[EM] Craig's question

[Bart Ingles]

Craig Carey wrote:
>
>    I ask people that post to state their position on truncation resistance.
>    This asked to all subscribed, provided they post a lot or they want to
>    make a statement on this (yes/no) question.
> [...]
>       A definition of Truncation Resistance:
>       That is the principle that says that for each candidate, and for all
>       STV style ballot paper collections, the alteration of preferences
>       after than candidates preferences makes no difference to the win-lose
>       state of the candidate. [Papers not holding a preferences for the
>       candidate can't be altered. More than one paper can be altered.]


I consider that property unnecessary, and probably counterproductive. 
The opposite might be needed to help overcome shortcomings of ranked
systems.


My view regarding truncation:

Assuming three candidates A, B, and C, where B is always ranked lower
than A, and C's ranking is unspecified,

If a voter raises B's ranking, regardless of whether he/she raises B
above A's ranking:

 1) It is acceptable if this causes B to win (thus causing either A or C
to lose), so long as this possibility is predictable & known to the
voter ahead of time.

 2) It is unacceptable if it causes C to win if C would not have won
otherwise.