[EM] Re: What respected IRVists are saying about Condorcet, Approval, etc.

[Dave Ketchum]

On Thu, 10 Mar 2005 16:07:53 -0800 Ted Stern wrote:

> On 9 Mar 2005 at 11:06 PST, Eric Gorr wrote:
> 
>>I attached a document, which I came across recently, being distributed 
>>by an IRVist (Jim Lindsay) to fellow IRV supporters and found it quite 
>>interesting.
>>
>>Thought others might be interested as well in what is being said about 
>>Condorcet, Approval Voting, etc.
>>
> 
> Who is Jim Lindsay?  Where is he located?  I know someone by that
> name.
> 
> Surprisingly, Approval looks better in the table compared to IRV except
> for already being used in public elections and the "Majority Vote
> Wins" comparison (Approval: "usually").
> 
> But with Condorcet, he uses the potential worst possible case as a
> straw man.  These are the key comparisons:


Since IRV and Condorcet are looking at THE SAME ballots for the STRONGEST 
candidates, they will usually agree as to winner.  They usually disagree 
when it matters that IRV does not look at complete ballots or Condorcet 
sees a cycle.  Thus:

> 
>                                  IRV    Condorcet
> --------------------------------------------------
> Ties are unlikely or
> easily adjudicated               YES    no


True ties should be equally unlikely.

Adjudicated?  Why a difference?


> --------------------------------------------------
> Encourages 'honest' voting; 
> discourages strategic voting"    YES    depends


Tell me more.


> --------------------------------------------------
> Encourages cooperation and
> more positive campaigns.         YES    depends


Tell me more - I DO NOT SEE THIS.


> --------------------------------------------------
> Requires enthusiastic
> support to win                   YES    no


Tell me more - I DO NOT SEE THIS.


> --------------------------------------------------
> 
> I'm sure many list readers could argue the correctness or importance
> of these comparisons.  These are the only real contrasting criteria
> for preferring IRV to Condorcet.  Here is the remaining text of the
> article for those without a way to read pdf [Sorry, I wish I had a
> better way to convert the table to text!]
> 
> ,----
> |    Plurality is a very bad system. It is the easiest system to
> |    administer, but that is about all that can be said for it. It
> |    doesn't even require a majority vote to win, which is
> |    ridiculous. It strongly discourages alternative parties, so if
> |    one is a strong believer in the two party system, one might want
> |    to be in favor of plurality voting. Traditional Runoff
> |    (i.e. where the runoff is 1-9 months later) at least ensures a
> |    majority winner. However, if the runoff is held as a special
> |    election, it is common to have a a 50% dropoff in the vote, so
> |    only those people determine the ultimate winner. Alternately, it
> |    may be consolidated with the next major election date, in which
> |    case it isn't rare for 8 months to pass between the first round
> |    and second round of the election. Finally, they are quite
> |    expensive to administer, costing jurisdictions a lot of money.


Agree most of the way.

IRV and Condorcet do not require a majority vote (of all the voters) to 
win.  IRV discards ballots that only vote for rejects, and then brags that 
the winner got a majority of the remaining ballots.

> |
> |    IRV is probably overall the best single-winner system, based on
> |    the criteria above. It is a little harder to explain than other
> |    systems, but not too hard. It is a good default choice for
> |    single-winner elections. In IRV, a compromise candidate that does
> |    not have a strong base of enthusiastic support will not get
> |    elected. However, that candidate's voters will have a lot to say
> |    about who wins, because the winner of the election will probably
> |    be based on to whom that candidate's #2 votes transfer.


As I said above, IRV and Condorcet use the same ballot, and usually agree 
as to winner - please explain why these descriptions should not be 
rejected as fairy tales.

> |    
> |    Condorcet is a favorite of many mathmaticians. It is a very
> |    strong system if a tiebreaking procedure is used that discourages
> |    bullet voting. (Unfortunately ties would probably be common in
> |    this system, and good Condorcet tiebreaking procedures are very
> |    complex.) Condorcet tends to elect compromise candidates, so it
> |    might be well utilized for a "healer" type executive
> |    position. Let's say that a group has had a lot of internal
> |    fighting. Electing the President of this group via Condorcet
> |    might be good choice in this situation. Condorcet would be a
> |    perfectly reasonable choice for electing a mayor of a city,
> |    too.
> |
> |    Approval and Borda voting appear to be good choices for fairly
> |    homogenous organizational elections. E.g., the IEEE and the
> |    American Mathematical Association use Approval Voting, and to all
> |    accounts, it has worked well for them. In emotional public
> |    elections, however, these systems would probably devolve to
> |    plurality, as candidates would ask their voters to bullet vote,
> |    since voting for anyone else other than ones' favorite candidate
> |    might very well cause that favored candidate to lose.


Sounds to me like an argument for moving up from Approval to Condorcet. 
Have not been seeing claims that Borda is in this league.

> `----
> 
> Jobst Heitzig's recent Total Approval Chain Climbing (TACC) proposal,
> when combined with an extra "Minimally Qualified" fictional candidate
> as a way to handle the approval cutoff, appears to satisfy all of the
> negatives above (Thanks for directing me to this, Forest!):
> 
> http://lists.electorama.com/pipermail/election-methods-electorama.com/2005-March/014924.html
> 
> It appears to strongly discourage bullet voting and encourage full
> sincere ranking and generous approval cutoff.  I think it may be as
> easy or easier to explain than any other Condorcet Completion method,
> including Ranked Pairs.
> 
> Ted

-- 
  davek at clarityconnect.com    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.