Copeland's method: Difference between revisions
Content added Content deleted
(Categorize criterion compliances and failures, add independence of covered alternatives failure) |
(Add more criterion failures.) |
||
Line 48: | Line 48: | ||
==Criterion failures== |
==Criterion failures== |
||
=== Schwartz === |
|||
In this election: |
|||
{{ballots| |
|||
1: A > B3 > B1 > B2 > C |
|||
1: B3 > B1 > B2 > C > A |
|||
2: C > A > B2 > B1 > B3 |
|||
}} |
|||
The Schwartz winner is C, but Copeland elects A. |
|||
=== Independence of clones === |
|||
Copeland's method is vulnerable to crowding. This alternative is [[w:Independence_of_clones_criterion#Copeland|due to Wikipedia]]. |
|||
First consider the election |
|||
{{ballots| |
|||
1: A>B>C |
|||
1: B>C>A |
|||
2: C>A>B}} |
|||
C is the Condorcet winner and thus also the Copeland winner. Now clone B into B1, B2, and B3: |
|||
{{ballots| |
|||
1: A > B3 > B1 > B2 > C |
|||
1: B3 > B1 > B2 > C > A |
|||
2: C > A > B2 > B1 > B3 |
|||
}} |
|||
The Copeland winner changes to A. |
|||
Since the Copeland winner is unique in both cases, every method that elects from the Copeland set must also fail clone independence. |
|||
=== Independence of covered alternatives === |
=== Independence of covered alternatives === |