[Phys-L] how research is done : exploring a maze using only local information

[John Denker <jsd@av8n.com>]

In life, it is very often necessary to give up on an unimportant
sub-goal in order to make progress on the important overall goal.

In the game of poker, if you try to win every hand, you will lose the
game, and lose miserably. I am reminded of the wonderful song written
by Don Schiltz and made famous by Kenny Rogers:
    You’ve got to know when to hold ’em,
    know when to fold ’em,
    Know when to walk away,
    know when to run.

When exploring a maze, no matter how smart you are, you cannot run
directly to the goal. You have to spend a goodly amount of time
exploring branches that turn out to be dead ends. The fact that
such-and-such avenue is a dead end is valuable information, and the
only way to earn that information is to do some exploring.

Non-scientists tend to have a grossly oversimplified view of how
science is done. In particular, please do not confuse the way science
is /published/ with the way science is actually /done/. The textbooks and
the scientific papers generally use 20/20 hindsight to give you an
overview of the already-explored maze, including the shortest path to
the goal. For various reasons (good and bad), they don’t tell you
about all the blind alleys the scientists had to explore before
finding the optimal solution.

For a discussion of these points, see
  https://www.av8n.com/physics/research-maze.htm
or equivalently (with less security)
  http://www.av8n.com/physics/research-maze.htm

In particular ... I cobbled up an interactive computer program
to demonstrate exploring a maze using only /local/ information.
This may be the nerdiest computer game ever, but it's a nice
metaphor for how research is done.

You get a bird's eye view of where you've been, but you have to /earn/
the information by exploring.  This is very different from the usual
printed-on-paper maze, where you get a bird's-eye view of the whole
layout before you begin.

The goal is to find the cheese.  Note that you don't even know where
the cheese is until you've found it.

To explore, move the mouse over the maze, starting from some place
that you've already seen.  Clicking the mouse is not necessary.
> From any place that you've visited, you can see one step in any
direction, unless a wall intervenes.

Special feature: The program keeps track of which spaces you have
visited, and the order in which you visited them.

The code is at:
  https://www.av8n.com/physics/maze-local-info.py

It's an interactive visual-python program.  For now, you have to
download it and run it from the command line.  It should run on a
reasonably wide range of systems: Linux/Mac/Windows.

  If you believe what it says at glowscript.org, it should be
  possible to get it to run as a web-browser app, but that's
  more work than I feel like doing at the moment.