From: Russ Paielli <[EMAIL PROTECTED]> Subject: Re: [EM] Re: simulating an Approval campaign/election
Rob LeGrand honky1998-at-yahoo.com |EMlist| wrote:You're simulating a DSV (Declared-Strategy Voting) election with Approval. My current research is on just that topic, though I'm also interested in using DSV with other point-count systems such as plurality, Borda and several others. That Approval strategy is identical to strategy A in the 3-candidate case.
Russ asked:
Interesting. Do you mind if I ask why you are interested in Declared-Strategy Voting as opposed to Undeclared-Strategy Voting?
Rob can give you his own reply on this, but here is mine:
Read Lori Cranor's dissertation on Declared Strategy Voting. [Just Google "Lori Cranor" or "Declared Stratetgy Voting".]
The original idea was to have voters submit their preferences (i.e. ordinal rankings of the candidates) and also their own strategy for plurality voting given partial results (the plurality totals for the ballots counted before theirs), and have the computer implement their strategy as their ballot came along for incorporation into the tally.
This idea might have been inspired by the U.S. Presidential election which is a plurality election in which voters on the West Coast can wait for partial results to come in before deciding how to vote.
Later the original idea was modified by using a common (near optimal) strategy for everybody.
Finally, batch mode was introduced to take out the arbitrariness of the order of processing of the ballots.
Rob and I felt that DSV would be more stable when based on Approval rather than Plurality. His simulations have borne this out. Furthermore Approval Strategy A is much, much simpler than the complex strategy needed for Plurality (which involves calculating intersections of normal distributions based on ever evolving estimated parameters).
It is important to note that the US presidential election is a kind of biased (towards the West Coast) DSV election. And since DSV Approval does yield the CW when there is one, adoption of Approval in the U.S. Presidential election would almost certainly result in a win by the CW when there was one, unless the East West bias was too great.
This could be improved upon by having all voters with SSN's ending in (say) 5736 vote two days before the rest of the voters, and the results announced immediately. (The special number would be drawn anew for each election.)
Forest ---- Election-methods mailing list - see http://electorama.com/em for list info
