[EM] (Fwd) Problem with Condorcet?

[DEMOREP1 at aol.com]

Using the Condorcet circular tie instant runoffs method of Mar 31, 1996 the 3
Apples-Chocolate runoffs are:

Chocolate loses (by not beating anyone head to head)
Second choice of Chocolate voters transferred
A1  20 + 15 = 35
A2  20 + 15 = 35
A3  15 + 15 = 30
Condorcet math redone (same as in original due to having only 3 runoff
candidates)
A1, A2    65, 35
A1, A3    35, 65
A2, A3    70, 30
Circular tie A1>A2>A3>A1 
A3 loses (having only 30 first and second choice votes)
Transfer the A3 votes
A1   35 + 15 (A3 1st, A2 2nd) + 15 (Chocolate 1st, A3 2nd , A1 3rd) = 65
A2   35
A1 beats A2 65 to 35.

Chocolate wins with the plain Condorcet standard of worse- pairing- defeat-
is- smallest due to a division of the 55 Apple (A1+A2+A3) majority first
choices among the 3 Apple candidates who all ranked Chocolate last.

The Apples/Chocolate example shows why there must be a disapproval vote
combined with Condorcet to have defeats for candidates not wanted by a
majority (which may well cause many more new elections with a different group
of candidates).

After the majority disapproval votes, Condorcet would pick a winner among the
candidates who has majority support.