Talk:IRV Prime: Difference between revisions
no edit summary
No edit summary |
No edit summary |
||
Line 87:
* For N>=2, a '''does not win''' in IRV-Prime (& thus the 2nd premise, that a wins, doesn't hold)
--[[User:Marcosb|Marcosb]] ([[User talk:Marcosb|talk]]) 00:57, 5 August 2021 (UTC)
So continuing my offline conversations, P4 demonstrates how IRV-Prime fails later-no-harm.
If we change cab voters to cba:
abc 3N + 2 acb 2N
bca 3N bac 2N + 1
cba 3N cba 2N
N - 1:
a: 5N + 2
b: 5N + 1
c: 5N
N (aka b vs a):
b: 10N + 1
a: 5N + 2
N-Prime (aka b vs c):
b: 8N + 3
c: 8N
b wins, which means later-no-harm fails.
Dang it! Thought for sure it satisfied it, but it does not.
Thanks everyone for the discussion & for helping out!
--[[User:Marcosb|Marcosb]] ([[User talk:Marcosb|talk]]) 03:50, 5 August 2021 (UTC)
I've transposed the examples on the page you listed (elections 3-5) [[https://github.com/marcosb/vse-sim/blob/master/methods.py to code]], but not sure what I'm looking for:
| |||