Coalition For Condorcet Voting

[Emsworth]

I'm afraid Condorcet suffers from many of the same theoretical flaws as IRV. It is possible for someone to rank A over B and thereby change the winner from A to B. One can hurt a candidate by voting for him, and help a candidate by voting against him. Thus, it is susceptible to tactical voting just like preferential voting.

[Emsworth]

I'm afraid Condorcet suffers from many of the same theoretical flaws as IRV. It is possible for someone to rank A over B and thereby change the winner from A to B. One can hurt a candidate by voting for him, and help a candidate by voting against him. Thus, it is susceptible to tactical voting just like preferential voting.

Can you give an example?

I'll try to come up with one tomorrow. From the research of voting systems I've done for school, I believe that the example involves who goes into the Schwartz set.

[Emsworth]

Alright, here's an example (numbers are based on another source - they're not my own):

The pairwise rankings are:
Supersoulty-Harry: 20-20
Supersoulty-Siege: 30-30
Supersoulty-JFK: 30-30
Supersoulty-King: 30-30
Supersoulty-Phil: 24-36

Harry-Siege: 34-26
Harry-JFK: 30-30
Harry-King: 30-30
Harry-Phil: 38-22

Siege-JFK: 36-24
Siege-King: 22-38
Siege-Phil: 30-30

JFK-King: 42-18
JFK-Phil: 30-30

King-Phil: 32-28

In this hypothetical, Supersoulty is the winner.

Suppose that three voters who support JFK tactically vote: Supersoulty, King, Phil, Siege, Harry, JFK. Although they preferenced JFK last, they've just made him the winner over Supersoulty.

[Emsworth]

It would scare away newbies and put counting votes into the hands of the forum members who are math geniuses. 

Shhhh... The math geniuses are the ones who make up the Forum Elite...

[Emsworth]

It's not Condorcet, and so relies on exactly what type of Condorcet voting you use.

I know that there is no Condorcet winner - Supersoulty wins the hypothetical because he has the appropriate beatpaths.