[From Dag Forssell (950614 1220)]
I follow, appreciate and enjoy the discussion of reinforcement theory. As
Rick expressed yesterday, I too recall that the probabilities of favorable /
unfavorable tumbles were discussed and settled late last year. I do not
quite recall how I visualized the spacial/probability discussion in my
kindergarten physics mind, but if as Rick now suggests the distribution is an
artifact of the geometry, how about this geometry:
Change the target to a line.
This will make it easier for E. Coli to reach in the physical world. It
should be esy to turn into a version of the manual game TUMBLE on the
demodisk. It brings symmetry to the question of "away from target" or
"towards target" with the in between, parallell to target, direction being in
the middle of a random distribution of directions, after a random tumble.
I trust this will be quite easy to model. Rick's awkward discontinuity of
random change of location is avoided.
Best, Dag