Another flaw in monotonicity

[Blake Cretney]

On Fri, 30 Oct 1998 09:44:25   Charles Fiterman wrote:
>
>In violent movements your worst enemy is likely
>to be close to you not far away. If someone

I tend to hope most voters aren't members of violent movements.

>votes
>
>A>B>C>D>E>F>G
>
>It is likely that A and G will be trivially
>different people in the same movement.

If you really believe that not even 2nd preferences are meaningful, how
can you favor any method but plurality.  Except that that wouldn't
violate monotonicity.  How about a Second Past the Post method where
the winner is the person who gets the second highest number of
votes.  This seems like the natural method to advocate based on
your standards.

>
>When they judge figure skaters the high and
>low scores are discarded. This means people
>who would have voted a 10 may say 9.9 to
>prevent having their votes thrown away. 

That isn't a rational strategy and does not represent a violation of
monotonicity.  If I vote 10, and it gets thrown away, I am preventing
another high score from being thrown away instead.  To prove a
violation of monotonicity (or practically anything for that matter) you
should include a 2 election example.

>Dishonest voting.
>
>Since the system is easy to change the fact
>that it is kept means it must be broadly seen
>as working well. Indeed I haven't heard anyone
>suggest a change in that rule.
Actually, if we were to change the rule, I would suggest using the median
score.  This is more in keeping with majority rule.  Interestingly,
the median is also the Condorcet winner for a linear example like this. 
>
>
---

Blake


-----== Sent via Deja News, The Discussion Network ==-----
http://www.dejanews.com/  Easy access to 50,000+ discussion forums