Hypothesis: IQ and Chess

** double sigh **

 I do not think this hypothisis is completely correct.

Hey Paul211,

Yes, I agree that it might be a good thing to get your 5 y.o. child a professional IQ test.  It may help you to understand your child's development needs, and would give you a valuable cross-check on information developed by teachers or school administrators. 

You also seem to be saying that childrens' intelligence is less "biased" or contaminated by prior learning than is adult intelligence.  I think I understand what you mean, but I think there is another interpretation here (which is actually the interpretation held by most cognitive scientists).

The other interpretation is, in a nutshell, that there are two great kingdoms of intelligence: what scientists refer to as "fluid" or "performance" intelligence, and "crystallized" intelligence.  Fluid intelligence is involved in responding to novel situations or problems, not encountered in the past.  It can be considered the more fundamental (or what software engineers would call "lower-level" machine language) intelligence.  In common with crystallized intelligence, fluid intelligence is symbolic reasoning, only the items being reasoned about, and the solution process itself, have not been encountered before.  Thus fluid intelligence relies more upon speed, flexibility, and the raw capacity to retain temporary information so that it can be manipulated. 

It could be theorized that children have primarily fluid intelligence, since they haven't learned very much either in terms of facts about the world, or in terms of problem solving approaches or logical reasoning procedures.  As children, and people generally, acquire both factual and procedural knowledge over long periods of time, as well as a rich library of distinct symbolic primitives (words and numbers) that can be used to represent objects in the external world, they begin to acquire another kind of intelligence -- the crystallized intelligence. 

Both kinds of intelligence contribute to intelligent behavior and to the ability of IQ tests (which test both fluid and crystallized types) to predict outcomes in school, work, society, and in health.  Both kinds of intelligence also develop and become stronger up to about age 25, when fluid intelligence peaks (after age 25, performance on speeded, memory intensive, and attentional/perceptual "fluid" test items flattens, and then, starting about age 40, begins to decline).  Crystallized intelligence is tested  with straightforward items like verbal analogies and vocabulary recall, though it actually operates in the real world with far more complicated factual and procedural constructs, especially in intelligent people and people with jobs involving complex analysis or calculation -- or, for that matter, chess masters with retained repertoires of playbooks.  Crystallized continues to increase until about age 60, when it too begins to decline.

The important point here, Paul, is that both crystallized and fluid intelligence are instrumental to intelligent behavior and the rewards that come from it in modern economies.  Crystallized intelligence is not a bias in IQ tests; it is a fundamental form of intelligent behavior. Both types contribute to the latent, general factor of intelligence called "g".

CB

too..many...words...brain...hurt

I find it highly suspect that everyone who enters in these kind of debates, including myself, claims to have over 130 IQ as measured by a scientifically reputable test. It'd be interesting to see people furnish proof, though it's extremely likely that, as the TC said, you are only given a percentile and not a raw score.

Also, ozzie, I doubt there's any way to stop these silly faux study hypothesis threads, or even debate them because they're loaded with so much jargon as to be nigh incomprehensible without expending quite a bit of effort when clear english might have been more beneficial given the target audience.

In the end, we can all pat ourselves on the back for being able to score genius-level IQ on the test of your choice, however it's highly unlikely that anybody here will achieve a GM title(statistically speaking), nor is IQ or chess ability going to correlate on the actual important things in life. See Kamsky's recent personal troubles as to why he wasn't at his best during the Topalov match.

Goodsoul --

You have a great idea about ways of testing IQ (and other mental ability) correlations with IQ.  Thanks.

I do know there are a couple of places that offer valid IQ tests online (one of them is www.wonderlic.com), with safeguards to ensure test security and the identity of the taker.  These safeguards aren't terrific, because they involve going to a test center, but possibly, proctoring arrangements could be worked out with librarians or chess clubs or something.

Another approach would be to have chess players participating in the study submit prior test scores via their testing services or schools or the military.  The SAT, MAT, GRE, LSAT, WAIS-III, Ravens, Stanford-Binet, and many others from Europe are de facto IQ tests and can be interconverted in terms of scores, so many people have IQs at their fingertips.  This is what some of the "high iq" societies like Mensa and ISPE do, except when someone is willing to sit for a new test offered by the society.

Thanks again.  I would like to see the hypothesis tested.  And, I'm prepared to see it go down to defeat.  We would all learn something either way.

Regards,

CB

We could do without this self serving topic...

I think some people are misconstruing the hypothesis I put forward.  For example, one poster claims an IQ of 140, and then says that "I'm not some grandmaster".

As should be made clear from the case of Kasparov, extremely high intelligence is NOT required to attain the highest levels of mastery in chess.  Kasparov as 1 in 100 IQ, which is very superior, but he has 1 in 1,000,000 chess chops.

I think an IQ in the range of 125-139 is sufficiently superior, in the presence of prolonged and intensive practice, as well as extraordinary situational memory ability (which is distinct from IQ), to achieve grandmastery.  Thus it should not be surprising that people with genuine IQs of 140, 150 or 160 would not all be grandmasters.  Chess is largely an expertise that intelligence HELPS one acquire. 

CB

While I PMd you about this, maybe others will find it interesting too:

"...your hypothesis is based on the dubious assumption that IQ is actually intelligence. In fact, it is no more than an incomplete approximation at one point in time. For example, "idiot savants" who can play the piano very well definitely have a specialized form of extreme intelligence in a direction which is not likely to be revealed by IQ tests. This is less applicable to chess due to the diversity of thought processes needed but I'm only making a point."

Great caution needs to be observed when meshing intelligence and IQ in a coherent framework. They are correlated but not equivalent. For instance, this statement seems sloppy: "As should be made clear from the case of Kasparov, extremely high intelligence is NOT required to attain the highest levels of mastery in chess." First of all, it's not clear which test to believe, the 185-190 or the 135. Secondly, categorial statements cannot be made based on such a small and varied sample size. People are not consistent enough (poor test takers or bad days) to justify such certainty. Of course, the downside is that a reliable sample size may make the tests less valid due to familiarity (retest bias). Thus, a diverse range of IQ tests would be necessary. Lastly, even if Kasparov's IQ is indeed closer to the 135 range (which some already consider extremely high), it does not preclude other phenomenal qualities not measured by the test. IQ != intelligence and equating the two by implication is fallacious. The most you can do is make a reference to IQ. Still, even this is not totally accurate unless a particular test is specified. After all, an IQ value on the Cattell tends to be different than one on the WAIS or Stanford-Binet - even for the same individual. This is somewhat like how chess ratings OTB are not the same as those on chess.com, chessclub.com, playchess.com etc.  Moreover, different IQ tests have different scoring ranges and hence may not even be applicable for some extremely high or low scoring individuals. It's evidently a complicated issue.

BTW: Little is absolutely required but much becomes pratical necessity. For example, doctors don't have to have at least normal IQs but yet their average IQ is in the top 1%. While there are certainly exceptions, they do not make the rule. Thus, if there exists a qualified doctor with an IQ of ~75 (likely an indication of a faulty or improperly administered IQ test), it still wouldn't be responsible for me to tell another person with an IQ of ~75 that the odds aren't stacked against him becoming a doctor.

Exigentsky - I agree with many of your points:  different scoring systems used on IQ tests, etc.  In the confines of our newsgroup-style format, though, and given the nonspecialist audience, parsing the extensional meaning of IQ,  or debating  the construct validity (what is the model of intelligence assumed by IQ test authors) may get us into an interesting and informative conversation, but probably not the one we started with, which was the relationship between IQ-referred intelligence and chess achievement.  I do give you credit for PMing me on this, and I apologise for failing to see that.

Your comment about basing an argument on a single chess master's IQ score would be valid, certainly, if I were attempting a proof of my hypothesis.  But I am not, and I was not.  I was merely using Kasparov's score as an illustration of why I am putting forward my hypothesis.  There is no scientific obstacle here that I'm aware of.  OTOH, if your criticism has more to do with the style of my newsgroup writing, then I do accept it.  I don't always like it when others use the word "clearly", so it is good to take a second look at my own usage of that word.  Thanks for pointing this out.

As for your last, but I think most important, point:  quite a high IQ (or the intelligence referred by it, anyway) is a practical necessity in may fields, I completely agree.  So is Kasparov supporting this view, or not (at a WAIS-III equivalent IQ of 135)?  I believe that beyond IQs of 125-139, IQ-intelligence ("g") increases contribute incrementally less and less to chess mastery, while the contributions of crystallized expertise and specialized (non-IQ) talents like situational memory come to the fore, and differentiate the truly profound chess players from the merely great chess players.  Of course I could be wrong, in which case IQ's might contribute linearly or even extra-linearly to chess mastery all the way up the line, as per the Levitt formula.

As you suggested, physicians and attorneys have average IQs in the 130 range, but are the top 10%ages of physicians and attorneys defined by the top 10%ages of their IQ distributions?  There are good reasons to believe that the answer is "no".  One of them is that the predictive validity of graduate school admission tests are almost uniformly lower than the predictive validity of undergraduate admission tests.  Likewise, pre-employment thinking tests have greater predictive ability for first line management and technical jobs than they do for senior executive and advanced technical jobs.  All these exams are IQ surrogates and some are even more "g"-loaded than some tests that openly refer to themselves as "intelligence" tests (the Miller Analogies Test is a case in point; it may as well have been called the "IQ-test-for-graduate-schools-who-don't-want-to-call-it-that".  ).  These considerations are the logical basis for the hypothesis I have presented in this thread.

Thanks for your comments, exigentsky (gotta love that screen name!)

Regards CB

Dear AumTerre --

Thanks so much for your interest in exploring the working memory - chess hypothesis. 

If you're interested in moving forward on something that would be worthwhile, I would ask you hold off on any study-related formal trainining using the program for now.   The training effect on chess performance could be rather subtle, so we will want to maximize the power and sensitivity of the statistical tools available to us.  This requires careful study preparation, and by-the-book protocols and design of experiments.  Any comments or suggestions you have in this regard are very welcome.

In our study, we will undoubtedly either conduct:

a) before and after experiments, in which individual chess players' ratings are taken before training, and again after training.  This would include a "placebo" group whose chess ratings are tracked, but which does not get 2-back training;

b) treatment versus control group experiements, in which a treatment group receives the  2-back memory training, and the control group, nothing.

In the meantime, no harm would be done if you were to "play" with the memory program, think about how you'd like the study done, etc.  The main concern I'd have for anyone who wants their personal progress included in the study results is that significant training before the study actually starts will tend to limit the treatment effects that our study can detect.

I TOTALLY agree on the necessity of passionately devoted practice and study in order to achieve mastery in virtually anything.  It is a common misconception that great musicians, scientists, chess players, and the like get where they are primarily on the basis of their talents.  In fact, as almost any highly accomplished individual will testify, it was intensive and prolonged practice that won the day.  Perhaps talent (and this includes IQ-general intelligence) is something that in part makes it possible for people to tolerate focussed and repetitive practice; to be more annoyed than most by flaws and defects in a performance or result; and to appreciate, more than most, the naked beauty of a perfect or near-perfect performance.

Cheers,

CB

Contrablue, I believe the connection between so called"genius" and chess mastery may be less than you think . Look up the Polgar sisters and their fathers theory about "genius".

this is one of the greatest threads on this site!  im gonna put my two bits in...i tied for first in my senior class on the ACT scores, the valedictorian and salutatorian both played chess  (i placed 33 out of 356 in my class overall).  their IQ'swere reportedly 145 and 150, mine is an estimated 125...i whipped up on them both every time we played, why because i studied chess and they studied for school!  Intelligence helps no question there but its not the end all, you need knowledge to go with your smarts!  im not a genius by anyone definition but i paid attention to what i read.  I was the best chess player in my school after reading one chess book!  so my hypothesis is that IQ + knowledge = good...in college i played a guy whos IQ was reportedly 165 and he couldnt play his way out of a paper bag!  but talk to him about philosophy, mathematics, theology, chemistry, psychology, biology or any other erudite subject and he'd blow me out of the water...Fischer said that he could take anyone of average intelligence and turn them into a master, so in his opinion knowledge and training were more important than IQ, thats one of the few things he and I would agree on!

I don't believe there is a relation between IQ and chess skill.  I have met highly intelligent people who are poor chess players, great chess players (state champions) who are morons, and what about idiot savants?  A person could have an IQ of 70 and yet be a natural chess genius. 

I'm going to have to take issue with the heavy use of the word "predict" in this thread. I've looked through the referenced APA article on IQ mentioned as definitive. It's use of the term predict is in the context of correlation. This is technically true in a statistical sense, although statisticians tend to shy away from using such words. That's because they understand that correlation is not causation. In a colloquial setting (like an internet forum), predict contains a sense of causation that is not present in correlation. A correlates with B does not mean A causes B. It might be that B causes A. It might be that C causes both A and B. It might even be a coincidence.

High intelligence might cause better chess skills. Or learning chess might cause better intelligence. Or a hard work ethic and supporting environment might cause both.

Without being able to show a causation between IQ and chess (and I'm not sure where a correlation has even been shown, rather than assumed), I would also have to question the validity of any hypothesis stating that IQ limits chess ability.

The question re idiot savants and strong chess play has been brought up before. We can find accounts of mathematical and musical savants who, say, can't tie their shoes but I can't find any instance of even a  moderately strong (let's nominate a USCF Expert level) chess player who is otherwise an idiot.

Why? Who's the strongest such person you ever heard of? I'm into chess trivia and I can't recollect anyone at all.

Maybe it's that chess is knowledge based and learning isn't something the savants do. Maybe the social aspect of the game screens them out from even starting?

Anyone have any thing here?

At the lower levels, I'd agree since cognitive capacity limitations are only evident once the necessary work has been put in for sustained periods. For example, anyone can get over 1000 USCF just based on studying a little more and understanding the basics. There is no intellectual barrier. However, going from Expert to Master or Master to International Master is not even in the same realm. In a sense, one can ONLY make limited positive extrapolations based on chess skill since high accomplishment is difficult to achieve by anyone, regardless of work put in, but low accomplishment can be explained by a multitude of factors unrelated to intelligence. Regardless, comparing players' intelligence in any respect directly based on chess skill is dubious unless they have had EXACTLY the same background as well as amount and type of chess instruction. In other words, it's not possible to be scientific about this. In some ways, this applies to IQ too. IQ is a reflection of practical intelligence in the limited categories tested rather than a measure of raw intelligence. IQ testing does not take into account a person's background (education, practice in similar environments, etc.). In fact, all current psychometric testing can only show practical ability, not innate potential.

I'd recommend everyone interested in this thread to read these: Relationship between Chess ratings and IQ and Becoming a GM, too late? There are many insightful posts and my thoughts on this subject are explained in much more detail there.

BTW: I've never met a "dumb" chess player USCF Class A or higher. I've met a few stubborn ones with deplorable opinions, but this is not the same. The idiot savant idea as it related to chess is suspect for the highest levels of performance.

ill make this quick

"and an extremely large fund of chess expertise and knowledge"

this statement here that you said previously relating to being good at chess would be the logical equivalent of saying 'running fast' to be good at running.

is then your statement not also a tautology? to be good at chess you need chess expertise!

im not saying your hypothesis is wrong, just that it is an obvious one that people have know for a long time.