The Hungarian mathematical giant Paul Erdős would meet any reasonable definition of genius. He is so revered in his field that the “Erdős Number” refers to how many degrees of separation one is from having collaborated with the man. A number of 1 is assigned to those privileged enough to have co-authored a paper with him, a person who worked with that co-author would have a number of 2, and so on.

Besides incessant work habits which produced more than 1500 papers, Erdős was also known for his minimalist, transient lifestyle. He had almost no possessions, no significant interest beyond mathematics, and not even a home. Not that he ever wanted for a roof over his head. He traveled extensively to seminars, during which world-class mathematicians competed for the honor of having him stay with them so they could engage in problem-solving pursuits with Erdős.

During one such sojourn, a heavy thunderstorm sent rain shooting through an open window, which caused the somewhat-panicked Erdős to awaken the homeowner and express his alarm and confusion. This, as opposed to shutting the window. A man whose trophy case and walls of accomplishments would be absurdly expansive were he the type to have trophy cases and walls was unable to do what the average soaked dimwit would have done in the situation.

This amusing anecdote highlights one of the problems with Intelligence Quotient tests. They focus on specialties like problem-solving, reasoning, and planning. Erdős would have scored extremely high on such a test, perhaps achieving the most stratospheric number ever. But the test would fail to account for his ability to manage common-sense actions like weather-dependent room adjustments.

Similarly, an IQ test subject may have ingenuity but produce only mediocre grades in established academic classes. Another may struggle with slightly advanced mathematical principles but know how to recognize and exploit business opportunities. The idea that there is a single notion of intelligence, much less a way to adequately test everyone, in untenable.

In the early days of IQ tests, the quotient referenced the subject’s mental age, divided by the actual age. So a 10-year-old who reasoned at what the test considered average for a 15-year-old would score 150.

Later adaptations of the test graded on a curve so that the number represented a placement within the distribution of aggregated scores. So the “quotient” in IQ is no longer literal, although the term is still used.

But the tests fail to adjust for cultural differences and some critics argue that the testing more measures social class than intelligence. There is also the issue of those who don’t “test well,” while having a better ability to analyze and solve problems in real life.

Another drawback is the IQ tests revert perpetually to a normalized measure, with 100 being forever average and 68 percent of testers always scoring between 85 and 115. This keeps the focus on maintaining norms more than it does the stated goal of determining brainpower.

One needn’t be Paul Erdős to know all this doesn’t add up to a meaningful test.