Re: Waves of Customers?

Rich Ulric wrote:

"I wonder how they did the training, and how much, for people
and (which) animals. "

------------

Well most of the animals seemed to be pigeons. But I think a rat also
featured.

------------------

Rich Ulric wrote:

" wonder how they did the training, and how much, for people
and (which) animals.

"I wonder how indistinguishable it is, to be "nearly
indistinguishable".


"A couple of years ago, I heard a National Public Radio
report on the outcome of a big "paper-scissors-stone"
competition. In Japan, I think, for thousands of dollars.
The eventual winner apparently shocked and psyched his
last opponent by playing "paper" as his final six moves.


"I don't know the tourney rules. "

-------------

I think there is a problem with knowing for sure whether any sequence
of numbers is random. If one tries one can always find a rule that will
fit.

I think diffeent experimenters used different measures. The later more
sophisticated experimenters combined three or four different tests of
randomness to come to their conclusion. For example, the pigeon
training offered rewards to sequences of pecks that had to meet several
criteria for randomness.

Oddly enough your example of rock paper scissors games was discussed in
the Annual Review paper.

---------------

Rich Ulric wrote:

"I think I imagined that mosquitoes, at slow speeds (like, while
near a target) suffered from worse aerodynamic instability than
helicopters do, and had far worse sensor systems.
- or maybe you are suggesting, That's what works for them.... "

---------------

1. The mosquito movement is my example, not the author's of the Annual
review paper.

2. I think there are insects as small and light as mosquitos that don't
move in such a random way - so I am not sure it is Brownian movement or
aerodynamic instability. Mosquitos evolved from species of insects that
were hunters like dragon flies (IIRC) - so it would be unlikely that
they have defective flight.

3. Nevertheless perhaps the randomness is indeed through the selection
of unstable filght mechanisms rather than random neuron firing. (The
original paper in the Annual review spends much time showing that
neurons can fire completely randomly.

Thanks for your reply.

Lance

.