It is known that humans are worse than computers at random number generation, and that even computer's random numbers are often imperfect and predictable.

What strategies could a human employ to generate truly random numbers from a given range, and how uniformly distributed will they really be?

P.S. One trick I use, when generating a number from 1-9, is to select a random large number, say 2349, then find its digital root (2+3+4+7 = 16, 1+6=7, so digital root is 7). I don't know how random it truly is, but it's definitely better than just instinctively selecting a number from 1 to 10.

  • $\begingroup$ @RobertColumbia I know human-generated numbers are not random and hence predictable. I'm asking if there is a way humans can avoid this weakness. $\endgroup$ Mar 5 '17 at 4:44
  • $\begingroup$ So your random number between 1 and 10 will never be 1? $\endgroup$
    – Shokhet
    Mar 5 '17 at 14:13
  • $\begingroup$ @Shokhet you're right. My bad. $\endgroup$ Mar 5 '17 at 14:27
  • $\begingroup$ I am pretty sure 42 and pi/2 are random numbers $\endgroup$
    – StrongBad
    Mar 6 '17 at 17:09

One strategy for ranges that are powers of 2 is to split the range in half and flip a coin. If it's heads, the upper half is your new range. If it's tails, the lower half is your new range. Repeat until you have a single number. This gives you a truly random number.

Another strategy comes from George Marsaglia, on this Google Groups page: https://groups.google.com/forum/m/?hl=en#!msg/sci.math/6BIYd0cafQo/Ucipn_5T_TMJ

Choose a 2-digit number, say 23, your "seed".

Form a new 2-digit number: the 10's digit plus 6 times the units digit.

The example sequence is 23 --> 20 --> 02 --> 12 --> 13 --> 19 --> 55 --> 35 --> ...

and its period is the order of the multiplier, 6, in the group of residues relatively prime to the modulus, 10. (59 in this case).

The "random digits" are the units digits of the 2-digit numbers, ie, 3,0,2,2,3,9,5,... the sequence mod 10. The arithmetic is simple enough to carry out in your head.

I don't know what the distribution is, but the numbers look random, at least to me. You could also concatenate digits to create larger numbers. You could then get a pseudo-random number for any specified length of number (i.e. 2 digits, 3 digits, etc).

A similar question with many potential answers in the comments can be found at https://philosophy.stackexchange.com/questions/1961/are-people-capable-of-generating-a-random-number.

  • $\begingroup$ That doesn't give uniform distribution if your range is not an exponent of 2. $\endgroup$ Mar 5 '17 at 13:33
  • $\begingroup$ This does not provide an answer to the question. Once you have sufficient reputation you will be able to comment on any post; instead, provide answers that don't require clarification from the asker. - From Review $\endgroup$ Mar 6 '17 at 7:24
  • $\begingroup$ @RobinKramer It does answer the question. I imagine there are better answers, but it does answer the question. The first part (the pseudo-comment) is completely separate from the answer. $\endgroup$ Mar 6 '17 at 7:26
  • $\begingroup$ I hope that one of the moderators can move your answer to the comments because the philosophical perspective is very interesting. The solution you gave is, in my opinion, not what is expected at CogSci. At CogSci, we seek to have answers that are grounded in psychological sciences with scientific references. Your answer seems to be more or less a life-hack. $\endgroup$ Mar 6 '17 at 7:31
  • $\begingroup$ The question was what strategies can a human use to find random numbers, correct? I'll edit my answer to make it more clear that the link to the philosophy question (which has many practical examples of how a human can generate random numbers) is not the point of my answer. $\endgroup$ Mar 6 '17 at 7:35

Not the answer you're looking for? Browse other questions tagged or ask your own question.