Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Blain]

Is it me, or does this method of voting seem foolish?  Why play games with it?  If there is a favored state, it should win, right?  

[RidleyReport]

I just sent my ballot (which Jason thoughtfully  e-mailed to me when nothing showed up in my mailbox).

I voted:

NH>WY>ID>MT>SD>DE>AK>ND>ME>VT

[varrin]

Nope.  NH does not lose; therefore, NH is the winner.

Ahhh.... I get it.  I'm glad you're here  Maybe I should rerun it with the real data...  Or maybe I'll just wait until we have just a bit more data...

V-

[LeRuineur6]

Is it me, or does this method of voting seem foolish?  Why play games with it?  If there is a favored state, it should win, right?  

Condorcet's is the best method I've ever seen of finding the true desires of voters.  It is difficult to grasp at first, but simple to understand how it works once you realize the basics.

For example, if more voters rank A over B than rank B over A, then A should win, right?  Under Condorcet's, A will win!

Under some other methods, people can manipulate the vote by placing C, D, and E between them.  Under Condorcet's this method of manipulation does not work.

Condorcet's takes the true desires of all voters into consideration.  A Condorcet's vote cannot be manipulated.  It is simply unfeasible to do so.

[Blain]

Once I understand how it works, I am sure I can find some flaw in the system.  

[Zack Bass]

Once I understand how it works, I am sure I can find some flaw in the system.
 

If you could see others' votes before you cast yours, you might be able to manipulate it.  But not the way we're doing it.
The only thing you can do to try to manipulate it is to put one preference before another... and how could saying you prefer one over another, when you don't, possibly help you?

[Blain]

because 10th places represent undesirables (or they should) and thus they should hold less weight than 1st, 2nd, and 3rd choice picks.  If they do not it is just messed up and WRONG!!!!!!!!!  If they do than that is one way you can manipulate it, rank the states you do not like the lowest.  Otherwise, it don't matter what order you rank states in.  

[Zack Bass]

because 10th places represent undesirables (or they should) and thus they should hold less weight than 1st, 2nd, and 3rd choice picks....  If they do than that is one way you can manipulate it, rank the states you do not like the lowest.  Otherwise, it don't matter what order you rank states in.

Then you are indicating your preferences.  That's not Manipulation, that's Voting.

[Elizabeth]

The sentence I use when explaining Condorcet to a reporter is, "It's a system that allows for the state that's ranked the highest by the most number of people to win."  This is overly simplistic, but gets the point across.  It's the best system for a compromise candidate, and to prevent spiteful and strategic voting.  

There are already a lot of threads on how Condorcet works.  I just merged them all, you can read it here:
http://forum.freestateproject.org/index.php?board=3;action=display;threadid=2225

I should note too, that we consulted with election experts and game theorists around the world, and Condorcet was the clear answer we were given.

[Justin]

because 10th places represent undesirables (or they should) and thus they should hold less weight than 1st, 2nd, and 3rd choice picks.  

Nope. Condorcet uses pair-wise comparisons.  Using a ">" list is just a convenient notation.  For example, the more accurate notation of B > D > A > C would be:

Options	Preference
A or B	B
A or C	A
A or D	D
B or C	B
B or D	B
C or D	D

As you can see, the number of choices more preferable than C but less preferable than B have no impact on the magnitude of how C compares to B, (B or C)... B.

[Summerlin]

Under 50 people have voted on this thread and a few of you are gloating that "it looks like NH wins!"

 

[rdeacon]

You're right.  We should save our gloating until after the press conference.

 

[varrin]

Under 50 people have voted on this thread and a few of you are gloating that "it looks like NH wins!"

I don't think it's a foregone conclusion that NH has won.  Of the votes posted here, there were 21 that contained enough data to rank NH, ID and WY (I'm now including my vote since I posted all of it a couple posts back).  I believe these are the three front runners, though MT may well stand a chance also.  But due to lack of data, I could only get 21 good votes.  Here's the raw data for those 21 votes:

2 WY=ID>NH
3 WY>ID>NH
3 WY>NH>ID
5 NH>ID>WY
2 NH>WY>ID
6 ID>WY>NH

Here are the total first place votes, first without the ties, then with the ties:

WY: 6 (no ties)  8 (with ties)
ID:  6 (no ties)  8 (with ties)
NH: 7  (no ties)  7 (with ties)

Either way, they're relatively even in first place rankings.  However when you run Condorcet's method, there is a clear winner.  The winner is *not* NH.  Here's the table:

   WY      ID      NH   Losses
WY   X      8      14   1
ID   11      X      11   0
NH   7      10      X   2

ID comes out undefeated.  WY loses to ID.  NH (believe it or not) loses to ID *and* WY.  How can NH lose to ID *and* WY?  Look at the raw votes.  5 of 8 (including ties) WY voters put NH below ID and *all* of the ID votes put NH below WY.  ID does well because WY voters are exactly split on the ID/NH ranking but NH voters heavily preferred ID to WY (5 to 2).  Hence, ID beats WY (decided by the NH voters) and ID beats NH (decided by the WY voters).  WY winds up beating NH primarily because *all* of the ID voters preferred WY to NH.

I suppose it's difficult to predict how people will vote, but outside of this forum (and this has been discussed before) there is one clue as to how people may vote: where they're from.  There are 5,000 potential voters.  They may not all vote and there's probably no way to predict which of them will simply not vote.  But assuming the ratios are correct, I'll go out on a limb here and speculate as to how a few of them will vote.  Remember, we don't need to know exactly how they will vote.  If (and this is a big 'if') we assume WY, ID, and NH are the front runners, predicting a winner between those three will only require us to predict how people will rank them in relation to each other, not their entire vote or where any specific state falls in their vote.

Also, it's obvious that these are just pure speculations and that using these numbers is applying a single vote to all the members from a specific state (which will certainly not be the case).  I'm going to assume that half of WY-first voters will vote WY>NH>ID and half will vote WY>ID>NH.  I'll also assume that all of the ID-first voters will vote ID>WY>NH and that 5 of 7 NH-first voters will vote NH>ID>WY and the other 2 (of 7) will vote NH>WY>ID.

Here's what I think (give or take) will be the majority of first place votes out of these states with more than 100 members:

AZ   112   WY
CA   527   ID
CO   117   WY
FL   307   ID
GA   176   WY
IL   140   NH
MA   113   NH
MI   113   WY
NC   145   NH
NH   154   NH
NJ   121   NH
NY   193   NH
OH   112   NH
OR   102   ID
PA   206   NH
TX   274   WY
VA   123   NH
WA   170    ID

WY = 792
ID = 1106
NH = 1307

WY>NH>ID = 396
WY>ID>NH = 396
NH>ID>WY = 934
NH>WY>ID = 373
ID>WY>NH = 1106

The table looks like:

   WY      ID      NH      Losses
WY   X      1165      1898      1
ID   2040      X      1502      1
NH   1307      1703      X      1

There is a tie with one loss each. The three losses were: WY<ID=1165<2040, ID<NH=1502<1703, and NH<WY=1307<1898.  The smallest magnitude loss is ID<NH hence ID wins the tie and the competition.

I'm lousy at predicting things, but I got a buck says it's going to be closer than this thread illustrates.  As you can see from the above (obviously highly realistic ) table, a few votes one way or another could push any one of the states over the edge.  Also keep in mind that with the highest number of first-place votes, NH did *not* win the election.  

Okay, I've wasted enough time on this one..

If more people want to post their entire rankings, I'll check back and update the score...

V-

P.S.  Did I do it all correctly?  Any Condorcet expert, feel free to critique my work here (the tables and outcomes, not the guesses as to how people might vote ))))

[Bonner County]

IIRC
        ID>AK>WY>MT>SD>ND>ME>DE>VT>NH

[Blain]

You did it INCORRECTLY, because most people only listed their first/top choice, like I asked them too!  I ranked Idaho like 8th.