[EM] Proposed method of cycle resolution

[Kevin Venzke]

Diana,

It seems to me that a cycle participant could benefit, with your methods, by running 
along with a candidate who is there only to be crushed.

It seems to me that cycle resolution methods should assume that any combination
of candidates may be clones.  Huge, crushing defeats shouldn't necessarily 
reward the victor.  The method probably should do no arithmetic with defeats (such 
as adding up defeat strengths, or number of defeats) other than comparisons.

For instance, RP sees the strongest defeat and merely bars its loser from winning.
Other methods try to elect a candidate whose greatest defeat is the mildest.  You
can't benefit from crushing weak candidates with such rules.

Kevin Venzke
stepjak at yahoo.fr



___________________________________________________________
Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en français !
Yahoo! Mail : http://fr.mail.yahoo.com