PDA

View Full Version : I'll take Sklansky's bet, against any MIT student's real smarts


mosta
07-13-2004, 01:55 AM
Sklansky says (in this thread: http://forumserver.twoplustwo.com/showflat.php?Cat=&Board=books&Number=812875&page=0 &view=collapsed&sb=5&o=14&fpart=all):

Those who think that an MIT graduate would not be much more likely to correctly analyze the profitabiliy of poker plays, (if they put their mind to such a task), than members of the general population, are engaging in wishfull thinking. I am not talking simply math here either.

[and]

Simple precise question. You hear two people arguing about a poker play. One where the right answer can eventually be determined. You know only a few things about the debaters. One is a math grad from MIT who has made a study of poker and is quite certain he is right about this particular argument. The other person once wrote that the size of a jackpot should never change your strategy.

At this point you must make a price as to who is right about the argument. What would you say that is? And as long as you agree it is over 50% in favor of the MIT guy you cannot say I am making a personal attack or appealing to authority. (My figure is 98% by the way. What's yours?)
----------------------------------------------------------

I'd sell you all the 98 you want, Mr. Sklansky, and hit your next bid coming out, again your size. You have a profound and, to me, astounding misimpression of the nature of undergraduate education. Far and away the most significant characteristic inculcated by undergraduate "education" is obedience. Repetition, memorization, and imitation is the whole game in the vast majority of classes in any school. Howard Gardner, MacArthur fellow cognitive science researcher, studied junior undergraduate physics majors at MIT, Hopkins, and other top schools and found that the majority of them, when taken outside the context of a formulaic textbook problem (1. identify relevant equation from obvious cues 2. translate word problem term for term into equation 3. crunch numbers) and forced to think on their feet in unfamiliar situations, failed to apply even Newton's basic three laws of motion and fell back on the same intuitive Aristotelian wrong physics that ten year olds use. Anyone who's been in a top level PhD program in a technical field knows the best one can generally hope for from incoming students is that they know essentially nothing--that they haven't been permantently harmed. (Now of course there are complicated, analytical procedures that have to be learned, which are a foundation for a graduate student, no doubt, and are difficult and perhaps beyond the reach of many people. But those procedures are essentially mechanical. Learning them doesn't make one a physicist--someone who thinks physics to work through an unfamiliar situation.)

So if the bet was who would do better on the SAT, or some other crap like that, then by all means I like your side. But if you wanted to bet on who's going to make a better poker player one year in, an MIT undergrad or an options market maker who never went to college and never even heard of differential caculus, Bayes, or finite discrete combinatorics--well, I'll give you all the MIT guy you want with odds.

(Please note, I don't mean to take Lee Jones's side in any debate with Ed Miller, or imply that Miller's understanding of poker is in any way deficient (quite the contrary), nor that his understanding of anything he may have studied in college is--on that have no idea and no comment. I just mean to take issue with the particular ludicrous statement that an MIT student is a 1:49 favorite over a random guy at figuring out something new, unfamiliar, and complex on their own.)

Gardner, The Unschooled Mind Basic Books 1993.

p.3 a by now overwhelming body of educational research that has been assembled over the last decades. These investigations document that even students who have been well trained and who exhibit all the overt signs of success--faithful attendance at good schools, high grades and high test scores, accolades from their teachers--typically do not display an adequate understanding of the materials and concepts with which they have been working.

Perhaps most stunning is the case of physics. Researchers at Johns Hopkins, M.I.T., and other well-regarded universities have documented that students who receive honor grades in college-level physics courses are frequently unable to solve basic problems and questions encountered in a form slightly different fom that on which they have been formally instructed and tested. In a typical example, college students were asked to indicate the forces acting on a coin that has been tossed straight up in the air and has reached the midway point of its upward trajectory. ... 70 percent of college students who had completed a course in mechanics gave the same naive answer as untrained students: they cited two forces...

p.4 Students with science training do not display a blind spot for coin tossing alone. When questioned about the phases of the moon, the reasons for the seasons, the trajectories of objects hurtling through space, or the motions of their own bodies, students fail to evince the understandings that science teaching is suppposed to produce. Indeed, in dozens of studies of this sort, young adults trained in science continue to exhibit the very same misconceptions and misunderstandings that one encounters in primary school children.

p.152 Such students in fact get high scores on standardized tests of physics knowledge, and they are likely to earn honor grades when they are tested at the conclusion of a semseter or year of college physics. ... p.153 So described, the game sounds simple enough, and, indeed, both naive elementary school children and college physics students approach it with enthusiasm and confidence. Yet nearly everyone at both levels of expertise fails dismally. the reason, briefly, is that success at the game requires an understanding and application of Newton's laws of motion. To succeed, the player must be able to take into account the direction in which and the speed with which the Dynaturtle has already been moving. Whatever their formal training, however, players of this game reveal themselves to be dyed-in-the-wool Aristotelians. They assume that, so long as they aim the Dynaturtle directly at the target, they will succeed, and they are mystified when the KICK does not result in the desired collision. ... p.154 [of course it would be wrong to draw excessive conclusions from a computer game but] [this] behavior turns out to be quite typical of what is found when students with training in physics or engineering are posed problems ouside the strict confines of class--that is, outside what might be called the text-test context. Here are some further examples drawn from the large literature on the topic:

[objects propelled thorugh curvilinear tube are presumed to take a curvilinear trajectory upon exiting.]: "an object that moves through a curved tube acquires a "force" or "momentum" that causes it to continue in curvilinear motion after it emerges...After a while, this force dissipates and the trajectory eventually becomes straight.

p.154-5 [re more advanced students who have taken special and general relativity]: Students are able to repeat back accurately the principal claims of relativity theory, according to which ... . And yet students reveal in their responses that they in fact adhere to a belief in absolute space and time. Even a tutor in the course "shows a firm Newtonian commitment to a mechanistic view of the world, which require that objects have fixed properties such as length, mass, etc." Only when students and the tutor are made to confront the inconsistencies between the claims of Newtonian and Einsteinian models of the universe do they begin to engagte the problems in the proper fashion."

Richard Berg
07-13-2004, 05:12 AM
So people who study mechanics think like Aristotelians, and people who study relativity think like Newtonians. Clearly the answer is to put a philosophy course into the sequence after GR.

-RB, graduate of a top university in a technical field and student of Ro-sham-bo

kem
07-13-2004, 09:12 AM
[ QUOTE ]
Far and away the most significant characteristic inculcated by undergraduate "education" is obedience. Repetition, memorization, and imitation is the whole game in the vast majority of classes in any school.

[/ QUOTE ]

Memorization doesn't get you too far at MIT.. Can I go out on a limb and guess that you got rejected by MIT? You sound very bitter.

[ QUOTE ]
But if you wanted to bet on who's going to make a better poker player one year in, an MIT undergrad or an options market maker who never went to college and never even heard of differential caculus, Bayes, or finite discrete combinatorics--well, I'll give you all the MIT guy you want with odds.

[/ QUOTE ]

Can I get odds with a couple of my MIT friends? They each have a couple degrees and actually work as options traders. I'm pretty sure they understand Newton's laws too.

Blarg
07-14-2004, 12:26 AM
You guys attacking Mosta shouldn't take his quotations as proof of a character flaw or something. He didn't write the book either.

It's true that many things we do and give ourselves credit for thinking about involve far less actual creative thinking and far more recitation from memory and application of already learned rule sets than we by and large tend to think.

Some occupations have been intensively studied to determine the knowledge base and find out the "rules" used in decision-making processes that were once taken for granted to be the result of quite complex and even sometimes intuitive and creative processes, and that research has been used to build computer programs that have replaced people once thought irreplaceable -- not just assembly line workers, but white collar professionals. The attempt at making poker bots is an attempt to do the same thing, though really good poker is more complex than it looks.

The book Mosta quotes illustrates the mechanistic and relatively low level of thinking common not just to smart people or MIT-types in particular, but the entire human race Very little of what the average or even exceptional person does requires much deep thought or creativity. You learn the facts and the rules, and from then on, it's mostly a matter of whether you successfully apply them. It doesn't matter if you're doing brain surgery or tree surgery. Neither the need for nor opportunity for creative or original thinking arises much in life. Despite the potential difficulty of initially learning rule sets and acquiring knowledge bases, actually applying them is basically a mechanical process that often requires little in the way of judgment calls or thinking at all.

Mosta and that book are not attacking MIT, just asserting how surprisingly mechanical and limited the thinking processes of even very bright and successful people can be. Even if you can't or for some reason don't want to credit Mosta, it sounds like the authors he quotes have done a good bit of testing, and are reciting simple facts they've observed.