[EM] Methods based on sequential voting

[raphfrk at netscape.net]

From: Simmons, Forest <simmonfo at up.edu>

> In some elections not all of the ballots are cast at the same time, 
and
> furthermore, the partial results (from exit polls, say) may be 
available to
> voters later in the sequence.

> In any case, some voters  may prefer to vote near the beginning of 
the sequence,
> others may prefer to vote near the end, when the partial results make 
optimal
> strategy more obvious, and others may prefer to vote near the middle.

> (1)  Have each voter specify exactly how much weight would have to be 
attached
> to the first vote in order to make it equal (in her opinion) in value 
to the
> average of the remaining votes.  The voter that bids the lowest, gets 
the first
> position. Apply this procedure repeatedly among the remaining voters 
for the
> remaining postitions until everyone has a place in line.

A simplier system would be to just have polling split over say a week 
with the votes for
each day are weighted slightly differently.

The first day might be 1.2 and the last day 0.8.  People who don't care 
what others think
will vote on day 1.  People who want to know will vote on the last day.

This should be enough granularity.  If everybody votes on the first day 
or the last day,
then the range needs to be widened.

It means that you don't need to poll people about what they would offer 
etc.  People
just vote on the day that is best for them.

There would likely need to be rules about the counting.  It would be 
pretty annoying
if you waited a day and the count for the previous day wasn't ready.  
Also, the rule
for counting should be that the estimates need only be approx.
___________________________________________________
Try the New Netscape Mail Today!
Virtually Spam-Free | More Storage | Import Your Contact List
http://mail.netscape.com