
There is
considerable debate in the sudoku community regarding the use of tests
for uniqueness.
Solving techniques should be based on mathematical certainty. A
candidate can be eliminated as it is IMPOSSIBLE for it to exist in a
given cell (for example, regardless of puzzle validity, it is
impossible for a candidate to be present twice in the same house).
Tests for uniquess suggest
that a situation is impossible UNLESS the puzzle is invalid.
A puzzle is invalid if:
 It has no possible solution
 It has more that one solution
Uniqueness techniques can be used to prove that a puzzle has at least
one solution but it cannot rule out multiple solutions.
So if a user trusts that a puzzle from a reputable source is valid
then tests for uniqueness can be employed.
But skeptics who are fearful of making this assumption cannot use
these solving techniques with confidence.
(I'll leave it there. This subject have been beaten to death in
other forums)

