Talk:CPAR voting

15: A>B 7:A 33: B>A 6: C>B 38: C 34: ABC (or >ABC, or mixed between X>ABC and Y>ABC)

Note that this is not far from being a Condorcet cycle, C>B>A>C, but actually B>C barely.

In particular, the "cyclicness" is more than half the B>A margin. Which I think is the basic fact that must be true for FBC to be violated. Which means that while candidate A is the one to benefit from the lack of favorite betrayal, factions B and C must stand in an asymmetrical (quasi-cyclic) relationship in order for this to happen. There's no way faction A could engineer this situation; they could only at best try to take advantage in the rare cases that it happened to obtain (and they knew that it did with enough precision).