
Ex-kid who ate non-food items here -- I nibbled bits of paper as a kid for a long time. No real insight, but it didn't do me any long term harm, and I stopped after a while. School-age, though -- I was probably still doing it at 7 or 8.
it didn't do me any long term harm
That you're aware of.
Oh, and I chewed on my hair forever. I probably still do occasionally times when it's long enough to get to my mouth. I was a disgusting child.
And am a disgusting adult, come to think. I'm honestly not sure if I still do chew my hair -- it's been shortish for the last few months.
I ate all kinds of weird shit as a kid. Apple cores, corncobs, peach pits, stim-u-dents. Once I got sent to the nurse's office because my mouth was all blue because I'd eaten a blue colored pencil.
I still eat apple cores. They're organic, and it's easier than finding a trash can.
Does paper even count? I definitely ate paper late enough that I remember it (5 or 6). Also chewed on bits of clothing (hood drawstrings, sleeves) until at least 10.
I'm honestly not sure if I still do chew my hair -- it's been shortish for the last few months.
One presumes you are confused about your behavior in counterfactual currently long-haired LB worlds rather than the one we actually inhabit.
Kids should be careful with apple cores. It's fine for adults, but the arsenic content of a few apple cores is a lot for kids. Stone fruit pits even more so (especially if you break them open).
214: But do you pull it out of your head in clumps and then eat that? And chew your toenails? Honestly, the body stuff doesn't bother me as much as that she managed to eat half a potted plant at school and enough of her blanket that she needed a replacement. It mostly only happens when she's tired, so I'm working on helping her learn other things to do while relaxing and I'm really not too stressed about it since her lead levels are okay.
Honestly, I think it's a combination of genetics (her brother once ate a flip-flop, or at least the sole part, which amazes me!) and maladaptive self-soothing techniques she learned as a neglected baby, which is why so much of it happens when she's trying to fall asleep.
My dad's always eaten apple cores.
220: Well, I get a haircut about once a year or so. I cut it to chin length, and then it grows, and I like having it long for a while, and then I get bored with looking like a haystack and cut it off again. So I have had long hair recently, and probably will again soon, I just don't have access to it right now to trigger long-hair-related gross habits.
One annoying aspect of my life is that I'm mandated in the interests of co-parenting to enforce my ex's decree that the kid not drink bathwater during the bath. Unfortunately this is like her favorite thing to do in the world, and, while it is a little disgusting, it doesn't seem like a big deal to me and surely not one worth having huge fights over. Nonetheless, I am obliged to enforce the law, which makes bathtime not that much fun.
We did an iterated prisoner's dilemma exercise (with opposing teams) in grad school. The idea was to experience it ourselves before any theoretical foundations or predictions, but I persuaded my team to adopt the tit-for-tat strategy and we did very well.
. . . because you had read Piers Anthony as a child.
I worry about sticks/wood, dirt, and plants as things she eats that I don't want her to. Oh, and chicken bones, but I think we're over that one. But it turns out the one thing I didn't think I could handle her eating actually was manageable, though I still hate thinking about certain things in the abstract.
Who was it, in the archives, who was recommending eating orange peels and egg shells? Was that Tripp?
226: Bathwater is only for spitting, which is totally fun! She doesn't really follow that rule either, but at least in theory that's the plan.
223: Yeah, eating potted plants isn't safe.
Erm, just brainstorming, is she old enough that Tic-tacs or Altoids aren't a choking hazard? Like, if you made sure she always had a box of Tic-tacs in a pocket, could you maybe try to talk her into transferring the self-soothing oral behavior into sucking a mint? I'd say gum, but obviously falling asleep with gum is a mess.
225: I understand; I was just being an LB*. I have almost zero conscious access to my "automatic" behaviors--for instance during whatever that thread was I had to actually go take a piss to figure out the actual detailed fly/underwear dynamics involved.
*Not a you-variety LB, a nosflow-variety LB.
I'm honestly not sure if I still do chew my hair -- it's been shortish for the last few months.
I first read this as expressing uncertainty about whether maybe your hair is staying short because you're unknowingly chewing it (perhaps at night or something).
One might think that, looking at the kemptness level of my hair generally, but no.
I eat orange peels, doesn't everyone? Like, you have to bite the orange anyway to peel it, and you'd eat candied peel. SO.
My ankle is sore and I'm worrying about whether I should worry about deep vein thrombosis. I think it's just that long flights are exacerbating residual soreness from falling while skiing a couple-three weeks ago. But I would feel really stupid if I ignore it and fall over dead.
Chewed hair could be a look. You'd probably have to pay quite a premium to have a hairdresser chew your hair off instead of using scissors.
Lemon peels too. I always have a bad restaurant moment if they bring espresso with a bit of lemon peel, when I realize after I've started nibbling the lemon peel that that's probably not a normal thing to do in a nice restaurant.
Honestly, I can't take me anyplace.
232: She has chew toys meant for kids with autism that she uses and that helps. I forgot to say that I'm definitely not down with eating small rocks, either, but her teeth are strong and healthy and so I guess she's not chewing them. Helping her learn to self-soothe without chewing is part of the reason is I still rock her to sleep almost every night. One of her teachers has taught her to do a wig pat move rather than pull her hair and that's been great, especially since Lee specifically asked that I not teach her the wig pat, so ha!
But I would feel really stupid if I ignore it and fall over dead.
Unlikely--I've got to believe there's either no afterlife or its one in which you won't feel stupid about silly things like this.
Mmmm, lemon peels. Or at least mmm, lemon zest. Not so sure about the entire peel.
People should carry around a bag of octopi (octopodes, whatevs) to chew on.
Oh, that sounds great, then -- I was thinking pacifier, but something purpose-made for chewing is even better.
And I also chewed the hell out of pens and pencils. Nothing like chewing on a ballpoint and accidentally getting ink all over your mouth.
(What I'm ineptly attempting to convey here, obviously, is that while you don't sound inappropriately worried, you shouldn't be. Not that I'm evidence that being a non-food chewer is compatible with growing up normal, but it's at least compatible with having a legal career.)
one in which you won't feel stupid about silly things like this.
YOU'LL HAVE MUCH MORE CONSEQUENTIAL THINGS TO REGRET.
apparently I am almost exactly like LB in my-non-food-chewingness, for I too have bitter bitter memories of the inky hell that is an overchewed ballpoint.
SFO's wireless is unable to find Google and yet it can connect me to Unfogged. I'm glad their DNS has its priorities in order.
My weakness is thumbs (chewing-wise). At least it's my own thumbs.
247: Or you did a hosts file at some time in the past.
The description of game theory above as the study of mindless rational utility-maximizing automatons sounds kind of like microeconomics in general, doesn't it? I have to admit I'm fuzzy on where the disciplinary boundaries are, since I only know about game theory at the level of a few silly examples. Is the distinction that game theory isn't about utility-maximizing robots in general, but only those that are competing with a small analyzable number of other utility-maximizing robots?
I think the Hershey kiss example is missing some kind of stipulation, like that the goal is to maximize the highest guaranteed number of kisses rather than the highest probable number or something like that. The robots don't want knowledge, they want certainty.
Yeah, all of that class of problem require not only that you be perfectly rational and utility maximizing, but that you know the other players are all the same. At that point, if there's a unique equilibrium, there's no probability about it.
252: Not necessarily. I believe you are over-focusing on that (admittedly large) part of Game Theory. See 135 for example.
But it really isn't assuming that, is it? If you were both perfectly rational and utility maximizing, you would trade off and each take half. Heebie's solution is assuming the worst-case behavior of your opponent, which isn't very rational if your shared goal is to get lots of chocolate instead of to "win".
251: no, if you're trying with each move to maximize your total expected utility (across all moves) and are operating under the assumption that your counterpart will also with each move maximise his or her total expected utility across all moves, then the rational first move is for player one to take 2 kisses.
Heebie's solution is assuming the worst-case behavior of your opponent that your opponent will make a utility maximizing decision on each and every move.
(Maybe that's another way to say the same thing...?)
Or follow this obscure link and look at application section.
Oops, that was supposed to be a strikeout, not italics.
256: No, it isn't. The utility maximizing decision is that you both get 49 or 51 kisses, not 2.
254: Not exactly assuming worst-case behavior -- assuming it's impossible by stipulation to communicate or make deals.
253: That's what I meant by 'that class of problem', though -- if you're handed a problem like that as a logic puzzle, the assumptions are always that's everyone's perfectly logical and utility maximizing, no communication, unless otherwise specified. That's what you need to get the fun counterintuitive result.
It's not all of game theory, but it's all of the game theory problems that get bandied about as brainteasers.
Heebie's argument shows that this isn't an "equilibrium," for some definition of equilibrium. The trouble is that equilibria are not, in general, maxima, and this pretty clearly isn't. It's like you have a curve with a peak and for some reason the exact peak isn't an allowed point, and rather than choosing a nearby point you instead roll down the hill. That's stupid behavior.
So, if you're talking about wrapped Hershey's Kisses, in their little foil suits, and nothing was said otherwise, why couldn't you just dig all 98 out of the trash container, wash your hands, then peel and eat them? Better yet, you could make those peanut-butter-and-sugar-and-Hershey's Kiss cookies. THAT would maximize your utility.
Even better, you could do the same experiment, but with the miniature Reese's Peanut Butter Cups. Including the making cookies part. That would be almost as good as pie.
146 to 261. Or, in other words, yes, I agree with you.
Similarly, no police force has ever offered a pair of actual Prisoners the Dilemma. Because in real life, they'd both say "Fuck you, copper" and walk.
259: The utility maximizing decision is that you both get 49 or 51 kisses, not 2.
If you can coordinate, that's right. If you can't coordinate, the utility maximizing decision is that you (player 1) get 2 instead of 1.
261: Sometimes rolling down the hill is fun though. You can't do it as often as sledding, but even here in the frozen north there are canonically 3 months of damned poor sledding.
No, it isn't. Utility is higher if one of you gets 49 and the other 51. That's the optimal utility for both of you. So even without coordinating, you should both pick it.
265: Of course, actual human beings who were prevented from coordinating, but weren't perfect utility maximizers, would probably actually get more chocolate out of the deal by taking an uncoordinated leap of faith on the other guy behaving well.
I don't get what you're arguing, essear. If you're just saying 'that's stupid behavior', then sure, you're right. (Or, as player 2 would more likely view it when cursing out player 1: asshole behavior. Stupid, asshole behavior.) But if you're suggesting the problem doesn't work, and that's it's not really the utility maximizing decision given the constraints of the problem, you're wrong.
Yes, I am saying this is not the utility maximizing decision given the constraints of the problem. It is only the utility maximizing decision if you assume that your opponent is out to screw you instead of maximizing their own utility.
268: right, sorry, it's not just that you can't coordinate, it's that you can't coordinate, are a utility maximizing robot, and you know that the other player is also a utility maximizing robot. Even you even suspect that there's a chance the other player might be partially human, then gambling by taking just one chocolate probably is the utility maximizing choice. But if you know it's a robot, if you only take one, you're going to end up with one chocolate, and he'll have 2, and it's game over. So you should have taken 2.
In other words, there's a perfectly good strategy which is "at every step, make the decision that will allow you the highest possible total outcome," and if you both follow it, you both come out roughly as happy as possible. One of you does slightly better than the other, but so what? At least you're not stuck with 2 or 0.
So, essear, how exactly would you advise player 2 to maximize his or her utility? Walk me through it.
274 before seeing 273? I don't know how to say it more clearly. You have a landscape of possible outcomes and you can both walk uphill instead of falling flat on your face at the bottom.
My laptop battery is about to run out if I don't find an outlet somewhere.
I'm with essear on this one; I think you guys are throwing in what is basically the unexpected egg paradox and getting it wrong. Actual rational maximizers would realize that the meta-analysis will lead both to cooperate.
(Actually, arguably player 2 ought to take just one chocolate every time, because if player 2 even gets a turn, that means player 1 has already tipped his hand, taken only one chocolate and exposed the fact that he's irrational. So taking just one every time is a good bet.)
(But in the real problem player 2 never gets a turn, because player 1 is a robot.)
taken only one chocolate and exposed the fact that he's irrational
This is totally ass-backwards.
I agree with your reasoning in real life (see 137). But the argument Heebie spells out in 128 works. If you could possibly get to the 99th turn, Player 1 would obviously, if she's a utility maximizer, have to take two Kisses and end the game, rather than leaving the last one for Player 2: she doesn't have a choice, she's bound by the iron law of utility maximization to make her highest value play, and finish with 51.
So on turn 98, if Player 2 only takes one kiss, she knows for certain that's her last one. Player 1 can't leaver her the last one. Player 2 has a choice between taking one, and finishing the came with only 49, or taking two and finishing with 50. Again, Player 2 has no choice as to what play to make.
And then roll it back up with the same reasoning to the first move. To get away from that equilibrium, someone's got to make a move at some point that's not their highest value move. If the players aren't able to do that, they can't do anything but end the game immediately.
This is beginning to sound a lot like the arguments we had in Lincoln-Douglas debate about the concept of "utilitarianism" as an meta-value for debate topics. Of course, hardly anyone went back and actually read Bentham and Mill, and so very often you would hear people appealing to "utilitarianism" to defend propositions that are explicitly rejected in On Liberty.
I really wish someone had steered me to Quiz Bowl/Knowledge Bowl, or even Mock Trial, as a freshman, instead of letting me waste 3 years in boring old Debate. Speech might have been okay, but the adviser was exceptionally creepy and did indeed wind up being let go for not being able to keep his hands off the merchandise.
Step 1: Analyze the situation.
Step 2: Assume the other player has done the same.
Step 3: Take one, expecting and being correct that the other player will take one.
In actuality with "similar" setups humans will sometimes take 2 because they get focused on either "winning" (my x minus your y), or assuming the opponent will do the same.
Lot of typos in 280. Sorry about that.
280: And I'm saying it's a screwy definition of "rationality" that only admits equilibria as allowed outcomes. Players who really want to maximize their utility should accept a slightly nonoptimal outcome that's as close to a maximum as possible if the alternative is to take a minimum.
Actually, wait, I retract everything--heebie did set up the problem wrong, I think. Player 2's utility maximizing choice at each step is to take only 1 chocolate. Given that player 1 knows this, player 1's utility maximizing choice is the same. So heebie screwed it up. But the general sort of problem works.
Like, you have to bite the orange anyway to peel it, and you'd eat candied peel. SO.
Is this performance art? Do you have no knives or fingernails?
I think if you tried to write down the precise mathematical assumptions that are going into heebie's argument, you would see that at least one of them is transparently silly in a way that isn't in accord with intuitive definitions of "rational" or "utility-maximizing".
285: I don't think so -- Player 2 is motivated to defect one move before Player 1 does. Player 1 is motivated to defect either on the 99th move, or failing that, one move before player 2 does. If you accept that Player 1 always defects on move 99, that gets Player 2 defecting on 98, and then it rolls back up to the beginning. (Assuming they're utility-maximizing robot morons.)
I probably shouldn't try to engage in extended arguments on the internet when coming off an 11-hour flight. Stepping back a bit, I do understand what HG and LB are saying, I just think it's nutty even if you assume you're talking about rational utility-maximizing robots. Those aren't the robots you want to consider to get meaningful answers.
Those aren't the robots you want to consider to get meaningful answers.
These aren't the droids you're playing against.
289: no, 280 is right. I un-retract everything. I got confused for a minute. I'm trying to eat a goddamn burrito here.
280: That backward up-the-chain logic is the root of the unexpected egg paradox. Someone gives you 10 boxes and says that if you open them sequentially you will "unexpectedly" find an egg in one. By similar logic to the Hershey thingie you can reason that it "can't" be in the 10th one you open and have it still be "unexpected", and so for the 9th, 8th, 7th etc., until in a burst of triumphant logic you say that it cannot be unexpected. And yet you in fact have no clue which one you will find it in.
Is this performance art? Do you have no knives or fingernails?
Generally no, actually --- isn't this the universal way of opening oranges?
293: I think it doesn't work for the unexpected egg paradox, but it does 'work' for the robot morons. Although I'd be really interested in a definition of rationality (as opposed to 'acting like a reasonable human being') that would let a utility-maximizing robot get more chocolate out of the deal.
I breach the orange peel by biting it, and peel the rest back with my hands. I can't say that I'd ordinarily eat much of the peel, but it isn't, like, toxic or anything.
"The Unexpected Egg Paradox" sounds like the name of a really terrible jam band.
295: I think you folks are constraining the robots to a particular form of "rationality"--but it is hard to see the way out with the maximizing re-analysis at every step. I'll think about that on the drive home.
296 is the procedure I refer to! I certainly don't always eat the peel, but it is definitely an acceptable foodstuff when pushed into service.
I do think there's something to 277: if robot 2 gets even a single turn, that tells him robot 1 isn't acting in the predicted game-theoretical utility-maximizing fashion. That's useful information, which he ought to utilize, and which ought to lead him to take only one chocolate. (Clearly, having taken one once, given the constraints of the game, player 1 is unlikely to take two on turn 2.)
And Player 1, knowing this is the rational response of Player 2, ought thereofre to be able to anticipate it, and to take only one chocolate on the first move. They both will realize that someone's going to get screwed towards the end, but they'll still be better off then shooting themselves in the foot right away.
295: For one thing, you could imagine that their utility function is relatively flat for near-maximal choices and less so for very non-maximal choices; they both walk uphill for a while, and then near the peak, one or the other might decide to screw the other, but they clearly agree that first they need to get near the top. Details left as an exercise for the reader, or something. (Sorry. I'm sleepy.)
Essear, heebie is right, and you're trying to sneak in actual human psychology into the game. If you knew that the other player was applying the reasoning that you describe, you would screw them in the last period by taking the last 2 kisses. If the other player figures that you're going to screw them in the last period, then they're better off screwing you in the second-to-last period. Any period you think you're going to get screwed, you're better off screwing the other player in the previous period. So by backwards induction player one ends up screwing the other player in the first period.
You can say that they're better off waiting to screw each other as late as possible, but whenever I have a prediction of the period you're going to screw me, I can always do better by screwing you one period sooner. So I will end up deviating from your waiting strategy.
Looks like urple and I have converged on an answer in 300.last and 301.
302: Again, this all depends on very particular definitions of "rational" and "utility-maximizing" that don't necessarily make a lot of sense. If you're going to try to make a mathematical model of something, you'd better define your terms precisely, and I'm free to pick them apart.
251: Another way to put it, I think, is that we stipulate mutual certain knowledge that the participants will in any situation perform an act if it would definitely maximise their utility (calculated in kisses) if they did it. So we're assuming that the probability the other guy might not defect (in situations where it would definitely be to his advantage to do so, like the one where there are two kisses left) is zero. So the highest probable number and the highest guaranteed number are the same, given the assumption. (In other words, what LB said at 146.)
I can't figure out why 300 isn't right. Is it right?
I'm sure biting works fine, but I've always used my fingers to open oranges - you certainly don't have to as per 236. And some people just cut them into slices.
Time to board my flight and finish reading Debt. Maybe there will be comity when I land.
It is a little more difficult to open the peel (and separate the pith from the flesh) with hands. Sometimes hands don't fully penetrate the pith, leaving you with a lumpy mess, with uneven pith and chunks of the flesh. Better to get deep enough to pull away the pith smoothly, and teeth do that easier.
I think I made a small mistake and the "maximize the possible outcome" strategy ends at 50-48. Pretty good for all concerned.
||
Judge Cebull news: http://www.ce9.uscourts.gov/absolutenm/articlefiles/437-Letter%20to%20President%20Obama.pdf
http://www.ce9.uscourts.gov/absolutenm/articlefiles/437-Letter%20to%20Judge%20Kozinski.pdf
Says the Judge to the President: Honestly, I don't know what else I can do. Given that he's a year or two short of the time period in 28 U.S.C. 371(c), I suppose he doesn't know. (I don't think the Jucidial Council is going to tell him he has to step down.) He'll make it during the second term, though, and I think no one will begrudge the President a bit of a smile as he sends a nomination up.
|>
Here's the wikipedia page, for the curious. (They're called "Centipede games", not caterpillar.)
For some reason, I love this:
The unique subgame perfect equilibrium (and every Nash equilibrium) of these games indicates that the first player take the pot on the very first round of the game; however in empirical tests relatively few players do so, and as a result achieve a higher payoff than the payoff predicted by the equilibria analysis. These results are taken to show that subgame perfect equilibria and Nash equilibria fail to predict human play in some circumstances.
Am I happy or sad that experiments were run to confirm that these games fail to predict human play "in some circumstances"? I'm not really sure.
You know who neatly predicts why the Nash equilibria fail to predict human play?
Kah-ne-man!
the "maximize the possible outcome" strategy ends at 50-48. Pretty good for all concerned
Of course, that's not as good for player 2 as 48-49, so why did player 2 pass over that opportunity when he had it?
315: highest possible. At that stage, 50 was still a possible outcome for him.
313: Thanks for the link. At some level I am clearly wrong.
I don't follow 316. In your version, why does at end at 50-48 instead of 51-49?
Says the Judge to the President: Honestly, I don't know what else I can do. Given that he's a year or two short of the time period in 28 U.S.C. 371(c), I suppose he doesn't know.
I could make a few suggestions.
It is a little more difficult to open the peel (and separate the pith from the flesh) with hands. Sometimes hands don't fully penetrate the pith, leaving you with a lumpy mess, with uneven pith and chunks of the flesh. Better to get deep enough to pull away the pith smoothly, and teeth do that easier.
I just cut my nails and I can't do a thing with them!
From 313:
Palacios-Huerta and Volij (2009) show that expert chess players play differently than college students. With a rising Elo, the probability of continuing the game declines; all Grandmasters in the experiment stopped at their first chance. They conclude that chess players are familiar with using backward induction reasoning and hence need less learning to reach the equilibrium.
This is really interesting, because I'd thought the logic felt very similar to chess reasoning.
303.last: No one is actually interested in defending the definition. You'll have to find actual game theorists to pick on for that one.
More from 313:
Another example Binmore likens to the Centipede game is the mating behavior of an hermaphroditic sea bass which take turns exchanging eggs to fertilize.[citation needed]
Oh, sorry, that's the wrong quote. I meant this:
Since the payoffs for some amount of cooperation in the Centipede game are so much larger than immediate defection, the "rational" solutions given by backward induction can seem paradoxical. This, coupled with the fact that experimental subjects regularly cooperate in the Centipede game has prompted debate over the usefulness of the idealizations involved in the backward induction solutions, see Aumann (1995, 1996) and Binmore (1996).
I'll sleep happier tonight if someone adds a link to this thread as a reference for this point.
Am I happy or sad that experiments were run to confirm that these games fail to predict human play "in some circumstances"? I'm not really sure.
See comment 151. Some variations make opening dick move what everybody would do.
This is the version of the "unexpected egg paradox" that I'd heard before. I'd never heard it as an unexpected egg.
324: well, right, but when the reward for immediate defection is vastly more enticing that the entire pot that could potentially be split, it's not a very interesting game.
Kah-ne-man!
Late to the thread, but it really is a good book. (Well, except for the "talking about" sections at the end of each chapter. Those sound so much like they're made up by a robot that I half-expect they're there to demonstrate some cognitive effect that'll only be revealed at the end of the book.)
I've actually only read about a chapter of the book so far, but I recently read some of his papers (his nobel lecture, some of the old stuff with tversky). I'm a little worried I won't like the level the book is pitched at, but we'll see.
326 to mean: 100 quarters on the table and a $5 bill on the side is more interesting than 100 bad chocolates and a $50 bill.
Especially since you can get more than 100 bad chocolates with $50, if that's your bag.
313: Am I happy or sad that experiments were run to confirm that these games fail to predict human play "in some circumstances"? I'm not really sure.
This made me laugh. Sad! Or rather, happy! Since some people needed to see experimental evidence (studies, cites) in order to be convinced. Or yet sad again, that we must produce such evidence before making a claim, any claim.
... which thoughts are, I see more or less pwned by 323.
313, 331: Of course we should be glad for the experiments! This is just the classic double-bind psychology always faces. If the outcomes confirm what 'everyone already knows', then (duh) psychology only tells us the obvious. If they're surprising, well, we predicted that too, since after all humans are such strange and contrary creatures. Either way, of course people do/don't behave like that.
But this game we know to be uninformative. Sometimes common sense gets it right and sometimes it doesn't, and there's only one way to find out. I doubt I'd even trust my opinion as to how *I'd* behave in a real-life game theoretic situation. (And the studies agree: I shouldn't.)
Isn't 300 right, though, and the game incorrect even on its own terms?
332: Experiments are obviously useful when there's reason to question folk psychology. However, the premises of the experimental framework should, I think, bear scrutiny from the get-go: if, as in 313, "perfect equilibrium" is being tested as, I dunno, a possible way to understand and explain how people behave, well, uh, why on earth do we think they behave that way in the first place? I mean, I can test whether people believe the moon is made of green cheese, and my results will (likely!) show that they don't ... but do we really need me to be able to show the results of my study before we agree that people know the moon isn't made of green cheese?
"There's only one way to find out" only goes so far.
I must ban myself for analogizing now.
I don't really begrudge them the experiments, Man Suit. It's just something where, you know, everyday experience sort of *is* an experiment, and pretty clearly shows that people do not behave this way. One might hope the progression of inference would be: (1) we've constructed a model that predicts that humans behave in bizarre ways, therefore (2) what's unrealistic about our model?, NOT: (1) we've constructed a model that predicts that humans behave in bizarre ways, (2) therefore, we assume that humans behave in bizarre ways, (3) we run experiments in a lab to test whether humans behave in line with our assumptions, (3) they do not!, therefore (4) what's unrealistic about our model?
But, sure, the experiments obviously don't hurt anything, and it's good to test our intuitions.
Oh, shit, that second set should be numbered (1) through (5).
the experiments obviously don't hurt anything
The experiments do fuck with our ability to judge when empirical data is actually needed.
I realize I'm being fussy and possibly obnoxious, but 335 is on point.
333: 300 assumes that Player 1 doesn't find the backwards induction argument compelling, so doesn't immediately defect. At that point, Player 2 knows she's not playing against a utility robot, so it's a different game even if she's a utility robot.
316: That doesn't work. 50/50 is not a possible outcome if player 1 is a utility robot -- on the 99th move there's no argument for not defecting. If they get to move 99 and player 1 doesn't defect, this is not the game as it's been defined. You agree with this: that's why you said 48-50 not 50-50. But if you can figure this out, so can player 2. Which means they're going to defect on move 98. And that goes back to 0.
I might be talked into an argument that the utility maximizing thing to do is pick a random number within some range and defect on that move. But any argument that give a deterministic answer for which player defects when has to fail, because if both parties follow the same train of thought the other one does better by defecting one move earlier.
300 assumes that Player 1 doesn't find the backwards induction argument compelling
I don't think that's really a fair response, insofar as it seems to be saying that player 1 in 300 is missing the actual compellingness of the backwards induction argument. You might rather think that what the strategy in 300 shows is that the argument actually isn't compelling, because the players can, in fact, communicate with each other by making certain moves not predicted by that precise argument.
Doesn't Russell Hardin have a line on this similar to LB in 338?
I think urple is correct that the solution stated in 128 to the problem posed in 116 has a hole. The solution depends on each player knowing the other will always act as game theory predicts. This is a situation where you would be better off if you could make a binding commitment (eg I will never take 2 kisses). You can't do that but you can prove you won't always act as game theory predicts by taking 1 kiss which means the argument falls apart.
Or to put it another way the position where you start with 50 kisses is not the same as the position where you arrive at 50 kisses from 100 as in the second case the players have been able demonstrate through their moves that they won't always behave as game theory predicts. So you can't reason inductively.
Similarly: "it's a different game even if she's a utility robot."
The game is defined by its rules, not by the psychology of its players. If it's better to lose with 48 kisses than with zero, and indeed better to win with 50 than with 2 (the whole point of the game is that kisses are something good, right? it's not just about winning the game, you're winning kisses) then I don't see why the reasoning in 300 isn't sound.
In a way 300 works because the backwards induction argument does hold for not-very-thoughtful utility maximizers: taking only one kiss thereby becomes a way of communicating to your (presumed thoughtful) opponent.
Of course if the second player, after the first player takes just one, figures, well, player one is a moron, then it won't work.
Wait, I got that wrong. You think player 1 is going to defect on move 97. I don't understand why you stopped there, but whatever your thinking, what keeps player 2 from following the same train of thought and defecting on move 96? If you can come up with which move makes sense to defect, a utility robot can do the same.
335: we've constructed a model that predicts that humans behave in bizarre ways,
See, this is the part I don't see as being one bit intrinsic to Game Theory per se--it is rather a relatively inappropriate application to human behavior in that situation.
Oh wait I forgot it's not even a matter of winning or losing; all the specification of the game says is that after a player takes two the remaining kisses are thrown out. In that case I really don't see why you wouldn't just take one, if you're the first player, and hope the second player is canny enough to catch on.
(Perhaps utility maximizers don't have hopes like that. But surely some of them have expected values for canniness of their opponents.)
But I don't understand why utility maximizing Robot 1 wouldn't be able to figure out that if it doesn't defect on the first turn everyone gets more. What's utility maximizing about a robot that can't figure out that its best first move is to signal cooperation?
335, 337: I don't disagree at all that if you already have tons of evidence that this is a bad modelr, you're better off focusing straight away on where it's defective rather than detouring through yet another demonstration that it is. As a sociological observation, it's not so easy to get a study published unless you present the reasoning in the second way, however, since referees are cranky and will indeed demand proof that the model is a bad one. That goes some ways towards explaining why the literature looks the way it does.
I'd push back a little on the idea that everyday life functions as a kind of experiment, since I doubt it does for most of us. Most people's naive physical intuitions get it wrong about simple cases they've observed a million times.
But moreover, I'm not sure that economic and game-theoretic models of rational agency really were ever meant to function as descriptions of real people--or even 'ideal people'. They're norms, so they're not straightforwardly falsified by the fact that people suck at conforming to them. The whole point of the 'rationality wars' was to show that people are actually really good at normatively right reasoning--if you keep your eye on the right set of norms. (Or, sometimes, that people can obey the norms just fine if you frame the problem right.)
Anyway, my apologies, since that's all not merely fussy but positively scholastic, which is obnoxiousness squared.
342
Of course if the second player, after the first player takes just one, figures, well, player one is a moron, then it won't work.
It works fine if player 2 thinks player 1 is a moron. As player 2 should not take 2 kisses if they think player 1 will repeat their "error" and only take 1. The problem is if player 2 thinks player 1 has had a momentary lapse of judgment.
What's utility maximizing about a robot that can't figure out that its best first move is to signal cooperation?
The counterintuitive results it generates mean more grant money for its designers.
Right. There is no stable equilibrium other than Player One: 2 chocolates, Player Two: no chocolates. Once you start playing the game it could end any time, and both players realize the endgame will be unstable. But cooperating to play the game at all very quickly makes both players better off than ending it immediately, and they're able to signal from the outset that they're willing to cooperate, at least for a limited period of time.
For essear, clew and others who were discussing women physicists:
The reference you're looking for is Traweek's book Beamtimes and Lifetimes, page 104.
Another way to say the problem I see with 300 is that on move 2, player 2 can say to themselves "okay, the backwards induction argument isn't compelling to player 1. So this is going to go on for some time. Which one of us is going to defect when?" And then I can't see what keeps player 2 from making the backwards induction argument from any endpoint she reasons her way to.
Admittedly, the logic in 300 seems less compelling if you imagine the payoff for defection being 10 chocolates instead of 2.
I'm not sure why, at that point, the second player has to make the argument at all. Why can't player 2 just say, "okay, the backwards induction argument isn't compelling to player 1. Great!"?
347 is fine, ManSuit -- and welcome, by the way.
it's not so easy to get a study published unless you present the reasoning in the second way, however, since referees are cranky and will indeed demand proof that the model is a bad one. That goes some ways towards explaining why the literature looks the way it does.
Understood. The rest of us find it tedious of you all sometimes, but understood.
I'm not sure that economic and game-theoretic models of rational agency really were ever meant to function as descriptions of real people--or even 'ideal people'. They're norms, so they're not straightforwardly falsified by the fact that people suck at conforming to them. The whole point of the 'rationality wars' was to show that people are actually really good at normatively right reasoning--if you keep your eye on the right set of norms.
This is weird, and I don't really understand. I'd have said they were intended to be models. Descriptive.
353: As I can work out (Amazon will give me page 104 itself, but not anything before that), the claim there is that physicists talk about their profession in ways that code masculine in their culture. E.g. in western culture there's a tendency to talk about physics in ways that involve geniuses having key ideas independently and fighting off the legions of doubters, and that this kind of narrative codes male. (This seems pretty true to me, I read a paper once on "The sexual politics of genius" that makes similar points, and it seemed pretty persuasive.) This is rather different from suggesting that many physicists would say that women can't do physics because they're too collaborative.
Most people's naive physical intuitions get it wrong about simple cases they've observed a million times.
Constantly! About everything! (Man Suit is great on this thread.)
I mean, that's the thing about folk psychology that's so nifty; it's simultaneously constantly wrong about almost everything and very close to a provably maximal solution to operating in the world.
Anyhow, 335 is kind of true and kind of not (this is hella going to recapitulate Man Suit); the real (1) is "we have a vastly simplified model that seems to do a pretty good job of predicting how people behave in the aggregate; can we test if it's good enough to do useful work with", and the real (2) is "we know our model is unrealistic in various ways, but which of those ways are important given what we would like to understand?"
359: Maybe it was a different edition, but this is what I'm looking at on 104:
"I am not suggesting th.at only biological males can participate in the cycle. I am claiming that in this cycle a certain cluster of characteristics is associated with success, a cluster that is part of our culture's social construction of male gender.51 These stories about a life in physics define virtue as independence in defining goals, deliberate and shrewd cultivation of varied experience, and fierce competition with peers in the race for discoveries. Indepen- dence, experience, competition, and individual victories are strongly associated with male socialization in our culture. By con- trast, recent studies in Japan suggest that these are the qualities associated with professionally active women, not men. Women are seen as not sufficiently schooled in the masculine virtues of inter- dependence, in the effective organization of teamwork and cama- raderie, commitment to·working in one team in order to complete a complex task successfully and consulting with group members in decision making, and the capacity to nurture the newer group members in developing these skills. It would appear that there is nothing consistent cross-culturally in the content of the virtues associated with success. We do see that the virtues of success, whatever their .content, are associated with men."
Right. I understand. Really, it was just the phrasing in the wikipedia article ("fail to predict human play in some circumstances") that triggered my "No shit?" response.
363: ah yeah I mean the empircal economic literature sort of triggers the same response in me.
360: Ha! Exactly. That's the horrible truth about 'folk psychology'. On the one hand we're forever misunderstanding other people, tragically and incorrigibly, including those allegedly closest to us, and being just as misunderstood in turn; and if we really grasped how wide the gap of understanding is, we would plunge into (in order): despair, and the deepest nearby pit.
But at the same time, almost nobody notices these facts because WE ARE ALL ABLE TO SHOW UP TO MEETINGS IN THE RIGHT PLACE AT THE RIGHT TIME.
As a social coordination device? Not too crappy. As a predictive/explanatory tool, or a medium for understanding others' inner worlds? *crickets*
That's the paragraph I read. Nothing there about Anglo physicist sayin women are bad at physics or giving reasons for it. Rather they're talking about it ways that code male, while in other cultures people talk about it in incompatible ways that also code male. An interesting point to be sure, and close enough that it's easy to misrember one as the other, while far enough away that the misremembered claim sounds implausible.
Folk psychology is constantly wrong about almost everything?
Maybe "folk psychology" in your mouths means something different from what it means in mine, but my impression is that folk psychology mostly gets things right.
I have to repeat my 358.last, which quoted Man Suit's 347.3, which suggested that it's not the case that economic and game-theoretic models of rational agency really were ever meant to function as descriptions of real people--or even 'ideal people'. They're norms, so they're not straightforwardly falsified by the fact that people suck at conforming to them. See the rest of 347.3.
Are we using descriptive/prescriptive (normative) in different ways here? Also don't know what the 'rationality wars' refers to.
264
Similarly, no police force has ever offered a pair of actual Prisoners the Dilemma. Because in real life, they'd both say "Fuck you, copper" and walk.
Really? So you think there is honor among thieves?
Wait, if the norms aren't even supposed to function as descriptions of ideal(ly rational, for instance?) people, what are they norms of? I mean, anyone can offer a norm, right? Surely we want norms of rationality that people by and large actually do conform to? [Selected works of Davidson hereby incoroporated by reference.]
358: Models are just devices of representation. So in principle they can model close to anything. Sometimes they're intended to be descriptive: whoa, look at this projection of Soviet missile production! Or, check out the rate of STD transmission among the unicorn population. (Where did you *think* they all went?)
But you can also model things like how simplified agent-like systems that conform to some seemingly rational principles would behave. If you thought they existed you could use that model descriptively. If you didn't, you could use it normatively: see, there aren't any systems that are really like this, but there are all sorts of benefits to doing what they do, so if you want to get the goodies associated with rationality, model yourself after the model. I guess that's how I'd think of these things.
(One can fudge this by saying that the models *do* describe fictional rational agents, but c'mon. That's just to repeat they they're representations. Whether they're descriptive or normative is in their use.)
There's more at the door.
Conform to the norm.
Conform to the norm.
Conform to the norm.
Conform to the norm.
Conform to the norm.
Conform to the norm.
Conform to the norm.
Conform!
367: I think I might be using "constantly" somewhat nontraditionally; folk psychology gets lots of things wrong in every domain of understanding on an ongoing, continuous basis, but still gets enough right to solve the fiendishly underdetermined inverse problem of figuring out what's in the world based on sensory input a good, what, 80-90% of the time? More? Hence "nearly provably optimal".
I have no idea what Man Suit means by "norms" in this context.
Meanwhile, 365's feeling that folk psychology, as a predictive/explanatory tool, or a medium for understanding others' inner worlds produces ... crickets? Oh my dear. What is literature, never mind relationships countless under the sun, never mind almost every conversation you've ever had? We do mostly understand one another, fairly well and sometimes quite deeply. Of course whatever it is we understand might not be called 'folk psychology'.
What are the error bars on Hamlet?
If you thought they existed you could use that model descriptively. If you didn't, you could use it normatively: see, there aren't any systems that are really like this, but there are all sorts of benefits to doing what they do, so if you want to get the goodies associated with rationality, model yourself after the model. I guess that's how I'd think of these things.
You could say that there are all sorts of benefits to acting the way the way the hypothetical entities this model would accurately describe would act, were there any such, but I'm not sure why you would then describe that model as "rationality", rather than "behaving like this".
(Rationality is not normally thought of as something that you have (additional) reason to pursue, a reason that you would have already to be rational to act on, whereas the way you've presented things, in fact that is the case: the reason we have to be rational is that being rational helps us get things we want (the goodies), so, you should model yourself this way, be rational—at least, I guess, if you're rational? Not that there aren't puzzles here which mostly I find beyond my ken, though maybe if I'd actually read the paper I just linked I wouldn't anymore.)
376: the original research determined his father to still be alive.
Based on 373 I don't think Sifu and I use "folk psychology" the same way at all!
Oh I might be using that oddly too at that.
Have you read Noë's Action in Perception or any of that so-called "4EA" stuff?
I think I was thinking of folk psychology as a subset of fast, intuitive cognition and then I was kind of using it as a synecdoche for such and then maybe I lost the thread.
381: I have not. The description here certainly sounds reasonable enough, potentially.
377: That's a better way of putting it. God help me if I get into a debate about the foundations of rationality. I'd rather revisit the justification of deduction if I'm in a mood for self-flagellation.
Anyway, back into the scholastic weeds. Calling the principles behind the construction of the model 'rational' is just a way of saying that they seemed to the modelers, on reflection, to capture some features of good reasoning, and the model is a way of drawing out some consequences of those principles in action. To the degree that those consequences are appealing, that presumably constitutes an endorsement of being like the modeled agents. I guess (hope) that agrees with your point.
374: I've been failing to find a pithy way to reply to this. I wouldn't want to disagree with any of those cases--not at all. But still, it pretty often seems that giving things the slightest little turn can reveal an awful lot of this understanding to be illusory, or at the least frighteningly shallow. (There are studies, but we're in the realm of attitudes more than evidence here.) Plus, we want pretty badly to find these connections, so we tend to over-estimate how deep and enduring they really are.
Then again, I'm equally baffled by why *I* think, desire, and do what I do. At the very least, these most basic facts elude articulation (but thanks for indulging the attempt).
I recalled the Prisoner's Dilemma-based exercise in human "irrationality" I was thinking of above. It was run on the first night of some team-building retreat and consisted of setting up games with two teams each who did a multi-round cooperate/defect game with payouts that would lead to a Prisoner's Dilemma situation. However, the catch was that the single explicit-stated goal was to maximize the total payout between the two teams--nothing on individual payout. And yet most ended up with defections. Somewhat interesting--although the facilitator laid a heavy thumb on the scales by having everything seem competitive despite the specific goal--but even more frustrating if your team would perversely refuse to get on board.
386: So not a game theory exercise at all, but one which used its trappings.
385: I'll be honest: I feel like I'm reading someone who's in the throes of realizing that we can't be inside one another's heads ... we can't really know what anyone else experiences ... we are all, each of us, isolated and alone, operating with an individualized 'theory', constantly revised, as to what's going on with others and the world, and goddamn that is shocking.
Except that, well, the notion that it could -- barring the advent of ESP or a group mind-meld or the singularity -- be any other way is a fantasy in the first place. There's really not anything fundamentally dysfunctional about the way we are with one another: we actually do share a great deal. Forms of life and whatnot, that not only inform and enable communication but shape our selves.
I'm not sure if 388 seemed rude. It wasn't meant to be.
388 -- Just watched Being John Malkovich. That's one strange movie. Has anyone else noticed a marked decrease in quality in the movies available for Netflix streaming?
Duncan Watts' Everything is Obvious is pretty good on people's intuitive explanations for stuff and why they often suck. Although he's hilariously wrong about historians at times. Didn't anyone tell him that the stereotype is that they always say that things are complicated rather than simplistically monocausal?
Have we agreed yet that optimization algorithms that don't distinguish maxima and minima shouldn't be called "rational"?
If Nat or others are still looking, I've heard this recommended, and it's free (if you register; it looks like you could give false info, so long as you use a valid email address).
The fable My/erson tells at the bottom of page 65 here is a nice illustration of the wackiness of game theoretic rationality.
393: That's cool. I can just use my general business/spam address.
313
Am I happy or sad that experiments were run to confirm that these games fail to predict human play "in some circumstances"? I'm not really sure.
I don't think the failure to predict is very unexpected. The predictions depend on game conditions that are not typical of real world encounters. If the games are tweaked to be more realistic the predictions change drastically. I don't think it is surprising that people apply heuristics that are generally useful but fail in some rare situations.
385: I'll be honest: I feel like I'm reading someone who's in the throes of realizing that we can't be inside one another's heads ... we can't really know what anyone else experiences ... we are all, each of us, isolated and alone, operating with an individualized 'theory', constantly revised, as to what's going on with others and the world, and goddamn that is shocking.
FWIW, I wasn't getting that freshman-dorm feeling from Man Suit's comments at all.
388: I wasn't trying to rehash other-minds skepticism, and fallibilism per se doesn't bother me. The original point was that 'folk psychology', meaning the bio-culturally structured ways we have of interpreting one another's minds, is very good at some things, like social coordination, and (often) shockingly bad at others, like giving you a genuine picture of what is going on with another person. I suspect that's because it's basically *for* coordination, not explanation. Moreover, the nature of the tool itself is that it can produce systematic illusions of understanding others, and there are extremely commonplace circumstances in which we can, if we are paying attention, become aware of just how much slippage there is here. But there are lots of reasons to think this is in the right ballpark, from our general susceptibility to poor causal reasoning to our addiction to satisfying but false narratives.
A quick last shot on literature: there is the phenomenon of what you might call the 'literature of disconnection'. Kafka and DeLillo are getting at *something* real in experience when they vividly show people as opaque and baffling symbols. See also Adam Ross's excellent 'Mr. Peanut' on the everydayness of not-getting-people.
And now, back to Real Work...
the nature of the tool itself is that it can produce systematic illusions of understanding [ whatever ], and there are extremely commonplace circumstances in which we can, if we are paying attention, become aware of just how much slippage there is here
(What I was trying to get at above (not in response to Man Suit, just trying to disentangle my mumbling from last night) is that this is true for all sorts of cognitive functions aside from folk psychology; it's best studied in perception, but there's pretty good evidence that biases and illusions are a universal feature of automatic cognitive processing (see, wait for it, Kahneman). This is all largely orthogonal to the ability of literature to illuminate and inform.)
I think the upthread dismissal of game theory as frequently trivial - "Yada yada Prisoner's Dilemma," in LB's words - overlooks the importance of the even the most trivial game theory concepts.
Lots of people lack an explicit understanding of collective action problems - even though they often have a solid intuitive understanding. I have a dream that universal teaching of the tragedy of the commons could do to libertarianism what universal vaccination did to smallpox.
Game theory works really well when the assumptions model how people actually behave. So, in the $50 centipede problem, or in certain versions of Prisoner's dilemma that model free rider problems very well. Game theory works horribly when the assumptions fail to model how people actually behave. (Like the original centipede problem, or other versions of Prisoner's dilemma.)
It seems to be a very low priority for some game theorists to use well-designed assumptions and well-designed problems. Therefore the whole field gets a bad rap.
Funny, I've always seen the tragedy of the commons used as a (fallacious) argument that collective action is impossible and everything must be privatized. That wasn't the intent of the original article, of course, but that was how it was taught in law school.
tragedy of the commons used as a (fallacious) argument that collective action is impossible
What?
400: It seems to be a very low priority for some game theorists to use well-designed assumptions and well-designed problems.
My remaining quibble with these repeated assertions is that it seems to confuse what actual game theorists do which is explore a set of relatively abstract formalisms (like math! or statistics!) and its specific application of it to human behavior (and from this discussion I think the cooptation of the term 'rational'* as a descriptor of the behavior the agents with its pre-exisitng cognitive and normative load is the real issue).
*I am confident that essear can work out a well-defined rationality that will make us right. Get on that essear. For me I'll take a Turing Machine
402: You haven't seen that sort of argument? The tragedy of the commons means that collectively held resources will inevitably be wasted and abused -- only private property rights can save us.
404: I certainly never have. Your comment was one of those baffling moments when I was like "wait, that's really how they think?"
I swear the argument's out there.
404: I've seen the argument that (in certain circumstances) you need well defined property rights to prevent a tragedy of the commons scenario. I've never seen an argument that those property rights couldn't be applied to public land (so that "everything must be privatized", although maybe you were being flippant on this point). And I've certainly never seen an argument that this somehow implies that "collective ation is impossible"--I'm not sure I even understand how that would theoretically relate to the tragedy of the commons.
I was certainly being flippantly hyperbolic about it.
beamtimes & lifetimes looks likely, as in '88 I was working in a reactor, thinking about grad school, & hearing stories about Japan from my physics-trained father who had just started working there. essear, if it's not true now, hurrah! U~, if you don't find that people ding non-gender-standard people, also hurrah, but d you think it's that good everywhere?
Halford, you've claimed not to believe me before, the second-last being about tithing. Also, you're exaggerating my statement even in this thread; saying women aren't in tech fields is not at all the same as saying women haven't the capacity. However, since I'm willing to admit I'm snmetimes wrong, touchy, & a jerk: when I find the thread,if you (in it) aren't ignoring my existence as a counterexample, I will both apologize & buy something DRM'd from anyone but Elsevier. If I'm right...?
(Is this a game theory case? Above, he assumes I've already read the archives. As it happens I'll probably Beautiful Soup them next week when I'm back at my desk. Must away again, sorry to leave mid-thread, I'll bring pictures of dead things.)
Halford, you've claimed not to believe me before,
In the interests of chilling this out a little, you really can't be offended by Halford's failing to take your word about an interaction he was involved in. You are, after all, failing to take his word about an interaction you were involved in. One of you is right and one is wrong, but until that's settled the mutual disbelief shouldn't be an independent ground of offense.
(I have to say that what you remember his saying seems unlikely to me; I'll be interested in what you come up when you find the thread you were thinking of.)
Okay, but when you say "I've always seen the tragedy of the commons used as a (fallacious) argument that collective action is impossible and everything must be privatized", what did you mean? What's the non-flippantly hyperbolic version of that statement look like? "The tragedy of the commons isn't necessarily always taught/presented in a way that would be likely to eradicate libertarianism"?
The non-flippant version is that I've seen regulatory, collective solutions to managing collectively held resources dismissed because the tragedy of the commons teaches us that collectively held resources will always be wasted and only privately held resources can be efficiently managed.
My remaining quibble with these repeated assertions is that it seems to confuse what actual game theorists do which is explore a set of relatively abstract formalisms (like math! or statistics!) and its specific application of it to human behavior (and from this discussion I think the cooptation of the term 'rational'* as a descriptor of the behavior the agents with its pre-exisitng cognitive and normative load is the real issue).
I'd counter by saying there is a lot of math which gets kickstarted loosely by econ or public policy considerations. Huge simplifications are needed to get the math going. I'm always suspicious that those simplifications are disregarded when the non-math people get pulled in, on the far side of the math theorizing. It's not specific to game theory, but that's a really obvious case of it.
Um, well I don't know what to say to that other than that this seems to me to be closer to support for politicalfootball's 399.2 than an objection to it: whoever is it that you've seen explaining the tragedy of the commons doesn't have the damnest idea what he or she is talking about. So, more universal teaching of the tragedy of the commons would help solve that.
414: I haven't read all the links on the google search in 406, but I swear people implicitly argue that sort of thing fairly frequently.
Aren't there widely acknowledged problems in both legal publishing and legal pedagogy?
Maybe 401 is an example. For instance, the inverse phenomenon, too many property rights crippling innovation, is also well known:
http://en.wikipedia.org/wiki/Tragedy_of_the_anticommons
A competent and honest discussion of the issues would mention both limiting cases.
Not sure what to say about ideas being tainted by unpleasant admirers. For me, the difficult cases are individuals who mix great insight and speculation in their work. EO Wilson and Richard Lewontin are examples. Cases like game theory, where a marginal idea is overused by idiots, I try to just ignore the idiots.
From the abstract of one of the top results in 406:
According to generally accepted wisdom, there are two potential solutions to the tragedy of the commons: 1) government regulation, or 2) privatization.
This seems bog standard to me. (The paper goes no to explain why the "privatization" solution isn't actually a solution at all under any plausible real world conditions--it works in theory, but that's it. Government regulation is the only real world solution to the tragedy of the commons. This is an underappreciated point, I would agree. But that's very different from what you say in 412.)
418: From wikipedia:
Libertarians and classical liberals often cite the tragedy of the commons as an example of what happens when Lockean property rights to homestead resources are prohibited by a government.[33][34][35] These people argue that the solution to the tragedy of the commons is to allow individuals to take over the property rights of a resource, that is, privatizing it.[36]
Obviously, I'm not saying this is right, just that people who brought up the tragedy of the commons at my law school tended to be using it to make this point.
I'm not sure why a moronic law professor serves as an indictment of game theory.
It doesn't. It's a counterexample to the idea that introducing more game theory concepts is necessarily going to improve the discourse: it is perfectly possibly to support stupidness with game theoretical jargon, and I've seen it done lots.
422: yes, and also some people very bad people were well-educated, so let's just go with our gut because in our hearts we already know all that is good and true. I mean, those fancypants snob professors are just sophists anyway. Thank you Mr. Santorum.
I join you in calling for jihad against higher education. Burn the intellectual snobs!
Them too. I'm not playing favorites.
In my experience embedded within The Vast Libertarian Conspiracy, LB is absolutely right about how game theory, as with econ more generally, is used; and I find the thought that this wouldn't be so if only it were taught properly kind of hilarious.
422: Well, that's a brief against the utility of education in general. It doesn't really say anything about game theory.
Indeed, it's not uncommon for libertarians to express the belief that if only everyone knew economics, game theory in particular, the whole world would be libertarian.
All I'm saying is that I have no faith in the use of game theory as a vaccination against libertarianism, as you suggested in 399, given that I've seen libertarian arguments thrive on it.
You know what's another idea that's abused by libertarian idiots? Individual freedom! They use this concept to make the most preposterous arguments!
Why should any school child be subjected to think about freedom? It doesn't necessarily lead to an elevated discourse, quite the opposite in fact. I mean, I went to law school and heard the stupidest things there.
Indeed. Everyone should be required to think about freedom, even if they have to be chained and held at gunpoint to make them do it.
||
Can somebody explain to me what an aneurysm is exactly? The descriptions on the internet talk about the brain or aortas.
I'm asking, because my boyfriend had a sonogram recently, and an abnormality came up in one of the ventricals. They think it's congenital, but he has to get an endoscopic EKG. I think they described it as a possible aneurysm or pseudoaneurysm.
|>
434: Basically a bulge or distension in an artery. This is useful. Common aneurysms are aortic and cerebral but in principle you can get them anywhere the vessel walls thin out. Hope he's okay.
433: I see the Law and Econ guys managed to brainwash you about public education as well. Oh well.
I think it's because the pleasure center floods our system with dopamine whenever we're triggered by auditory speech stimulus of our anal cortex.
Brilliant.
Everyone should be required to think about freedom, even if they have to be chained and held at gunpoint to make them do it.
That's what I've been saying!
436: Damn straight. I'm firebombing the kids' schools right now.
I don't think I'm engaging in "No True Scotsman" when I say that libertarians are characterized by their inability to comprehend collective action problems. No true tragedy of the commons would lead to libertarian conclusions.
Conservatives, on the other hand, often have a very good grasp of the tragedy of the commons, and won't be helped by a more accurate understanding. If everybody else is grazing that land, then I'd better do it first and most.
Yossarian articulated this best. If memory serves, the doctor asked him "What if everyone acted that way." He responded "Then I'd be a fool not to."
409 -- clew, I'm just going to ignore your comments from now on. Just please stop egregiously misrepresenting my words and/or position. I never have said, nor would I have said, nor do I believe, nor (to my knowledge) have I ever implied or wished to imply that there are no women in technical fields. That would have been both a really stupid and a really sexist thing to say, and it's not something I believe, at all. Also, why on earth would I have said that you weren't in a technical field when you said that you were?
As I say, I have plenty of relatives and friends who are women who are in technical fields, and would hope that my own kid would do the same, so I really am a little pissed off that you are going around calling me some kind of supersexist based on some kind of false impression of me that you're making up from whole cloth.*
*You mention the tithing thread, which happened just a little while ago, and say that I disbelieved you. That is completely false; I never said that I didn't believe you. How the hell would I know what you do or don't do? I'm happy to own my own words, even the really stupid ones, but you seem to be keen on reading a lot of things into my comments that simply aren't there.
Anyhow, that's it and I'll drop it, but please just say what you have to say without falsely representing my views or comments. Thanks.
excuse me, sorry to interfere, but my god, children being educated in their SCHOOL and mom commenting on blogs about firebombing it is an inacceptable situation
if you don't find that people ding non-gender-standard people
Where did I ever say that? I meant what I said, and not some totally different thing. I said that you'd misremembered what the research said and that what you said was somewhat implausible. And I said that what the research actually said strikes me as quite plausible: people in male dominated fields tell stories about their field in ways that code male in their culture (and that this fact is common across cultures even when the codes are inconsistent) and that this make it harder for women to succeed in the field because people mental image of a successful practitioner is male.
Late to the thread, as always, and not very knowledgeable in game theory, but the popularized conception of game theory is far from where the field currently is. First off, the model of things like prisoner's dilemma and the chocolate game have been expanded to allow things like wise grandfathers (or some similar term), who can provide both players with 'guarantees'. I don't remember where that takes these things, but the models are more developed now.
Also, the similarity to chess noted by others above is because I think the origins (or at least some of the fundamental first theorems) did treat exactly such problems: this is how we know, for example, that for every deterministic full-information game (like chess), there is guaranteed to be a non-losing strategy for one of the players.
Also, game theory is (I think) applied in lots of stuff relating to computers, like protocols of communication where you suspect at least one of the computers participating has been sending false information (due to a glitch, or intentionally). How many agents and what communication process do you use in order to guarantee that the information is preserved, assuming some small fraction of misleading agents, that kind of stuff.
And even more at the intersection of computability and game theory, we find things like this, which basically shows that the current trade in derivatives is fucked (not that we didn't know that), since evil players can always game the system so that even if you have complete information, you could be sold fraudulent assets and it would be NP-hard to detect it.
And also in the category of bad models but really interesting and counterintuitive proofs, there's Arrow's impossiblity theorem. I already wrote too much so just read the entry if this sort of thing interests you.
414
Um, well I don't know what to say to that other than that this seems to me to be closer to support for politicalfootball's 399.2 than an objection to it: whoever is it that you've seen explaining the tragedy of the commons doesn't have the damnest idea what he or she is talking about. So, more universal teaching of the tragedy of the commons would help solve that.
The dominant meaning of tragedy of the commons today is that common resources like fishing areas which anyone can use are prone to being overutilized because mutual restraint is unstable. A commonly proposed solution is private ownership. I am vaguely aware that a few lefties claim this is all wrong but even if this is true this may be a case where the original meaning has been hopelessly lost.
People should call it what it is: tragedy of unregulated commons. There's nothing tragic about a well regulated commons.
And without a functioning regulatory state, privatization is not only ineffective, it's non-existent. Outside of the minds of people who've not gotten past hs sophomore levels of perception.
As urple notes in 418, the point of the tragedy of the commons is that unregulated common pool resources are likely to be overused. The two possible solutions that are generally posited are government regulation and privatization, which are not actually mutually exclusive; most of what gets called "privatization" in this context involves some sort of action on the part of a government or similar authority. When libertarians talk about the tragedy of the commons they usually just ignore the regulation solution and focus on privatization, but the concept itself just illustrates the problems with a total lack of any mechanism for controlling access to resources.
445
And even more at the intersection of computability and game theory, we find things like this, which basically shows that the current trade in derivatives is fucked (not that we didn't know that), since evil players can always game the system so that even if you have complete information, you could be sold fraudulent assets and it would be NP-hard to detect it.
I wasn't impressed with this paper when it came out and I haven't changed my mind. I don't think the paper has much real world relevance.
And without a functioning regulatory state, privatization is not only ineffective, it's non-existent.
Painfully obvious, but I think it's also necessary to say it.
Economics was lost as a legitimate discipline precisely because people stopped telling the truth about it. If game theory is no longer a reputable area of study, it's because people have endorsed the LB-x.trapnel-Shearer view, not because of some inherent flaw in game theory.
Even in these debased days, honest scholarship is still possible.
444, Unf; your last sentence is most of the way to what I meant by 'dings'; women can do things "male style", and in heavily gendered-male fields will do so in emulating the successful; but in heavily gendered societies, women acting male are dinged for it. It's a fork.
re commons: Garrett Hardin's book on immigration is a piece of work; rich countries shouldn't allow immigration because it will just let the numerous poor act like the destructive rich.
Unfoggetarian: "Pause endlessly, then go in" (9) != Unf AFAIK. Unf has gone to a better place where they don't blog.