Insincere Equal Rankings

[Mike Ositoff]

I'm replying to Hugh Tobin's letter, writting while I was temporarily
offline.

Let's say that it's true that the Least-Margin-Of-Defeat method
(Young's method) really never violates No Show or Positive Participation.
By all means question Smith//Condorcet(EM) or Schulze, but are you
saying that Young's method is better than those if it meats the
No Show & PP criteria? 

As I've said, those are fine-tuning criteria. Failure of them
provides an opportunity for offensive strategy, to the person
with the virually impossible ability to predict when & how to
exploit those criterion-failures.

But where Young isn't as good as SC & Schulze isn't a matter
of fine-tuning. It's a quite gross problem.

You mentioned truncation. I've posted examples here where
Young allowed truncation to elect the truncators' favorite,
defeating the Condorcet winner. That's quite impossible in 
Condorcet(EM) or probably Schulze.
 
I've posted at great length, a long time ago when we discussed
Young vs Condorcet(EM) before, about the _gross_ failures of
Young, failures that don't happen with Condorcet(EM) or Schulze.

About vulnerability to truncation, yes that's a problem that
tends to haunt all pairwise count methods. But not Condorcet(EM)
or Schulze.

With those 2 methods, you won't need defensive strategy to protect
a Condorcet winner unless someone attempts the offensive strategky
of order-reversal (or maybe some complicated offensive strategyk
based on predicting the exact nature of a failure of No Show or PP).

With Young, where mere truncation can defeat a Condorcet winner and
make the truncators' candidate win, the statement in the paragraph
before this one can't be said. That means that Young, in a quite
gross way, doesn't do as well as SC & Schulze.

Mike