Re: [MLUG - DISCUSSION] Interesting mind experiment

[Jonathan King <EMAIL:PROTECTED>]

On Thu, 6 Mar 2003, Mike Miller wrote:

> On Wed, 5 Mar 2003, Mikhail Kovalenko wrote:
> >
> > This is priceless! Especially the "explanations" page:
> >
> > http://sprott.physics.wisc.edu/pickover/esp4.html
> >
> > "Quantum mechanics" and "synchronicity"! Yeah, right  ;)
> 
> My dad showed me this page, or another like it, several years ago.  I
> figured out the trick by using "back" on my browser the first time I did
> it.  It's just based on the idea that they've used such similar cards in
> both sets that no one will notice the complete absence of overlap.
> People who can't figure it out, especially after serious consideration,
> really aren't that smart.

I think that's a bit harsh, since there are a lot of interesting tricks 
out there that some really smart people have cratered on.  In particular,
the "Monty Hall Problem".  Erdos himself got bamboozled by that one.

The thing about this trick that's interesting is that it does an
outstanding job of playing on three specific weaknesses in people's
cognitive abilities.  

The first weakness is that people have a fairly
limited working memory capacity.  Each of the face cards they use is a
pretty complicated object that is, as Mike pointed out, very confusable
with any of the others used or not used.  I haven't done the study, but
I'm willing to believe that your memory span for them is sharply lower
than for cards in general.  This means that people won't be motivated 
(or incidentally very likely) to encode more than the one card they 
selected, or to retain the identity of more than one card after the 
distractor task (selecting an eyeball).  

The second weakness this exploits is that people are truly terrible in
detecting changes made to the part of a visual stimulus that they aren't
paying attention to (so-called change blindness).  In other words, you've
really only thoroughly encoded and paid attention to the one card, but
maybe you would recognize changes in the other cards.  Not likely, 
especially when they are this subtle.

The third weakness, and the star of our show, is the fact that people
trust background stories and moreover do not systematically attempt to
disconfirm their own hypotheses or the statements of others.  Everybody
tacitly assumes that the way this trick has to work is that one card is
taken away from the ones displayed.  Nothing else would "make sense".  
Everybody also assumes that this is somehow a test of ESP, even though
they "know" it is a computer on the other side.  So they end up testing
the rule, "if the guy has ESP, the card I picked will be gone".

Now, that rule will be violated only when the card you pick stays on
the "table".  This, we know, will never happen.  Somebody who was now
trying to disconfirm the ESP point will now have to generate a more 
sophisticated hypothesis, which is just what people tend not to do.
Most people in this case will take the easy way out of just trying it
again and again, and so build up what looks like a very strong *inductive*
case for ESP when they might instead have tried to test another rule, 
maybe "if the computer takes one card away, the others will stay the 
same".  Testing that rule would be much better in this case.

So Mike, I guess I'm saying that this is likely to trip up a lot more 
people than you might think.  It is *really* well designed, despite the 
cheesiness of the graphics and such.

jking


_______________________________________________
discussion mailing list
EMAIL:PROTECTED
http://mlug.missouri.edu/mailman/listinfo/discussion