Math ability is one of the best proxies for IQ so I decided to create a little math test for anyone who’s interested. It’s completely anonymous. There are only 10 questions ranging from extremely easy to extremely difficult. The easiest question can probably be answered by over 99% of the general U.S. adult population while the hardest can probably be answered by less than 0.1%. To enhance the g loading, I picked questions that emphasized mathematical intuition and conceptual understanding as opposed to brute number crunching.
There’s no time limit, but try to take the test in one sitting without use of calculator, pen, paper or other aids, though I doubt they would help much. Take the anonymous poll to let us know how you scored.
Pumpkin Person 
Are you sure about the answer to ‘How many two-inch tall cubes are needed to make a six-inch tall cube?’ I expect it to be 3 times what you accept as an answer, but maybe I am misunderstanding something.
you were right. the error has been fixed. apologies to anyone who got it wrong before correction was made.
I replaced the poll since the first 3 respondents took an invalid test
where the fuck are my comments about oprah’s love of yuge penises?
MOMMA’S BRINGIN’ HOME A BIG ONE!
What’s the norm?
I scored 8/10. I don’t have a high math IQ so whatever that is – assuming 120s – it’s probably accurate. Though maybe if I was earnest on the ones I didn’t understand, I would get 9/10.
I got an 8 lol. The two last ones did not make any sense to me but I guessed anyways. Where does this put me compared to the general pop Pumpkin?
i’m sure it puts you well above average but i don’t have any norms yet since you guys are the first to take this test
i got 10 out of 10 10 times in a row.
you rigged the poll, i knew it was impossible for that many people to have scored perfect
impressive!
Your last question assumes all attributes are independent which is not likely
it’s not particularly unlikely.
I just assumed the covariance matrix = identity matrix so we got the same answer.
Got a 9, so I must’ve have gotten at least one of the last two correct, thanks to reading this blog and learning some math about correlations.
If my theory is correct and Mugabe took the poll 10 times, they are, at least, 2 obvious liars in the 17 (26 – 9) persons who took the poll. It would be nice if I am right not to post my previous comment to see who else is bragging about a perfect score. Or even ask who scored a ten …
If I am wrong, obviously this comment is useless 😂
How do you do the last two? I couldn’t get those.
9. I misread the garden hose question like a retard but I know the answers to all 10 now.
Are you still accepting submissions for your heritability test?
If so what’s the best way to prepare?
I got 9 out of 10 on this, and like other people said I didn’t exactly understand the last two questions, but reading your blog a lot I learned a bit about this kind of math and figured what you were going for. I’m guessing my answer for the last one was too high.
I’m American, I never took the SAT because I dropped out of High School. However I took the GED and got a perfect score on the math portion, though I think that test has a pretty low ceiling in terms of the most difficult questions on there. Also, math was my best subject during my primary school years, but I’ve not studied it in any detail for maybe 7 years now. Highest class I took was Calculus AFAIK.
I’ve never taken an official IQ test, but I personally reckon my IQ is between 1-2SDs above average. My mathematical IQ is probably close to 130 while my verbal IQ is probably closer to 100-110. Could be pretty off on this dichomoty though… that’s just my idea based on experience in primary school.
I got 8 out of 10 but I wish I knew which ones I got wrong because I’d like to work on them.
I got 10 with same answers. I probably did it when there were the incorrect answer. As I thought all my answers were correct, I believed you had score it out of 9 just to check out false reporting 😊
I don’t think the math ceiling is higher than 135 (99%).
well if we ignore the ten fake tens by Mug of Pee, it seems about 10% of people here got 10. In the general population that’s IQ 120 but assuming this group is 25 points above average with normal variability, it might be IQ 145.
also seems bottom 4% score 5 or less. In the general population that would be IQ 74, but if we add 25 points, that would be IQ 99.
so 5 = IQ 99
10 = IQ 145
If we assume linear
6 = IQ 108
7 = IQ 117
8 = IQ 127
9 = IQ 136
just a preliminary guess
It’s difficult to know for me because I am quite trained in mathematics. I was proposed for Math Olympiad at 16 yo but didn’t do it because I used to go on holiday to Spain 3 months every summer and my parents weren’t interested by school grade or performance …
I remember LoB spoke about a test with 3 questions given to Harvard students who got only 25% 3 out of 3 when I seemed to me being of 125. So I may be 10 points short.
For whatever reasons, I have crystallized more intelligence on math despite being weak in spatial.
I got a 9. If 9 = 136 (converting to about 99th percentile assuming SD = 15) then that would be consistent with my SAT math score (770; 98 percentile).
Pumpkin this seems to relate to working memory (my intuition). 127 is essentially my working memory and math IQ probably relies a lot on it.
If this is 145, Taleb regularly asks question wich would be 20 points higher, thius 165. I have answered correctly only once because I hate the guy who is a fake mathematician.
IMO would be then 40 points higher at least when I thought there would be 172. But it would make sense because there is only 2 or 3 perfect scorers in the all world. I just thought the competition was not known enough to reach the super smarts kids lost in an Alpes village en somewhere in central China because it’s math teacher wouldn’t be neither good enough to notice nor knowledgeable enough to know the IMO (nor caring enough about pupils).
For me, my math teacher hated me but the dean of my high school had a PhD in math and proposed me to the French selections in Paris …
The difference between 1 in 1,2M and 1 in 40M, is that out of 120M born, you need to reach 4% in the first case (172) but 80% in the second case (182) …
it’s interesting how the french system selects for IQ so strongly yet the french elite are still so dumb.
bruno denies their system selects well for IQ
French system selects on knowledge based on a program you acquire in preparatory schools during 2 to 3 years after higher school. The higher IQ, the less you will have to work but due to the routinized testing, wich unlike MCQ doesn’t have an essential time constraint and basic tasks, you get something more like chess where rote memory is the most important factor.
Then elites are selected by one school ENA wich is based on law and public policies knowledge. So there, even if the selection is the highest in terms of attractivity, correlation with IQ is even lower.
They are two exceptions : students who enter Ecole Normale supérieure d’Ulm in pure math (40 people) tend do be very bright. This school compare to no other in termes of bachelor getting the field medal.
And students from engineering school Ecole Polytechnique who enter the corps des Mines (15 people out of the 400 students) tend to be very bright. So for those 55 people, probably their IQ is above 135. But that’s a very tiny group ans the power is in the end of another group of 80 people per year out of ENA.
Then the larger french elite 4000 people out of French management program and engineering schools have probably an average IQ around 125 and are very industrious. There is no positive discrimination for them. And they are in all corporate world.
3 out of top 5 management programs (business schools without professional experience) are French:
http://rankings.ft.com/businessschoolrankings/masters-in-management-2019
And our top engineering schools are more selective that our business programs, so they don’t have much competition in the world. And it’s because of the selection process : you select students based on high school grades, they prepare a vast program during 2 or 3 Years, and the Grandes Écoles select the best scorers they can (some have joint exams other have separate one, but they are all based on the same national program).
Out of 800k people each year , 10k prepare management schools and 15k prepare Engineering schools. The three best French schools select 1200 students. They also select 600 students with a bachelor (that’s were you get some legacies) and 600 from foreign countries (same). It’s the same selection process for engineering schools.
So in general, you have to sacrifice your 18-20 in France if you want to join this club (2400 people) and your 20-28 if you want to join the ENA and Corps des mines tiny group (80+15).
There are plenty of other small prestigious programs only known by French elites but that will give you prestige for life here …
Bruno what do you do for a living anyway?
I love Taleb’s 110 IQ reply guys who pretend to understand anything he says.
I follow him just to laugh at them.
taleb isn’t the genius he portrays himself to be.
his whole Incerto series can be summarized in one sentence. much like immanuel kant’s CPR can be summarized in one sentence.
taleb is interesting because he combines very high VIQ and SIQ with high numeracy.
by SIQ I assume you mean social, not spatial?
yeah.
the whole taleb shtick is just anti-autism:
1. some people habitually treat models of reality as reality, are habitually incognizant of “model risk” or that no model can ever be useful for long.
2. “gut instincts” are TOO often more accurate than models.
HERE MUGABE SPEAKS WITH AUTHORITAH…
HEARSAY AUTHORITAH…
…
i worked with a guy who’d given up being a futures floor trader to do actuarial shit…
i pecked at him.
“why did you leave?”
“we’d have to go drinking. (dude was an alky of the first kind)…
because the risks i had to take to make bank were too great. (it rhymes!)”
i said, “but you have the math and stats better than most such people.”
dude says, “the guys making bank in the pits have third grade educations.”
or so i was told.
For whatever reason, autism is virtually non-existent in the Levant. Ultra-orthodox jews are probably the only exception.
there’s actually a marxist reason for lack of freedom of speech in germany and austria. but the other countries have no excuse.
that is, german capitalists fear economic sanctions if neo-nazis gain some representation in legislatures.
Just like certain capitalists like Robert Mercer backed trump.
my IQ is too low to see the connection.
And the last two questions are for people who read your blog and can recognize the sd for height and weight, in order to guess the simple correlation. Same for skills & height in NBA.
they don’t have to know the SDs but yes, being familiar with the concept of regress is a hue advantage on the last 2 so they’re perhaps not good measures of IQ. still i’m shocked anyone got #10 as the reasoning is very esoteric. I wonder if they got it by accident
you might have to redact this but I just assumed that [redacted by pp, aug 4, 2020].
very impressive. exactly the reasoning i was looking for. I had assumed that one would have stumped everyone except bruno & marsha.
though the last 2 might be good measures of IQ for the vast majority of people who haven’t read about regression and just use their common sense to figure it out.
I’m not sure its easy to test mathematical reasoning – study has way to large an effect. In my experience that average IMO medalist is prob around 3 sigma, great, but nothing too exceptional. Practice is the key.
Solution for 10 that doesn’t require you to know SD. It gave the correct answer, though I’m not wholly confident in the method’s correctness.
You have to assume that the components account for 100% of the variance, which you haven’t specified but I don’t think it’s solvable otherwise, then since Var(Basketball) = Var(Height) + Var(Practice) + Var(Speed) + Var(Coordination) = 4Var(Height), the standard deviation is twice as big for basketball as it is for height (since variance is standard deviation squared) and the slope of the regression is 0.25 (since Var(Height)/Var(Basketball) = 0.25, i.e. variation in height accounts for 25% of the variation in Basketball playing. It essentially works the same way as heritability where you treat Height as G and the rest as E.). You can use that to find the correlation coefficient and solve it that way, but I hadn’t thought of doing that, so instead I treated the 12 inch gap as a gap of 1 arbitrary unit (some multiple of the standard deviation) which means there is a 2 arbitrary unit gap in-between the best and worst basketball player. Since slope is 0.25, for two arbitrary units of basketball there are 0.5 arbitrary units of height, i.e. 6 inches.
Again, you have to assume that the components account for 100% of the variance, but I think this is true for all solutions. Since you haven’t mentioned this, I don’t think the question is strictly solvable.
I treated the 12 inch gap as a gap of 1 arbitrary unit (some multiple of the standard deviation) which means there is a 2 arbitrary unit gap in-between the best and worst basketball player.
I’m not following the logic.
I will admit this might not have been the clearest solution, it was however what I had thought of while working on the problem.
Let’s say the gap between heights is of X standard deviations. Let’s say we set the SD for height to 1 (which is probably what I should have done), then the SD for basketball is 2. At this point, the more obvious solution would have been to say that there was a gap of X in height and therefore a gap of 2X in basketball, and proceed from there. Instead I invented a unit such that a gap of X SD in height corresponded to one unit. In this situation, one unit would correspond to X, but let’s say the SD was 2, then one unit would be 2X. It was just a way to solve the problem without setting the SD, but there’s not really a reason to do that.
rr reminds me of ben affleck in that exchange he had with sam harris. sam harris is gross but being anti-islam doesn’t make you a racist.
for one thing there are white muslims in the balkans and the caucasus.
but the kill shot is:
1. religion is NOT genetic.
2. UN-like ALL other religions, islam is a VERY explicitly POLITICAL religion.
Despite that decades-long legacy, the president insisted that, perhaps excluding Abraham Lincoln, “nobody has done more for Black Americans” than he had, including Lyndon Johnson, who passed the Civil Rights Act in 1964.
“How’s it worked out?” Trump retorted. “If you take a look at what Lyndon Johnson did, you think?”
It was unclear what the president meant.
Hahahaha. I hope trump gets his own radio or tv show after he loses in November.
I sometimes wonder blacks should be allowed vote personally.
of course one might claim that judaism is also explicitly political.
but in two senses.
1. ancient judaism DID prescribe a form of government. but israel doesn’t follow it.
2. modern judaism may prescribe a form of government too, but only in the sense that “whatever the form of government, jews control it”.
and islam is really just universal, militant, ersatz judaism.
Eh, as far as I know the ancient caliphs and their courts were essentially licentious and filled with topless dancing girls.
was thinking how the counter-cultural people today are alt-right.
that is, in 2020 being “alt-right” is like being a “hippie” in 1968, but less gay.
but in my case it’s more than just contrarianism.
i support the nation and the nation-state because i’m a small “d” democrat.
i’m an FDR Democrat.
^^^SUCH COMMENT WOULD BE REMOVED BY UNZ^^^
[redacted by pp, aug 4, 2020]
10/10
I did the first 8 questions easily,but I think I only got the last two questions right for reading your blog with some regularity.
The last two don’t seem very culture fair.
“the rise of the nation-state” is an historical phenomenon those who jive for american empire pretend doesn’t exist.
those who espouse american empire are ANTI-american.
Mug completely destroys the aesthetics of this blog. Especially with his beige gravitar…probably a pretty good representation of his whiteness and his overall blandness in terms of persona.
Is he called Mug of Pee because he likes calling pumpkin peepee?
Yes, I thought this obvious to anyone. He used to go by ‘Robert Gabriel Mugabe,’ so Mug is an homage to his legacy using that pseudonym while converging on the fact that indeed he calls Pumpkin “peepee.”
Mugabe
MUG A BEE
MUG OF PEE
I thought the horse one was a trick question. (6)
7/10, which seems to be consistent with my high average iq. However when I first took the test I misunderstood question 7 and scored 6, I don’t know if I’d be penalized for that.
Have you tested your iq?
https://www.dailymail.co.uk/femail/article-8593165/Do-traits-genius-Tricky-15-question-quiz-reveals-IQ-higher-151.html
Scores of 9 equate to a 151. I got all 15. Super easy test for anyone on this blog tbh. May have to consider that it was in the “Denali” section of the DM lol.
Haha i have no idea what a “denali” is probably some Indian bullshit but I meant to type “femail” hahahah.
I ended up here by chance via Cochran’s blog, saw the test, though I’ll bite, and took it.
First try was 5/10; turns out I hadn’t read the questions carefully lol. 2nd try got 10/10
I find it hard to explain how I ended up with the correct answers for questions 9 and 10 (English isn’t my first language and I’m not good with math terms in English) but I supposed they’re related and the same logic applies to both. Does it?
For the record, since you’ve written about IQ tests not yielding consistent results, I’ve taken various Mensa-grade IQ tests multiple times over the years in different states (fresh, tired, sleep-deprived, hungover) and the result is always somewhere around 130. I can never answer the final questions in the allocated time so I know there’s a hard limit on what I can grasp of this world. Meeting more intelligent persons is both fascinating and depressing!
8/10, Smoothness in increasing difficulty from item to item is always a problem for tests, so nobody should go nuts . I mean, as many might notice, there must be, say, one more item between n8 and n9 . Was fun anyways👍
wow. interestingly, since in n5 I have mistakenly read it as “5 times more” instead just “five more” I got it wrong. It means that my guessing on n9 or n10 was right 🤦♂️
instead of*
it was the 9
Got a 7/10, what’s the norm?
10/10
First 8 are very easy with no time limit. I don’t know if I would have intuited the final 2 without learning about regression from this blog, though.
Some math stats:
34 on ACT math section
129 on WRAT-4 math section
19(perfect score) on WAIS-IV arithmetic
A 19 on WAIS-IV Arithmetic equates to IQ 145 (U.S. norms), very similar to your 10 on PATMA which I’ve tentatively assigned IQ 147 (white norms)
I scored a 7
Some other relevant stats:
IQ: 139 (SB)
SAT math score: 550 out of 800
Perhaps it is just sour grapes on my behalf but I tend to agree with Marc Lamont Hill that the SAT is a measurement of how middle class you are rather than a measurement of intelligence. I come from an extremely low income family and attended shitty public schools my entire life. I don’t blame my mathematical mediocrity on educational quality, however, so much as the attitude of the underclass towards education. my cavalier attitude towards education likely reflected that of the single !other who raised me, who was an 8th grade drop out. Personally, I treated the SAT like an eye exam; something you just walk in off the street and take to get your guidance counselor off your back. I attribute my stellar 770 verbal score to a love of reading more than to family background or formal education.
Just took the test a second time and scored an 8. I completely misread the question about garden hoses the first time through. Apparently my reading comprehension isn’t as strong as I tell myself it is. Ha.
I might have about the same IQ score and scored an 8 as well. Also scored similarly on the SAT Math, and my reading score was also higher, though not as high as yours.
Is your SB CPI significantly lower than your GAI? That might have something to do with poor math scores and reading comprehension.
Pumpkin, practice effects from ravens only minimally affect wais matrix reasoning right, did to different types of items?
Pumpkin, since 40% of Hispanics are monolingual, if the VIQ PIQ gap is 2 points, if the sample size was 100% bilingual, it would be more than 2 points right?
Pumpkin, I got a 7 on this test. I finished in 8 minutes. I’m too lazy to read anything now but tell me what all this means
THANKS
Interesting. So based on the test’s you’ve taken here:
Gestalt IQ 137
PATMA IQ 116
COMPOSITE IQ 132
which makes sense since it’s in between your WAIS-IV (IQ 120) and your SAT (IQ 138)
GondwanaMan had also a 173 LSAT wich must be higher no ?
On his second attempt iirc
Yo 132 Composite IQ??? I’ll take it!!
Pumpkin, I scored 29 on ACT Math. I hadn’t studied for the test and was mostly or completely unfamiliar with some of the types of math it covered, particularly plane trigonometry. My composite was 33, with 36, 35, and 33 on English, Reading, and Science respectively. How high do you think I could have scored on ACT Math with the proper preparation to reach my maximum score?
For reference, I got 16 scaled on the WAIS-IV arithmetic. I think I underperformed on that due to sleepiness, but it is what it is.
i.e., at least one of the arithmetic questions I missed was because I zoned out while the question was being read and the administrator wouldn’t repeat it
If you got a 29 with no prep and very little familiarity, you’d easily get a 36 with a little prep. Your 29 is already ridiculously high, especially with little familiarity with the content.
Pumpkin, my highest SAT Math was 780, even if everyone took it multiple times and studied hard, would it still be the case that the 780 is at least a 115 math iq?
Pumpkin, can a good matrix reasoning score make up for a lower figure weights score, hence still retaining the Math ability.
The link is doesn’t work. Can you please make the test accessible again?
it should be working again
Thanks!
8/10. 11/20 on the number sequence test but I had fallen asleep after sloppily answering some of them and forgotten which they were when I woke up, also skipped one of the first 10 (which generally were easy?). So I did rush.
With regards to this test, for the 9th question I had both the correct answer and an incorrect one in mind, I submitted the incorrect one. I guessed 12 inches (same as tallest person) though I knew that answering like this would not minimize the error across these kinds of questions but that’s entering statistical territory that I’m not familiar with (I never went to high school). Take a bipedal robot factory, are you seriously going to guess the heaviest robot is only 6 inches taller than the average one and not just a beefed up version of the tallest one? So to answer this question I’d have to get into the mindset of statistics and genetics. My other guess was 6 inches but that was based on a visualization of 2 containers holding 12 inches and 50 pounds maximum, and as 25 pounds fill to 50 pounds, 12 inches empties to 6 inches. No idea how nonsensical or not that idea was.
Link is down! Same for the TAVIS 😦
Any way I can get access to either? Thanks!
I took the KAMIKAZE and got 11/17, interested to compare.