I recently got the following email from commenter “Aint tellin”:

Hello,

I don’t want to bother you with such a mundane question, but this past week I’ve been seeing more and more of one particular comment in regards to comedian Norm Macdonald. I figured you would be the best person to validate this claim. He is frequently cited by many comedians as being a genius and having an incredibly high IQ. These same high-profile comedians are quick to label him the greatest living comic alive.

While I don’t know if these comments are correct, I do have to concede that from my listening of him out-of-character that he appears to be well read (particularly in Russian literature) very witty and articulate.

I know that you’ve previously estimated the IQ of the average comedian to be in the realms of +1.5SD above the norm. How high above the norm (no pun intended) would a comedian have to be to produce work worthy of the title: ‘funniest man alive’? Would you consider the claims of Norm Macdonald’s genius accurate or hyperbole?

 

I don’t know if there’s any consensus on who the funniest man alive is, but certainly Norm Macdonald is up there as one of the greats.  Comedians are fascinating because most are obviously very smart, but often act dumb for comic effect,  and unlike other high IQ occupations, they seldom have elite or advanced degrees, or even any degrees at all.

One sign that they have very high IQs is the extreme overrepresentation of the high IQ Ashkenazi population.   Depending on how you define them, Jews are 2 or 3% of America, and yet a third of Comedy Central’s “100 Greatest Standups of All Time” are Jewish.  Interestingly blacks are also overrepresented: they are 18% of the list despite being 13% of America.

Expected IQ of “funniest man alive”

We first need to know the correlation between IQ and comedy.   A 2011 study found a 0.27 correlation between Raven IQ and humor ability (as measured by the rated funniness of captions you can think up for cartoons).

But there are two reasons to think the correlation is an underestimate.  It was calculated on college students who vary less in IQ than the general population.  Correcting for range restriction would likely increase the correlation by about 0.1 or so.  Also, a very abbreviated version of the Raven was used and the lower reliability likely further depressed the correlation by perhaps  an additional 0.1.  Correcting for both factors, the correlation is likely 0.44.

However the Raven likely has a g loading around 0.72  (the g loading of the Matrix reasoning subtest on the WAIS-IV).  Some might say the Matrix reasoning subtest is less g loaded since it’s shorter than the Raven, but this is perhaps counterbalanced by the fact that it’s individually administered which ensures everyone understands the task.  Dividing 0.44 by 0.72 gives 0.61, which is the likely correlation between humor ability and general intelligence (g).

However great achievement requires more than just raw talent. It also helps to have 10,000 hours of practice, among other things. Raw talent seems to explain 66%  to 70% of the variance in various cognitive performance, suggesting talent correlates 0.82 with performance.

So multiplying the 0.61 correlation between g ad humor ability by the 0.82 correlation between ability and performance gives a 0.5 correlation between g and comic performance.

Now when people say “funniest man alive” they probably really mean  “funniest American of either sex” (Macdonald had to move to the U.S. though may not be a U.S. citizen).  Assuming there are 215 million Americans over age 25, a perfect correction between IQ and comedy would give the funniest person in America an IQ of 186 (U.S. norms).  In other words, 86 points above the U.S. mean.  But since the correlation is perhaps 0.5, we’d expect him to be 86(0.5) = 43 points above the U.S. mean, or IQ 143 (U.S. norms), with a 95% chance of being from 118 to 168.

How close did the estimate come?

Given the huge margin of error associated with this estimate, it’s always nice to have some empirical confirmation.  Usually no such confirmation exists, but I was extremely lucky to discover that back in 2000, Macdonald had appeared on the hit TV show Who wants to be a millionaire?  Since general knowledge is among the most g loaded measures of intelligence,  I decided this could serve as a rough proxy for his IQ, though he was asked only 15 questions (the WAIS information subtest has nearly twice that), and they were multiple choice which makes it easier to get lucky.

Although Macdonald relied on help from the audience and friends for some questions , and was talked out of giving a final answer to the last question, in my judgement he knew or would have correctly guessed the answer to all the questions except for the one about the guitar auction, giving him a score of 14 out of 15.

To see how this maps to IQ, I asked the questions  to my readers, who self-reported the following scores (out of 15):

poll2

 

Distribution of  the 48 self-reported scores as of sept 24, 4:24 pm Eastern:

15,15,14,14,14,13,13,13,13,13,13,13,13,12,12,12,12,12,12,12,12,11,11,11,11,11,10,10,10,10,10,10,10,10,9,9,9,9,8,8,8,8,7,6,5,5,4,2

The mean is 10.5 with a standard deviation (SD) of 2.8, so Macdonald’s IQ (as crudely measured by this quiz) is 1.25 SD above the average reader of pumpkinperson.com.

Since previous research suggests my readers average a borderline genius IQ of 127 (U.S. norms) with an SD of 15, that would put his IQ at:

1.25(15) + 127 = 146

Of course, modern IQ scales normalize the distribution, but that makes little difference for Macdonald’s score.  Either way he scored in the mid 140s, exactly as we’d statistically expect from his comic talent.

Despite his great performance on the game show, some might dismiss Macdonald as a mere clown and doubt he could be so brilliant.  And admittedly, general knowledge is only a rough proxy for IQ, and the game show was only a rough and perhaps rigged proxy for general knowledge (it’s not in the show’s interest to make us celebs looks dumb) and my equating to IQ relied on self-reported data from people on the internet.

However I saw an interview where Macdonald made a comment implying he’s smarter than he seems:  He said that unlike Bill Maher and other cerebral comics who want to show how smart they are, David Letterman is smarter than all those guys but smart enough to know that everyone hates a smart guy.  Perhaps Macdonald was projecting onto Letterman his own dumbing down strategy.