[EM] Intro to list (etc)

[Dave Ketchum]

I read these two threads thru 2130 EST on Monday, but choose to respond to
this original.

Looks like a GREAT idea, though a few details disturb me.

You talk of a primary customer, who would give their users some experience 
with Condorcet, but as individual voters with no choice as to details of 
the implementation.

Doing that well takes what I consider to be an almost complete program, so 
I would lay out design for a more complete program, planning to implement 
only the subset for starters, and then completing the rest if/when that 
seemed worth the extra effort.  Some details:

List of candidates - done by whoever sets up election.  Can have long 
names but MUST have a unique label of not more than 4 characters, to allow 
for max quantities of candidates to be displayed in vote matrix.

Vote - every voter can vote, so this must be easy - perhaps permit voter 
to use either full or above short labels - there are other 
possible methods.  For a simulation mode, let "voter" vote as multiple 
voters choosing a vote pattern.

Display ballot count matrix, as if this was last voter before polls 
closed.  Could make sense to display this while the voter is voting, 
stepping as the voter goes thru ranking candidates from first to last 
voted by this voter.  This ONLY starts with matrix as of voter starting to 
vote, incremented according to voter's current proposed vote.

Calculate winner - always doable based on above activity, but meaningful 
only after polls have closed.  In simulation mode this is the major 
opportunity for offering various Condorcet-class methods, for taking turns 
with various methods to process a single results matrix.
      Standard method:  I LIKE winning votes and Ranked Pairs.
      Simulation methods:  Throw in all the choices anyone can make 
sufficient case for.
      Simulator supplied:  Perhaps allow for simulation users to provide 
their own calculation method.
      IRV:  Could make sense to include this as a simulation choice - 
tempting, but demand a STRONG case before investing this effort, for IRV 
demands records of how many voters voted each voting pattern used, while 
Condorcet needs ONLY the information placed in the pairs matrix.

Sort and redisplay ballot count matrix, starting with winner in
row 1/column 1?  This would place most of the numbers of any true interest 
in the upper left corner, ready for comparison with adjacent numbers - 
even if you supported having more candidates than display width could 
display at once.

Do graphs?  I am not against this, but not clear to me how to do anything 
useful.  One thing I insist on is that stepping and winner calculation 
SHALL be valid Condorcet operations using whatever method was promised, 
and that the graph subroutine be restricted to doing whatever is 
consistent with that.
      BTW - We seem agreed on ordering candidates in one dimension from 
winner to worst loser.  Just hard to see how to get to such a graph from 
the 2-dimension results matrix (remembering that cycles make this even 
harder to translate).

True results matrix gets stepped according to voter's final proposal only 
when voter says record this as final (remember that each voter could spend 
minutes - even hours - voting while other voters are active on the 
Internet at the SAME time).  Note that this is the step that needs defense 
against double voting in a true election - which still leaves this on the 
internet and thus too dangerous to use when temptation to cheat gets too big.

HELP:  DO provide this.  Major topics:
      What is Condorcet.
      What, in detail, is each method supported (more than one for 
simulation).
      What are cycles?  I suggest "near ties" as less confusing to these 
outsiders, and meaningful to insiders - A>B and B>C while C>A implies 
being close to a tie.

Borda?  Keep this and other election methods and words OUT of this support 
for Condorcet.

Dave Ketchum

On Sun, 26 Oct 2003 11:15:49 -0800 Rob Brown wrote:

 > Hi all,
 >
 > I have lurked on this list on and off for a few years (the whole Nader
 > thing in 2000 really got me interested in how thorougly broken plurality
 > systems are).  Now I'm working on a web-based Condorcet based election
 > system, so I figured I'd drop in and introduce myself, and see if anyone
 > has any thoughts on the issue I'm having.
 >
 > I'm a programmer (c/c++/java/javascript) with an industrial design
 > background, so I tend to be big on UI and graphical stuff. I have also
 > dabbled in fuzzy logic, collaborative filtering, people matching and the
 > like for a good while.
 >
 > Here is a UI I am working on for doing for ranking candidates:
 > http://weblogz.com/voting/2000pres.html
 > This demo is of course based on the 2000 presidential election, and
 > allows you to rank candidates with a (hopefully) friendly UI.  I avoided
 > having people manually assign numbers to candidates  (after all, they
 > are sitting in front of a perfectly good computer which can do that sort
 > of thing well!), and I tried using a little animation, which seems to
 > help in making it easy for voters to follow what they are doing.
 >
 > On the back end, I don't have any problem with figuring out how to
 > tabulate the results, in terms of who is the first choice, who is the
 > second, etc. I am using the Condorcet method, and using  ranked pairs to
 > break ties.  No problem there....
 >
 > Now I need to figure out how to display results, in a way that makes
 > sense to people.  The people I have talked to who are likely to use this
 > system want to see more information beyond just the final ranking of
 > candidates -- they want to see some kind of "score" or a graph.  They
 > are used to web based polls where you can see a nice little graph,
 > showing how many voted for which candidate.  Knowing whether an election
 > was neck-and-neck or a landslide is relevant information people should
 > be able to see, I think.
 >
 > I have tried showing various things:  for instance, I can show a Borda
 > count score.  But of course the Borda score does not always correllate
 > with the Condorcet outcome:  you could have a higher score for a lower
 > placed candidate, and that will confuse people.
 >
 > Another thing I tried was showing a score which is the sum of all
 > margins by which they beat (or lost to) every other candidate.  For
 > instance, if 20 voters placed B above A, and 10 placed A above B, that
 > would add 10 to B's score and subtract 10 from A's score. Again, though,
 > it may not correllate with the Condorcet ranking, although for some
 > reason my feeling is that this type of score is "better" than a Borda
 > count. (is there a name for such a way of scoring?)
 >
 > So has anyone tried showing bar graphs or numerical scores for Condorcet
 > elections?  I suspect that if this is completely impossible to show
 > something reasonable, I won't have much luck selling a condorcet based
 > system to a mainstream web audience....people really seem to want to see
 > something to help them understand the results.  On the other hand, my
 > gut feeling is that there MUST be some way to have Condorcet results
 > equate to numerical values.
 >
 > Any thoughts?
 >
 > -rob
-- 
   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.