Re: non-random free will

jillery wrote:

The following is a link to a Veritasium video.  It claims that when a
statistically large number of people are asked to identify a random
integer between 1 and 100, their choices are anything but random.
Instead, the most popular choice is 69 followed by 37.  Not only that,
but multiples of 2, 5 and 10 were chosen much less often, while prime
numbers were chosen much more often than random chance allows.

<https://www.youtube.com/watch?v=d6iQrh2TK98>

The video goes on to provide a running histogram of 200,000 choices
made in a Reddit survey.  It shows how the distribution remained
remarkably consistent over time. 

These patterns suggest there is something that makes people choose,
and not choose, numbers not just non-randomly, but selectively,
without being aware of it. 

I have to strongly disagree with that analysis.
To illustrate, consider 3 distinct prompts:
1 Pick a random integer between 1 and 100
2 Pick your favorite integer between 1 and 100
3 Pick any integer between 1 and 100
Would you expect different results? How would you be inclined to
interpret the differences between the ensemble of answers?
Now let's add a variant. Before question 1 a person is asked, ask respondent to read a short discussion of a definition of randomness.
 This is offered to prod thinking about if the experimental design
was well constructed to actually address the sort of answer being
supplied as an interpretation. I have to add that it's a very difficult thing to design a study
around.
Part of that difficulty is the fundamental problem with even defining what random actually means. Seemingly, the analysis used is presuming
that "random" picking would generate a histogram with equal frequency
for each integer between 1 and 100 (and does that "between" mean 1 and
100 are legal choices?) A second problem, as indirectly alluded to above, is whether or not people are interpreting the instructions the same. These things are
very
poorly controlled. That makes interpreting the results rather dubious. I do think the whole thing provides a good lesson in the importance
of proper experimental design respecting to the questions one is hoping to address with the data.

The video suggests that most people "feel" prime numbers are more
random than non-prime numbers, perhaps because there is no exact
formula for calculating the nth prime number, and also perhaps because
the real world provides few representations of prime values greater
than 7.

So why should 69 and 37 be selected most often?  69 has some cultural
associations.  As for 37, the following offers a number of
suggestions:

<www.thirty-seven.org>

For those who might ask what this post has to do with evolution vs
creationism, I would say it has about as much to do with it as do
posts about free will.

--
To know less than we don't know is the nature of most knowledge