Why the unrealistic example is unrealistic:
40A 35C>B 30B
This example is not realistic because it is extremely unlikely that all the supporters of one party would express a second preference whilst none of the supporters of the two other parties would.
The example does not remain fundamentally different if I merely change it to:
40 A>B>C 35 C>B>A 30 B>A>C
As such, it is irrelevant whether I gave the second or third preferences. The fundamental flaw in IRV still holds.
B is eliminated first, while A wins even though > 60% of the population preferred B over A.
Why this example is not that unrealistic:
49 A>B>C 3 B>A>C 48 C>B>A
Bedfordshire South-West British General Election 2001
Conservative 42.1% Liberal Democrat 14.8% Labour 40.4% Other 2.7%
In every British general election since Feb 1974 you could probably find approximately 100 similar results ( out of 635- 659 constituencies).
I agree, it is realistic, I simply don't recognize the problem here. The unambiguous winner should be B since:
> 50% of the voters wanted B over C > 50% of the voters watned B over A
Of course, 97% of the voters wanted someone else other then B, but this statistic is rather meaningless as those 97% cannot agree on their first place pick. So, what are they to do? They truly have no good option but to attempt to find a compromise and that option was obviously B.
Why should they not be allowed to find a candidate mutually agreeable to everyone since neither of two groups can agree on their top choice and clearly hated the others top choice?
Preventing compromise (I'm headed off the deep end here) is the surest way to civil war (back on land now).
So, again, I ask, why should not the candidate with the broadest support win when there is no clear top choice?
-- == Eric Gorr ========= http://www.ericgorr.net ========= ICQ:9293199 === "Therefore the considerations of the intelligent always include both benefit and harm." - Sun Tzu == Insults, like violence, are the last refuge of the incompetent... === ---- Election-methods mailing list - see http://electorama.com/em for list info
