[EM] FBC, center squeeze, and CD

[Michael Ossipoff]

I'd said:

>
> If A was the CWs, then plumping by even 100% of the A voters wouldn't have
> foiled the burial by the B voters.
>
> You wrote:
>


> Mike,
>
> Thanks for pointing that out!
>
> 46: A
> 44: B>C  (sincere is B>A)
> 10: C
>
> 100 ballots.  C>A 54-46,   A>B 46-44,   B>C 44-10.     Top Ratings:  A46 >
> B44 > C 10.    Approvals: C54 > A46 > B44.
>
> A is the sincere CW.  MDDTR elects B.
>

Sure. I said that.


I'd said:


> With wv-like strategy, truncation from one side can't take victory from
> the CWs & give it to the truncators' candidate.
>
> ...and plumping by the CWs's voters makes it impossible for burial to
> succeed.
>
>
>
>


> From this we see that MDDTR  *doesn't* have "wv-like strategy".
>

Incorrect.

Here's what I just finished saying:

"Your "CW" (Candidate A) is a peculiarly unsupported CWs.

"   "Protecting" such a CWs wasn't what wv was proposed, offered &
advocated for.

"When I first proposed wv in the late '80s, I showed its "protection" of a
middle CWs who is supported from both sides. I never claimed that it could
"protect" a CWs who is entirely unsupported by preferrers of any other
candidate. "Protection" of an entirely outside-unsupported CWs was never
the intended purpose of wv.

"You "CWs" is entirely unsupported by preferrers of any other candidate."

"wv-like strategy" can only refer to strategy-properties that I claim for
wv. At no time did I ever guarantee that wv will "protect" the CW under the
conditions of your example.

No such guarantee can be made, not have I made one.

"But wait, wv _does_  thwart & penallize burial in Chris's example!"

...but not because of any guarantee that can be made for wv. Not because of
any guarantee-able & guaranteed wv property.

In your example, wv elects C, thwarting & penalizing the burial, because of
an accident of faction-sizes.

(Try reversing the size-relation of A & B.)

Sure, by suitable adjustment of faction-sizes,any pairwise-count method
could achieve (whatever you think is) the right result, in a cycle.

So, not electing B in your example is not a test of "wv-like strategy".

Your example isn't a test of "wv-like strategy", and has no resemblance to
the conditions for which that property was defined, and for which wv was
proposed.

And yes, I made it clear, with many examples, what wv's desirable
properties are.


> A is the sincere CW and all of  A's supporters plumped for A,
> but nonetheless the Buriers' favourite, B, won.  Winning Votes would have
> elected C.
>

...by an accident of faction-sizes, not as an instance of any guarantee
that can be or has been made for wv properties. Electing a CWs with no
support from preferrers of any other candidate has never been claimed as a
wv property, and is not a "wv-like strategy" property.

And I'll remind you that you neglected to comment on vulnerability to
truncation. Could that be because, even in your example, truncation by B
voters wouldn't succeed in taking the election from A to elect B?

I'll also remind you that truncation from one side, which could be innocent
(careless, hurried, lazy or principled), will likely be considerably more
frequent than burial.

But, in MDDTR, voters should be advised to not rank short of the CWse
(expected or evident CWs), because indifference is the threat to MDDTR's
success. Any candidate rank-order distinction that you don't bother to make
when you vote, just might not be reflected in the outcome.

But, then, that's pretty much true of rank-methods in general.


>
> Also we see that MDDTR  fails the Plurality criterion.
>

Yes, in the standard chicken-dilemma example, MDDTR elects A, and that's a
violation of the Plurality Criterion. Try to forgive MDDTR for electing the
most favorite-popular candidate who isn't majority-beaten  :^)

(Of course that's a way of saying MDDTR's rule but it hardly sounds
unjustified, by general objective standards, independent of any rule.)



>
>
> So to sum up,  MDDTR  doesn't have "wv-like strategy"
>


Incorrect. See above.


> , but does fail  Mono-add-Plump
>

...but IRV's, Benham's & Woodall's failure of Mono-Raise is ok  :^)

Mono-Add-Plump failure is where X loses _in spite of_ your favoring X on
the ballot that you add. X is made to lose because you added a ballot, and
the fact that you favored X on that ballot has nothing to do with how or
why it happened. X would have been made to lose if you'd added any of many
other kinds of ballot.


Mono-Raise failure: X loses for no reason other than because you moved X up.

Notice the difference?




> and Plurality
>


I answered, above, about Plurality.


> and has a strong random-fill
> incentive.
>

Because of LNHa, you can safely rank your inbetween (neither strong top-set
not strong bottom-set) in sincere order, below top & above bottom.
...ranking them all with respect to eachother, and giving to all of them
the "protection" of wv.  Adding a candidate to that ranking can't harm
anyone whom you've ranked higher.



>
> I vastly prefer 3-slot TTR,TR  (aka ICT).
>

It doesn't have "wv-like strategy". When you demote a good candidate whose
voters you don't trust, you're giving hir no "protection" from burial, and
hir voters have no way of "protecting" hir from burial.  And s/he also has
no "protection" from truncation, including innocent, non-strategic
truncation.

Michael Ossipoff

>
> Chris Benham
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20161109/b93c23ba/attachment.htm>