[EM] Practical democracy as a serialized cascade tree

[Michael Allan]

Here's the Practical Democracy method (Fred Gohlke, in thread
"language/framing quibble").  I'll show just three process steps (or
levels), culminating in the final selection (AA) at step 3.

   Practical:
   ---------
 1
   (A) (B) (C) (D) (E) (F) (G) (H) (I) (J) (K) (L) (M) (N) (O) (P) (Q) (R)
    |  /    |  /    |  /    |  /    |  /    |  /    |  /    |  /    |  /
    | /     | /     | /     | /     | /     | /     | /     | /     | /
    |/      |/      |/      |/      |/      |/      |/      |/      |/
   (S)     (T)     (Y)     (U)     (V)     (Z)     (W)     (X)     (AA)

 2
   (S)     (T)             (U)     (V)             (W)     (X)
    |      /                |      /                |      /
    |  ---                  |  ---                  |  ---
    |/                      |/                      |/
   (X)                     (Y)                     (AA)

 3
   (X)                     (Y)
    |                      /
    |  -------------------
    |/
   (AA)


This looks like a delegate cascade (or delegable proxy), in which the
structure was serialized.  If we un-serialize it by gluing all three
steps together, we obtain the equivalent decision as expressed by a
cascade:


   Cascade:
   -------

   (A) (B) (C) (D) (E) (F) (G) (H) (I) (J) (K) (L) (M) (N) (O) (P) (Q) (R)
    |  /    |  /    |  /    |  /    |  /    |  /    |  /    |  /    |  /
    | /     | /     | /     | /     | /     | /     | /     | /     | /
    |/      |/      |/      |/      |/      |/      |/      |/      |/
   (S)     (T)      /      (U)     (V)      /      (W)     (X)      /
    |      /       /        |      /       /        |      /       /
    |  -----------          |  -----------          |  -----------
    |/                      |/                      |/
   (Y)                     (Z)                      /
    |                      /                       /
    |  -------------------------------------------
    |/
   (AA)

The difference with the cascade is that voters (all nodes) are free to
shift their votes (edges) at any time, in response to the current
state of the election.  Also, they can vote for anyone at all (A to Z
and AA) - they are not constrained to triads (binary order of tree).

Suppose we ran the practical method first, selecting AA.  Then we
translated the same votes over to the cascade method (as shown above),
then let it run from there.  Would the selection be *expected* to
hold?  Would a rational C, for example, be expected to keep voting for
T (or maybe directly for Y or AA)?  And so on, for all the other
voters and delegates?

Suppose not.  Suppose they deliberate for another 3 months.  At the
end of that time, instead of 100% of the votes cascading to AA, we
have 66% of them cascading to X (and 33% for others).  Something like
this:

   (A) (B) (C) (D) (E) (F) (G) (U) (I) (Z) (K) (L) (M) (N) (O) (P) (Q) (R)
    |  /    |  /    |  /    |  /    |  /    |  /    |  /    |  /    |  /
    | /     | /     | /     | /     | /     | /     | /     | /     | /
    |/      |/      |/      |/      |/      |/      |/      |/      |/
   (S)     (Y)      /      (H)     (V)      /       |      (W)      /
    |      /       /        |      /       /        |              /
    |  -----------          |  -----------          |  -----------
    |/                      |/                      |/
   (T)                     (J)                     (AA)
    |                      /
    |  -------------------
    |/
   (X)

Who then would be the better choice, AA or X?

(I say X.  Most of those under the power of the office have assented
to placing X there, by their votes.  They had not assented to placing
AA.)

-- 
Michael Allan

Toronto, 647-436-4521
http://zelea.com/