[EM] RE: assumption of sincere ballots

Simmons, Forest simmonfo at up.edu
Thu May 19 14:33:43 PDT 2005


 

________________________________

In the context of this thread Chris wrote ...
 
Normally if  there is a lot of ill-will between rigid majority and
minority factions, then the majority will not assent to a raffle, and if
there is one the losing side will suspect that it was
rigged.


I reply:

Under DFC a rigid majority would not have to assent to a "raffle," since a majority can pick the DFC winner with certainty by approving only one candidate.

In the scenario I suggested  (sincere 60 H>M>>T,  40 T>M>>H) the 60 Hutus could easily adjust their ballots to H>>M>T and thereby make H both the Condorcet Winner and the Approval Winner, hence the only member of P, the set from which DFC chooses.

But if the official assumption of  sincere ballots in the altruistic public interest is realistic, why is it less realistic to assume that they would reject a raffle that gave a 60 percent chance for their first choice and a 40 percent chance for their (also approved) second choice?

By the way, I ended up in Vietnam because of a draft lottery that wasn't run properly.  They put all of the birthdates in a container, stirred a little, and then drew them out, one by one.  Naturally the December and November birthdays came up first.   After that drawing some statisticians noticed the pattern and a new (more random) system was adopted for subsequent years.  Under the old system November 4th was the third birthdate drawn, so my draft number was three.  Nobody knows what my draft number would have been under the new system.

There was no outrage over this mistake even though many lives hung in the balance.

We now know of fair and transparent ways of doing random drawings.

Forest

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/ms-tnef
Size: 4962 bytes
Desc: not available
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20050519/ef23ad5f/attachment-0002.bin>


More information about the Election-Methods mailing list