Skip to content

Debunking I.Q. Test Claims Discussion (Parts 2 & 3)


Author(s): Chris Cole, Richard May, Rick Rosner, & Scott Douglas Jacobsen

Publication (Outlet/Website): Noesis: The Journal of the Mega Society

Publication Date (yyyy/mm/dd): 2022/10 (Issue #210)

Chris Cole is a longstanding member of the Mega Society. Richard May is a longstanding member of the Mega Society and Co-Editor of Noesis: The Journal of the Mega Society. Rick Rosner is a longstanding member of the Mega Society and a former editor of Noesis: The Journal of the Mega Society.

They discuss: major warning signs of something awry; the minor, or subtle, warning signs; 4 standard deviations above the norm; the successes and failures of the Mega Test, the Ultra Test, the Power Test, and the Titan Test; 4 and 5 sigma above the norm; the principal design of the Adaptive Test; other extraordinary high-I.Q. societies; associative horizon; the Mega Test; the claims about the Mega Test; legitimate testing; extrapolations well beyond the norms of the mainstream tests; the motivation behind making claims well beyond the norms of the most used mainstream I.Q. tests; the more egregious I.Q. claims in 20th century; and the big lessons in debunking phony I.Q. claims.

Keywords: Adaptive Test, Aleph Society, Chris Cole, debunking, I.Q., intelligence, Keith Raniere, Marilyn vos Savant, Mega Society, Mega Test, Power Test, Richard Feynman, Richard May, Rick Rosner, standard deviation, The Plurality IQ Society, Titan Test, Ultra Test.

Scott Douglas Jacobsen: You have all been around the block. Your membership in the Mega Society has spanned decades. So, you’ve seen controversies, failed high-I.Q. societies, and proclamations to this-or-that I.Q., even individuals who spun off into fraudulent activities, messianic posing, and criminal behaviour. As a note on collectives of high-I.Q. people, when it comes to claimed high-I.Q. societies, what are the major warning signs of something awry, not quite right, with it?

Richard May: The major warning signs of statistical and psychometric incompetence, fraud, or madness are usually quite subtle. Please see below.

Rick Rosner: You got to start with the disclaimer that most people in high-IQ societies are well-behaved relatively normal people who like taking tests and solving puzzles, and there are only a few lunatics. And because the ones I belong to don’t get together very often, you don’t have a chance to see any warning signs developing.

Although, in the case of one guy from many years ago, you could see a guy who was kind of being physically dominant and, I guess, mentally dominant getting increasingly frustrated that people didn’t understand him or believe his theories. So, it was just an increasing belligerence or pre-belligerence.

I guess, a skosh of megalomania.

Chris Cole: The major warning signs are the ones you list: fraudulent activity, messianic posing, and criminal behavior.

Jacobsen: Following from the previous question, what are the minor, or subtle, warning signs?

May: I get slightly suspicious if someone comes up with the most brilliant Theory of Everything ever, explained in a newly invented language of neologisms, which only the inventor of the theory himself can understand, especially if the theory makes no falsifiable predictions and none of those few who claim to understand the theory can explain it in their own words. I’m also slightly suspicious of, e.g., taxi cab drivers or barbers, who have conclusively proved Einstein’s theory of special and general relativity wrong.

If someone claims to be the most intelligent person in the history our solar system or to be the actual God of the Bible, then this level of measured intelligence may be beyond the current development of psychometric science, even with the Flynn effect. I’m probably too skeptical sometimes.

Also, branding of one’s associates by high-IQ types is often unnecessary in my view.

Rosner: Again, I don’t hang. I have no basis or nothing to talk about regarding this. It is not like I was living with a high-IQ person who slowly went crazy, besides myself. Really, in the last few years, I’ve gotten less crazy, more lazy. Lazy has replaced crazy.

Cole: The minor warning signs are incredible IQ claims.  As a rule of thumb anything above five sigma is not credible as is anything that has not been normed using regular statistical methods.

Jacobsen: Why is 4 standard deviations above the norm (e.g., mean 100, S.D. 15, I.Q. 160) such a difficult barrier to break in finding highly intelligent individuals?

May: Almost no one in the alleged “real world” is interested in measuring intelligence beyond the 4 sigma level. Where would you find a large sample of individuals beyond the top 1-per-30,000 level of intelligence to study? This level of intelligence is not a target level for standard IQ tests developed by psychologists. Why should it be? Which professions require IQs beyond the 4 sigma level? Even Nobels in physics probably depend more upon a mathematical ability sub-factor of general intelligence than upon super-high IQ per se. Two physics Nobel laureates didn’t qualify for inclusion in Lewis Terman’s study of the intellectually gifted, because their IQs were not sufficiently high! In addition Nature may sometimes not be ‘politically correct’. What if cognitive differences were discovered among various human sub-groups? For example, what if a growing number of trans-species individuals, who identify as advanced AI units, were found to be better at arithmetic addition?

Rosner: Several reasons, one, there aren’t that many people. 4-sigma level is one person in 30,000. Although, in real terms, it’s less rare than that because the average IQ of people on the street is like 105 or 110. The people with IQs of 35 are institutionalized. You don’t see them around. It’s rare. That’s one problem.

Problem two, it is hard to test. All the good high-end tests take dozens of hours to do well on. Thing two-and-a-half, many people who might score well on them might be successful and may not want to waste their time putting in 40 or 50 hours in something that doesn’t compensate them.

They could be trading stocks or coding or doing business deals or getting laid. None of which taking an IQ test helps.

Cole: High range tests require high range questions which are hard to create. Plus there is not much of a market.

Jacobsen: What have been the successes and failures of the Mega Test, the Ultra Test, the Power Test, and the Titan Test in identifying highly intelligent persons – despite being compromised?

May: There is evidence that uncompromised tests work better.

Rosner: Maybe, some smart people still trickle in. The Mega Test has been compromised since, probably, the late ‘90s or the internet made it possible to contaminate the questions by throwing around answers in chat rooms.

The Mega Test was the most successful in finding high-IQ people because the most people took it when it was published in Omni magazine. 4,000 people took it. It’s more than any other test ever.

Which means, though, more people have taken the Hoeflin tests than tests by any other author, though probably a strong second and possibly somebody who has overtaken Hoeflin because he has written dozens of tests is Paul Cooijmans, who has been writing tests for decades and has cranked out quite a few.

Some of his tests have certainly been taken by more than 100 people. In the aggregate, thousands of people must have taken Cooijmans tests. With the success of the Hoeflin tests, they have found, depending on the cutoff, hundreds of high-IQ people.

Some of those people got together and some people were mentored by other high-IQ people, and had their lives improved, including myself. So, the success of the Hoeflin tests is the large numbers of people who have taken them.

For years, I, and sometimes with partners or being asked to consult, pitched TV involving high-IQ-type competitions. The same kind of shit as Project Runway or American Idol. A talent search, but instead of for fashion designing or culinary skill or singing skill, it was for raw intelligence.

This is an idea that comes to people not infrequently, but just has never been turned into a show. But if you had a show that did that, that would be the most successful project ever to find high-IQ people because millions of people would see the show and tens of thousands of people, if there were high-IQ tests associated with the show, would try those tests.

But that project has never happened, which I think is stupid because reality shows are about following assholes around with cameras and there are plenty of high-IQ assholes. Not as a percentage of high-IQ people who are, as I said, mostly decent, normal-ish people.

But if out of 100 people who have managed to score 160 on an IQ test, there are probably a half-dozen who you could productively, entertainingly follow around with cameras.  

Cole: First of all Ron Hoeflin is a talented question framer.  Next he spent a lot of effort validating his questions.  Finally he normed them several different ways.

Jacobsen: In principle, what is realistically needed to test between – let’s say – 4 and 5 sigma above the norm, reliably and validly?

May: Perhaps advanced AI can be used to develop significantly improved high-range intelligence tests. Other neurobiological methods of assessment of the general factor of intelligence, ‘g’, may eventually make IQ tests obsolete. For example, measures of biological traits such as pitch discrimination ability (of sound frequencies), among other such physical measures, have been found to have surprisingly high correlations with general intelligence. This may be the way of cognitive ability assessment in the future.

Rosner: You need experienced test-builders. You need a decent amount of people to norm the problems on, to make sure the problems can actually measure high-IQs. You need their other scores to see what scores getting those problems right correspond to.

As I said, you need some kind of widespread exposure. You have to let hundreds of thousands of people know that the test exists. Ideally, that it’s something fun and/or cool to do.

Another condition is that it would be really, really helpful if the test took less than 20 hours to take. It would be helpful if someonecould spend 20 hours or 10 hours on the test and score near the ceiling, which is not a common thing among these tests.

Cole: To avoid spoilage you need question schemas, not single questions.  Then you need a way to automatically collect many samples.  Presumably this would be on the Internet.  A group of Mega members is working on this.  Contact me if you’d like to help [Ed.].

Jacobsen: What is the principal design of the Adaptive Test, inasmuch can be stated at this time? (Is this series the first announcement of the test, by the way?)

Cole: Cf  There are some articles in Noesis.  Let me check with the team.

Jacobsen: What other extraordinary high-I.Q. societies have been observed by you – the highest, most inclusive, most exclusive, the most multi-planetary, least reliant on D.N.A. prejudice, most non-carbon-based, und so weiter?

May: The Plurality IQ Society

Top 0.0000000000000000000000000 … % of Multiverse

Previously the highest-IQ group founded was the Aleph Society, which sought to have at most fewer than one member per Multiverse potentially qualifiable. However, the Aleph is found to be insufficiently selective in its admissions criteria for several reasons. First, it only considered 3 dimensions of space and 1 dimension of time per universe. We feel that it is necessary to include all theoretically possible multiple dimensions of spaces and of times per universe of the Multiverse. (For multiple-time dimensions see, e.g.: , , ysicist-states/ )

Secondly, the Aleph only sought the highest IQ ‘individual’, including AIs, in the Multiverse ‘now’, i.e., at only one point in ‘time’ relative to one (1) observer, the Wormhole Officer (formerly called the Membership Officer). To remedy this we ‘now’ recognize that to whatever extent possible technologically, the Wormhole Officer must be a time traveler.

Thirdly, it is not sufficient that our psychometric instruments selecting at the Aleph level be culture free. Our IQ tests must also be genome free, i.e., free of any genetic influences upon performance. Speciesism is even more common than racism and gender-bias. We seek genetic justice in our member selection testing criteria. For example, in the past and even today, species with brains are unfairly advantaged over species without brains, including, of course, AIs. Why should an Isaac Newton have an IQ advantage over a slug, simply because a Newton has a brain? This obvious bias must be eliminated.

NB: All of the non-members of the Plurality IQ Society are Full Non-members and Official Non-members.

Jacobsen: What is the system of thought or the psychometric philosophy behind associative horizon?

Rosner: In my mind, when you get hit with a hard problem, one that might take more than ten hours to figure out. Part of it is how many different angles can you come up with on the problem. How many parts of life can you apply? How many possible analogies can you apply? How many keys are on your key ring to approach the problem?

When he talks about associative horizon, it is how many associations can you possibly come up with, with the symbols or whatever, that constitute the problem. To some extent, taking one of these high-range tests is profiling the author, trying to figure out, maybe, them, Hoeflin problems have a Hoeflin flavour to them, let you know if you are on the right track. Other test makers have flavours similar to them too.

It may be similar to their culture, say. The person building the problem found something in their world and boiled it down to an analogy. There is a popularish puzzle that is 7 d in a w.” You have to figure out what the “d” and the “w” are. It’s ‘days in a week.’ The problems can get tough. Another easy one. “5,280 f in an m,” ‘feet in a mile.’

So, “106 billion p who e l.” The “e” “l” is tough. You have to figure out. It is ‘people who ever lived.’ So, for a lot of IQ problems, they have at least some aspect of that. Decoding, figuring out what the symbols represent. Then it is an exercise in figuring out what could the “p” and the ‘p in e l’ stand for.

“6*10^23 As in an M.” My numbers might not be right. But ‘atoms in a mole,’ it is a test of cultural literacy. Often, there is further manipulation done to the symbols, so you have to work through two or three transformation or link two or three transformations to figure out the problem. It is how much cultural literacy do you have or do you give yourself, and then the flexibility for combining these things.

It is how much different stuff can you bring to bear on a fairly obscure or convoluted problem.

Jacobsen: How did you first come to find the Mega Test?

May: Actually I don’t remember. It was about 40 years ago. I probably met Ron Hoeflin through my membership in the Triple Nine Society. This was probably my initial connection to the Mega Test.

Rosner: Some guys in my dorm told me about the Mega. I must’ve already been IQ braggy. Yuck.

Cole: Saw it in Omni Magazine.

Jacobsen: What were the claims about the Mega Test – and your score(s) in each section on it – by Ronald Hoeflin, the media, and others?

May: Ron Hoeflin told me that I was the 2nd person to obtain a perfect score on the 24 verbal analogies, I believe. I think Marilyn Vos Savant was the first. I certainly didn’t tell many people, beyond my girl friend. I remember showing a copy of the Mega Test to one young woman, thinking she might be interested. She just laughed and laughed. Neil Blincom of Mr. Pecker’s original, illustrious National Enquirer tried to interview me once when I was Membership Officer of the Triple Nine Society. I pondered this offer deeply for a fraction of a second. I remembered Chris. (never forget the decimal point) Harding’s interview, “World’s Highest IQ Genius is an Unemployed Janitor” and decided not to be interviewed. I avoided the media.

Rosner: So, the claims were the Mega was the world’s hardest IQ test. By hardest, having the highest ceiling, the score a perfect score would get you, for instance. I think after the sixth norming, after Ron looked at 4,000 test submissions that came through Omni. I think the ceiling became 190 S.D. 16 or a little over 5.6 sigma. The first time I took it, I got a 44, which was 23 verbal problems right and 1 wrong and 21 math right and 3 wrong. I took it a second time and got a 47, which was 1 math wrong, I think. It doesn’t matter whether math or verbal; I got 1 wrong the second time.

What does that translate into for me, after the fourth or fifth norming, my 44 wasn’t high enough to get me into Mega. Marilyn herself turned me down for admission. My score might have corresponded to 172. Then after the sixth norming, after all these scores came in, I think a 44 got you a 180. I think the Mega cutoff is a 176. There you go. The 1-in-a-million level. Next question.

Cole: Omni called it the “world’s hardest IQ test.”   Interpretation of scores can be found in Hoeflin’s normings.

Jacobsen: How does the internet complicate legitimate testing in the high-range?

May: The internet facilitates cheating on tests and meeting other cheaters to work with.

Rosner: The Mega came out in ’85. The Titan, the sequel to the Mega, came out in ’90. Most people got on the internet in the mid-to-late-‘90s. For those tests, it complicated and contaminated them because people went on message boards and threw answers around. Some of which were correct. That was problem one. Problem two was once Google came along; you could put in the words to the analogy and the fourth word would pop up. The analogies were half of the Titan and the Mega.

The 24 verbal problems were all analogies of the type “find the fourth word.” Most of those could be instantly solved using a decent search engine. Tests are different. The Cooijmans tests, which I consider the most challenging of the internet era tests can’t simply be solved by plugging things into a search engine. You still have to figure a lot of shit out. The most general issue with these tests and the internet is just sharing answers. Beyond that, it is a pain in the ass to make sure that the problems on the test can’t be solved through easy searches.

Chris (Cole) and his group of people, who are working on this test that are resistant to having answers shared, are working on tests that give each test-taker the same general problem, but the specifics of the problem are fresh. So, somebody else’s answer on this problem is not going to help you because, even though the problem should score the same – getting it right should reflect the same IQ level, you can’t just post what you got on answer 12. They’ve been working on that for well over a decade.

It’s coming along. Anyway, next question.

Cole: The Mega and Titan tests have been spoiled on the Web.  The Power and Ultra tests are at risk.

Jacobsen: Some, in fact more than a few, claim extrapolations well beyond the norms of the mainstream tests, e.g., the WAIS and the SB, which cap out at or around 4-sigma. Assuming legitimacy of the claims, then, the individuals would be highly intelligent, but the claims can range between a little over 4-sigma to 6-sigma. How is this extrapolation generally seen within the high-I.Q. communities at the higher ranges?

May: I don’t know how other others generally perceive unsound or bogus extrapolations of IQ scores.

Rosner: I think the skepticism of super-high scores is generally more for specific claims than for the entire idea of being able to have an IQ that high. I think most people in the high-IQ community believe it is possible to have an IQ close to 200. But I think most people also have a reasonable idea of the rarity of scores like that. Adult IQs, the deviation scores, are based on a bell curve, where between 0 and 1 standard deviation, you have 34% of the population in a bell-shaped distribution for something like height. Between 1 and 2 SDs, you’ve got 14% of the population. Between 2 and 3, you’ve got about 1.5% of the population. Between 3 and 4, you’ve got roughly one-half percent of the population.

Let’s see, about 4 SDs, that’s only one person in 30,000 should score above 4 SDs. One person in 3,000,000 above 5 SDs. What is it? 1 person in 750,000,000 above 6 SD or so; somewhere, I’ve fucked it up, according to the standard bell curve. People also like to say that at the very far ends; there are more outliers than on the normal bell curve. That there are more high-IQs than would be given if it were a perfectly bell-shaped distribution.

But even so, you shouldn’t see more than a half-dozen or ten or twelve or whatever, people, with scores above 6 SDs. So, Paul Cooijmans has the Giga Society, which has 7 or 8 members. It is for people with IQs that are supposed to be one in a billion. So, there are 8 billion people on Earth, 8 members of the Giga Society, so that makes a certain sense, but not really. That’s as if everybody who could score at that level has taken one of his tests. That’s just obviously not true. So, way too many people scoring at the one in a billion level. It’s not like the Giga Society has 300 members.

Cooijmans is pretty rigorous in his norming and testing. So, if you have taken a Cooijmans test and scored at or close to the Giga Society, legitimately, Cooijmans has written in the past about people’s attempts to cheat on his tests, but I don’t think there has been a successful attempt in decades. So, people are pretty accepting that if you get a Giga level score on his tests; that you’re legitimately pretty smart. The claims of super high-IQs, there are legit claims based on performing well on ultra-high IQ tests or kicking ass as a kid on a test like the Stanford-Binet or the Wechsler. Someone can say, “As a kid, I scored a 200,” or something.

That’s another thing I won’t go into. People who claim high-IQ scores and are lying are generally not sophisticatedly lying. They’re saying something that cannot hold up at all. I don’t know if there are many or any sophisticated lies about having a super-high-IQ. So, then there are people outside the high-IQ community who are skeptical about the whole thing, but no one is really worried a lot about it, because: who gives a shit?

Also, if you want to say something, or know something that I’m not aware of, that contradicts what I’m saying, go ahead.

Cole: Hoeflin’s norms all involve some extrapolation.  I find it reasonable up to the mega level (about 4.75 standard deviations).

Jacobsen: Following from the previous question, what seems like the motivation behind making claims well beyond the norms of the most used mainstream I.Q. tests?

May: It’s a shame Einstein did physics. He could have been on Facebook (now called Meta, I guess).

Rosner: Going off my own experience, I kind of felt like a loser based on when I was about 20. I’d fucked up a lot of opportunities for myself. Then somebody told me about the previous world’s hardest IQ test, which was a Kevin Langdon test. It ran in Omni or Games Magazine. I took it and scored 170. I went, ‘Wow, that’s a good score.’ When Mega came along, I took that. I liked that validation that it gave me. Even though, it is a ridiculous thing. I kind of feel like it might be analogous to a guy who can bench press 500 lbs.

It’s kind of a goofy thing. You wouldn’t tell that guy it is goofy to his face, but the Sven Magnason. He is 6’4” and weighs 310 lbs. and eats 200 grams of protein a day to get that or support that huge bench press and has hypertension and his joints will be fucked in 10 years. It’s a kind of a goofy thing. It is amazing the guy can bench 500 lbs. It is this ridiculous thing. It is a very obscure sport. Sven Magnason is not playing in the NFL for 1.8 million USD a year. He probably works in a warehouse and does strength training on the side.

It doesn’t translate into the kind of fame or success that you might want. So, it is a niche kind of sport.

Cole: Vanity is one motivation.

Jacobsen: What are some of the more egregious I.Q. claims in 20th century by groups and by individuals? This is a free forum.

May: In the 20th century — maybe being the smartest man in America was a fairly egregious claim. Top 1 per billion high-IQ societies may qualify if such came into existence in the 20th century.

Rosner: I don’t know. Anybody can go on the internet and type whatever they want. One of the craziest claims I saw I mentioned before. Somebody had a site or has a site claiming Jesus had an IQ of 300. The idea that somebody with the deep wisdom of Jesus meant Jesus had a huge IQ. His estimate based on nothing: If smartest people have an IQ of 200, then Jesus must have an IQ of 300. William Sidis, people claim 259 based on extreme achievements as a young person, at least it is based on his history and is a fairly earnest attempt to estimate a very smart young man’s IQ.

It is kind of egregious and not based on him being tested. Oh! Some of the most egregious are in the last 15 years; some insane moms, one mom out of Colorado, maybe 18 years ago, got a hold of the answer key to an earlier edition of the Stanford-Binet. Stanford-Binet gets revised every 15 or 20 years. I don’t know. You can still find psychologists who will give an earlier version. In the stacks of libraries. Probably, the Norlin Library at the University of Colorado, she found an earlier editions, found an answer key. Then taught her kid all the answers, so, that kid scored, at age 3 or 4, like a 10-year-old, which, the way they calculate childhood IQs, gave him an IQ well over 300. She tried to get herself and her kid famous off this.

It, eventually, fell apart because the kid did not have a 300 IQ. So, that is pretty egregious. But! Doable if you’re not an idiot about it, I believe. But anybody who would do it would be a kind of idiot. First of all, I don’t know. How much would a 4-year-old be into it? But if you took a 6-year-old and got a 6-year-old into it, “We’re going to ride this pony into a T.V. show, your acting career.” It has never happened, but it is not impossible. Because Alicia Witt was a child actor, an actor now. Great actor and great kid actor, one of the things that makes for a great kid actor is a 4-year-old who can read.

Because if you can give a 4-year-old – Alicia Witt could read at 3 – a script and the kid can read the script and memorize the script rather than having to be told shit line by line, and if the kid is smart enough to do that, then the kid is smart enough to take direction. Alicia Witt was at least a kid actor because she was super fucking smart. So, I’m thinking if you had a motivated 6-year-old and a creepy parent. I even started working on a screenplay on this or thought about it 30 years ago as a good plot. Like a lot of shit I do, I didn’t do anything with it, except the mom did it and a shitty job in real life.

The right combination of psychopathic parent and bright, motivated kid. That team could believably sustain the bullshit that that kid has an IQ of 300+ for quite a while. Although, nobody has done that. Yes, that would be egregious.

Cole: Before they were banned by Wikipedia, there were many articles by groups making incredible IQ claims.

Jacobsen: What seem like the big lessons in debunking phony I.Q. claims from the 20th century?

May: “The first principle is that you must not fool yourself and you are the easiest person to fool.” — Richard P. Feynman

Rosner: [Laughing] A lot of stuff underlying a lot about high-IQ is “Why?” Why claim to have a high-IQ? Why work your ass off to get a super high score on these tests? Why sweat debunking it? In retrospect, you can see why you might want to hold people who might claim super-high-IQs up to scrutiny, at least given Raniere. The NXVIM sex cult, swindler of the Bronfman’s who is in prison for life now. One of the pillars of his duping people was using a high score on the Mega Test to claim to be one of the smartest people on Earth, though he didn’t really push it.

Because once he gathered enough acolytes, I don’t know enough about him to know how often he dragged out his IQ. But it seems that once he was surrounded by dozens of followers; that he didn’t need to do that. He could rely on his charisma and manipulation skills, and also being at the top of a pyramid of people with good manipulation skills. He was smart enough to recruit charismatic actors, TV stars. A couple actors from Smallville. People with actual show biz careers. One of his selling points and one of the selling points of Scientology can help you succeed professionally in shit where what it takes to succeed, like acting, can seem nebulous.

So, he didn’t need to haul out his IQ a lot because he was surrounded by TV stars who were helping him recruit other people into his cult. He, certainly, deserved a lot of scrutiny, perhaps a lot sooner than he got the scrutiny. There’s another guy who is pretty culty who has a bunch of acolytes who espoused a bunch of scary shit. So, that’s one reason to scrutinize claims of super-high-IQ because people can be up to no good, but those people are fairly rare. Of the 60, 80, 100, people who have qualified for the Mega Society over the past 40 years, 95 or more percent of them are completely normal, undangerous people.

The biggest danger might be that they might be really funny, like Richard May, is a completely decent guy who happens to be extra smart and extra funny. Super-high-IQ people mostly aren’t to be feared. What were we talking about? I always talk myself way away from the question. [Ed. Question repeated.]That, I guess, let the babies have their bottles for the most part, let high-IQ people be high-IQ people, it doesn’t hurt anyone, except for a few cases. Those involved in IQ fraud, the fraud is pretty transparent.

Most of the high-IQ lying is some desperate asshole who is 25 and going to undergraduate parties at his school. That guy finds a freshman girl and says, “Oh, people don’t understand me. I have a 205 IQ. I graduated high school at age 5.” It’s that abject bullshit. There are more sophisticated attempts, but not that much more. Because the payoffs are pretty low. Even lower than getting a hand job from a freshman girl, the end.

Cole: “It’s hard to be right.” — Richard Feynman

Jacobsen: What would you define as fraudulent activity in a high-IQ community or an individual?

Rick Rosner[1]*: Making claims that you know aren’t supported by your performance on tests.

Chris Cole[2]*: Fraud takes many forms just as it does in common law. Because of the Internet, tests with fixed questions are particularly vulnerable to cheating.

Richard May[3],[4]*: I have nothing to add.

Jacobsen: What would you define as messianic posing in a similar regard?

Rosner: If you end up with a cult, that’s messianic posing.

Cole: The common language definition of messianic behavior will serve. 

May: I have nothing to add.

Jacobsen: Similarly, what about criminal behaviour?

Rosner: If you end up in jail for the rest of your life, if the FBI has a thick dossier on you because you are considered a potential threat in certain ways, that’s criminal behaviour. The FBI has dossiers on lots of people because, historically, the FBI has done good things and asshole things.

So, if they have a dossier on you, because you’re a legitimate psycho who has the potential to do bodily harm to people for some weird political reason, then there you go.

Cole: Again I have nothing to add here to the common language definition of criminal behavior. 

May: I have nothing to add.

Jacobsen: On the Mega Test, why was the three interpenetrating cubes problem seen as the most difficult?

Rosner: It is widely agreed that the three interpenetrating cubes problem was the hardest problem on the test. So, the problem that is agreed upon as likely being the correct answer has not, as far as I know, been proven to be the correct answer.

Interestingly, you can look it up. It depends on what shit is online. But at various times since the ‘90s, it has been agreed upon that the correct answer is floating out there. But you can’t be sure that you’ve found the consensus correct answer.

But the figure, the geometric figure, that corresponds to the consensus correct answer can be found in popular culture, but I won’t tell you where.

Cole: It’s the only problem on the test where the answer that Ron accepts has not been proven. There are a few of these on the Titan.

May: It was the certainly most difficult, but my spatial ability is not sufficiently high to understand why this is so.

Jacobsen: Above 4 standard deviations above the norm, why should there be more scrutiny more than any other cutoff?

Rosner: Isn’t there some claim that “extraordinary claims require extraordinary evidence”? You could argue that because claiming to have one of the world’s highest IQs gets you more than claiming to have a 120 IQ.

In practical terms, not so often, it can get you on a quiz show. It can get you on the cover of Esquire magazine. It can get you interviewed. It can get you on TV. It kind of got me laid once. I was going to get laid anyway. But it was part of that package that got me laid, I guess.

Cole: A credible high range score requires credible high range test questions, which are hard to formulate and norm.

May: I have nothing to add.

Jacobsen: What was the hardest IQ test you’ve ever taken in the high-range? What lesson can be learned for test-makers from this?

Rosner: I say that I’ve had a lot of success, but I’d say that I’ve had the most difficulty with Cooijmans’ tests. Because he brings in stuff from a lot of areas. I don’t want to say too much about his tests because he doesn’t want people talking about his tests and helping other people.

But by the time the Mega Test had been published in Omni, it had been through a number of revisions with hinky problems getting knocked out or revised until they were clear and bullet-proof. The answers were tight. I think Cooijmans talks about the pleasure of when an answer clicks into place. That click of satisfaction of when you know you found the answer.

I would say that on some of Cooijmans’ problems. The click is, maybe, not as loud as on some Hoeflin problems. On Cooijmans’ problems, you can find some really good answers that aren’t as good as the intended answer. That’s, maybe, the mark of one type of really good ultra-high-IQ test.

That there are stopping points. On multiple choice tests, those are called distractors. There are answers among the choices that seem right for various reasons if you’re taking desperate stabs at an answer.

On high-IQ tests, you can come up with answers that make a lot of sense. But do they make as much sense as the intended answer? No. But you’ve fallen for an inferior answer. On tough tests, a lot of problems on hard tests are finding the signal among the noise.

I’m writing a book in which somebody or the recipient of what he thinks is a coded message, thinks that it is a true message because it is based on the first letters of four consecutive sentences. That spell out a word.

The odds that this would happen by chance are 26 to the 6th power, which is 676 squared, which is 400,000 to 1. Then you have to knock that down because there are a zillion four-letter words. So, anyway, the odds are tens of thousands to one that it’s not a coded message, especially since it is specific to the character situation.

So, the character reasons that it is likely a true signal. And on a tough IQ problem, you’d like the numerical coincidences to have an unlikelihood of, at least, 1 in a 1,000. When you look at a number sequence, you see a pattern. Then you say, “What are the odds that this pattern would arise by chance?”

On some super-hard IQ problems, there are more than one pattern to be found. Again, you have to ask yourself, “Was this intentional or accidental?” A tough-ass IQ problem really pushes the limit in finding the signal among the noise.

Cole: The only high range test I took was the Mega. 

May: The Mega Test and the L.A.I.T. are the only high range tests I’ve ever taken.
I did not distinguish myself on the latter.

Jacobsen: Is IQ declining in importance now?

Rosner: IQ as IQ is declining in importance because it is a product of the middle of the 20th century when people really believed in it and used it to skip kids a grade, or not, to put them in gifted classes, get admission to magnet schools.

At some point, probably in the ‘50s, you might be able to get laid by your IQ. Since debunked, it has a greasy feeling about it, weirdo, creepazoid. The Cal. State schools, today, decided to get rid of the ACT and SAT altogether and the SAT is an IQ surrogate.

They decided it is not helpful, not worth the shit people go through to prepare for the tests. We can see enough about a student without some IQ surrogate in their admission packet. I’d say intelligence is increasing in importance because we are tiptoeing up to artificial intelligence.

That when we talk about AI – and AI is a misnomer right now; AI means “machine learning.” Eventually, AI will mean “Artificial Intelligence.” We will need ways to mathematicize and to come up with metrics of the power of thought in brains and in other stuff.

So, old school IQ declining; new school AI shit increasing.

Cole: IQ seems to be about as important now as it was when I was young. The SAT has some problems because it has become easy to improve a score via tutoring, but that is being addressed.

May: There is a theoretical possibility that Nature, specifically natural selection might not be entirely “politically correct.” Theoretically there could be differences among human groups that evolved under different conditions. E.g., If only females could bear children, then males would be the expendable ‘gender’. A small number of healthy males could impregnate a large number of females and the group would survive. A large number of males, if males did not bear children, and a small number of females would not allow the group to survive. Hence, there could be more variability among males, including cognitive variability, because males would be more expendable, than among females, i.e., there would be more male ‘geniuses’ and more male idiots.

Fortunately we now realize that there are no biological differences between males and females. Gender is a purely social construct. We now realize that men can menstruate and have babies too, if given a chance. The only important differences are among large numbers of pronouns, all referring to identical nouns.


In-Sight Publishing by Scott Douglas Jacobsen is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. Based on a work at


© Scott Douglas Jacobsen and In-Sight Publishing 2012-Present. Unauthorized use and/or duplication of this material without express and written permission from this site’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Scott Douglas Jacobsen and In-Sight Publishing with appropriate and specific direction to the original content. All interviewees and authors co-copyright their material and may disseminate for their independent purposes.

Leave a Comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: