Re: [Election-Methods] Top 5 primary
On Thu, 27 Dec 2007 21:30:08 -0800 Don&Cathy Hoffard wrote: > Dave Ketchum Said > > >Best to leave US Presidential out of this debate because of its > peculiarities. Time enough to go there once the basic topic is resolved > for general use. > > > > >Also matters that Plurality is the major election method in the US. > > Plurality DESPERATELY needs Primaries to try to avoid parties having > multiple candidates to divvy up the vote of party backers. > > > > >This discussion is based on Condorcet, which has no difficulty with a > party having multiple candidates, and backers, should they choose, to > vote for more than one such. > > > > Reply: > > In my example I lumped all of the candidates in one primary election. > This is called a ?Blanket Primary? which the US Supreme Court has rule > as unconstitutional (freedom of association principle). I was just > trying to making a point. > > The Party?s rights to ?Their candidate? (s) is a 1st amendment right. > The reason parties were formed was to avoid the ?spoiler? problem > inherent with plurality voting. If a party had 2 candidates and > another had only one the other party would always win (under plurality > voting). But if we change the rules to Condorcet in the General > election there would be no need to limit their candidates to just one. > Then it would be time to either find a new reason for doing Primaries, or give up on having such, since Condorcet does away with the current basic need. Right now many of us worry about Republican politicians making sure Ron Paul is excluded from the general election by seeing to it that he loses the Republican Primary. > > > We could also have a Party primary that would have a Condorcet winner. > If we are just talking about a general election IRV would work in 90% of > the cases. Why? It is that in most cases the election is between two > major candidates and some minor party candidates, who in most cases have > little or no chance of winning. However, when you have 9 very similar > candidates in a Party primary elections (Democratic primary or the > Republican primary) you need a very different type of processes (like > Condorcet/Range). My point is that Plurality does not would work in the > Democratic or Republican primaries. Although there are no ?head-to- > head? polls for those primaries we can guess the results. Clinton may > be the Plurality winner but Edwards may be the Condorcet winner (Best > candidate) in the Democratic Primary and McCain may be the Condorcet > winner in Republican primary. Which was kind of my point. > IRV will even do fine in most cases of multiple candidates - it only flunks on certain patterns of voting. > > > Let us assume that a State passes a law that the ?winner? (Pledged > delegates) would be the winner of the states Condorcet Party primary. > For example we have 9 candidates in the New Hampshire Primary and the > Sate of New Hampshire decides to uses the Condorcet method to determine > how the State will cast its vote in the Party convention. -- [EMAIL PROTECTED]people.clarityconnect.com/webpages3/davek Dave Ketchum 108 Halstead Ave, Owego, NY 13827-1708 607-687-5026 Do to no one what you would not want done to you. If you want peace, work for justice. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [Election-Methods] Top 5 Primary
Best to leave US Presidential out of this debate because of its peculiarities. Time enough to go there once the basic topic is resolved for general use. Also matters that Plurality is the major election method in the US. Plurality DESPERATELY needs Primaries to try to avoid parties having multiple candidates to divvy up the vote of party backers. This discussion is based on Condorcet, which has no difficulty with a party having multiple candidates, and backers, should they choose, to vote for more than one such. On Fri, 28 Dec 2007 00:39:13 +0200 Juho wrote: > On Dec 27, 2007, at 22:39 , Don&Cathy Hoffard wrote: > >> In this example I will uses the 2008 US Presidential Election only to >> illustrate a point - you could assume a state Governor or Senators race >> >> >> >> We currently have about 25 candidates running for President of the >> United States. >> >> 9 Democrats >> >> 9 Republicans >> >> 3 (assumed) Green >> >> 2 (assumed) Libertarian >> >> 2 (assumed Constitution >> >> >> >> Bases on the current election laws we will have 5 candidates in the >> General Election. >> >> 1 Democrat >> >> 1 Republican >> >> 1 Green >> >> 1 Libertarian >> >> 1 Constitution >> >> >> >> Voters will be asked the vote for one of these candidates. >> >> >> >> Why do we have to choose among the following preferences (using >> head-to-head match ups – i.e. Condorcet) among the 25 candidates running? >> >> >> >> 4 most preferred (Democrat Party/Clinton-Plurality winner) >> >> 5 most preferred (Republican Party/Giuliani-Plurality winner) >> >> 15-19 most preferred (Green Party - assume Nader) >> >> 21 most preferred (Libertarian Party -unknown) >> >> 23 most preferred (Constitution Party -unknown) >> >> >> >> If we have to have 5 candidate in the general election why not have >> the TOP 5 (based on the Condorcet method) Every time I see a magic number I choke. We do need to keep the counts below astronomical, but specific numbers mean unreasonable methods of obeying the restrictions. Ditto for limiting by number how many a party may have. Condorcet cares not beyond needing to have space for its arrays. Election officers need to attend to ballots. Voters care little - they care not beyond being able to rank the handful they desire to rank. So - how might we reasonably limit the candidate count to something manageable without using unreasonable methods? >> > Is the plan is to first arrange a Condorcet election between all > candidates and then between the top 5 candidates? In that case if the > voters have not changed opinion between the two elections (and the > voters vote sincerely) then the results of the second round should > usually be a copy of the results at the first round. In that sense the > results of the first round could be considered also final. But if the > law requires to arrange an election between 5, why not then. > > Too many candidates in an election may be a problem since that makes > voting tedious. One option would be to allow large parties to have e.g. > 2 candidates each and small parties to have only one candidate. In this > case also republicans would have a say on which one of the democrats > will win (assuming that democrats will have majority). I'm not sure if > democrats want that or if they prefer to first elect the "best democrat" > among the democrats and name only that candidate in the final election. > Having several candidates in the final election may make the probability > of electing some of them higher. > > My point is just that running primaries and the number of candidates > each party wants to nominate and the maximum number of candidates in the > final election do have impact on how the system works. > > Condorcet should work ok in all phases (I don't expect strategic votes > to be a major problem). > > Juho Laatu > ... -- [EMAIL PROTECTED]people.clarityconnect.com/webpages3/davek Dave Ketchum 108 Halstead Ave, Owego, NY 13827-1708 607-687-5026 Do to no one what you would not want done to you. If you want peace, work for justice. Election-Methods mailing list - see http://electorama.com/em for list info
Re: [Election-Methods] Top 5 Primary
On Dec 27, 2007, at 22:39 , Don&Cathy Hoffard wrote: In this example I will uses the 2008 US Presidential Election only to illustrate a point - you could assume a state Governor or Senators race We currently have about 25 candidates running for President of the United States. 9 Democrats 9 Republicans 3 (assumed) Green 2 (assumed) Libertarian 2 (assumed Constitution Bases on the current election laws we will have 5 candidates in the General Election. 1 Democrat 1 Republican 1 Green 1 Libertarian 1 Constitution Voters will be asked the vote for one of these candidates. Why do we have to choose among the following preferences (using head-to-head match ups – i.e. Condorcet) among the 25 candidates running? 4 most preferred (Democrat Party/Clinton-Plurality winner) 5 most preferred (Republican Party/Giuliani-Plurality winner) 15-19 most preferred (Green Party - assume Nader) 21 most preferred (Libertarian Party -unknown) 23 most preferred (Constitution Party -unknown) If we have to have 5 candidate in the general election why not have the TOP 5 (based on the Condorcet method) Is the plan is to first arrange a Condorcet election between all candidates and then between the top 5 candidates? In that case if the voters have not changed opinion between the two elections (and the voters vote sincerely) then the results of the second round should usually be a copy of the results at the first round. In that sense the results of the first round could be considered also final. But if the law requires to arrange an election between 5, why not then. Too many candidates in an election may be a problem since that makes voting tedious. One option would be to allow large parties to have e.g. 2 candidates each and small parties to have only one candidate. In this case also republicans would have a say on which one of the democrats will win (assuming that democrats will have majority). I'm not sure if democrats want that or if they prefer to first elect the "best democrat" among the democrats and name only that candidate in the final election. Having several candidates in the final election may make the probability of electing some of them higher. My point is just that running primaries and the number of candidates each party wants to nominate and the maximum number of candidates in the final election do have impact on how the system works. Condorcet should work ok in all phases (I don't expect strategic votes to be a major problem). Juho Laatu They would be the following (assuming the Primary Election matches the Dec. Polls): Edwards (would beat all Republicans and Democrats based on Condorcet rankings see below) (tie) McCain (tie) Obama Clinton Giuliani We could then chose the winner (and next president) using the Condorcet method (or use range, approval, or IRV methods) If the General Election voting matched to polls then Edwards would be elected President I’m not necessarily voting for him, this is only the Voters Preferred President. Thanks Don Hoffard Poll results: Margins of Victory (based on Dec. Polls) Vs. Giuliani Romney Huckabee McCain Thompson Average Clinton 6 11 10 -2 13 7.6 Obama 7 13 15 0 12 9.4 Edwards 9 22 25 8 14 15.6 Average 7.3 15.3 16.7 2.0 13.0 10.9 Condorcet ranking (assumed based on margins and averages) 1 Edwards 2 tie Obama 2 tie McCain 4 Clinton 5 Giuliani 6 Thompson 7 Romney 8 Huckabee Election-Methods mailing list - see http://electorama.com/em for list info Election-Methods mailing list - see http://electorama.com/em for list info
