IQ 170

170 is 4.67 standard deviations above the mean. On a bell curve that works out to about 1 in 653,000 people, which would mean about 100 in the UK and about 12,000 worldwide. The bell curve is a model, though, and at this height it is doing most of the work.

No test has enough people in its norm sample to check its accuracy at 170. In a 2,200-person sample the bell curve expects effectively nobody at 170 or higher.

Some tests publish extended norms above 160 for exceptional children. These rest on very small samples and on modelling, so they are best treated as rough rankings.

A 170 almost never comes from a modern, supervised adult test. It usually comes from somewhere like this:

• Ratio IQ. Early childhood tests divided mental age by real age. A 6-year-old solving puzzles like a typical 10.2-year-old got a ratio IQ of 170. Modern deviation IQs do not work this way.
• Unnormed online quizzes. A test with no norm data can print any number. Very high outputs usually say more about the scoring than the person.
• Repetition without a source. Numbers attached to celebrities are often copied from site to site with no test report behind them.

Isaac Newton and Leonardo da Vinci are usually given estimates of 160 or more, and they are among 42 such profiles here. Only 14 go as high as 170. All are judgements from biographies, not test results.

Estimates this high mostly reflect how extraordinary the achievement was. That is a different thing from how someone would perform on a timed reasoning test.

On a bell curve with a mean of 100 and a standard deviation of 15, about 0.00015% of people score 170 or higher. Applied to whole populations:

Going from 165 to 170 makes a score 4.8 times rarer, and going on to 175 multiplies the rarity by another 5.3. Near 170, each single point adds about 39% to the rarity.

What your 170 could really be. Even a good supervised test has a margin of error of roughly 5 points either side, and far more at this height. A true score from 165 to 175 would span rarities from 1 in 136,000 to 1 in 3.5 million.

Which scale was it on? Most modern tests use a standard deviation of 15, which is what this page assumes. If your 170 came from a test using an SD of 16, it is about 166 on the SD 15 scale. From an SD 24 scale, used by some Cattell-style reports, it is about 144.

Estimates this high are reserved for a small group of scientists, artists and thinkers. These sit near 170. Treat these as informed guesses from biographies rather than results. Each profile sets out what its estimate rests on.

How 170 stacks up against the scores just around it and the round numbers people usually ask about. Each score links to its own page.