[EM] "Strategy with RCV" on reddit

[Kristofer Munsterhjelm]

On 12/05/2021 01.39, Rob Lanphier wrote:
> Hi folks,
> 
> Interesting proposed strategy for RCV posted on reddit here:
> https://www.reddit.com/r/EndFPTP/comments/n7wqxa/strategy_with_rcv/

As I understand it, the strategy is that the Republicans support a
further-left party (the Greens) so as to engineer a center squeeze
against the Democrats.

But given that the republicans' sincere ranking presumably is R>D>G, it
doesn't seem like they gain anything by doing so, since getting the
Greens elected instead of the Democrats would be a worse outcome.

Maybe it could be used to discredit IRV and return to FPTP. That could
be an interesting perspective: any voting method where enough voters can
be convinced to vote against their honest preferences, and which then
results in a seemingly incomprehensible or bad outcome, would be
vulnerable to such an attack.

For instance, an Approval election could be attacked by splitting a
party into a better and worse party, and then through duplicitous
"advice", engineering a Burr failure.

However, there's a kind of tension. If the bad outcome is obviously bad,
it's going to be hard to get the voters to go along with the unwitting
sabotage. On the other hand, if the bad outcome is not obviously bad,
the voters might not get sufficiently angry to repeal the method.

The concept of an attack on the legitimacy of the method itself is an
interesting one, though. The more advanced early EM criteria (SFC, SDSC,
immunity to majority complaints, etc.) might help defend against such
attacks.

> Those of y'all that have reddit accounts should just post your
> response there rather than replying here.

I used to be on Reddit, but I didn't like what it was turning me into,
so I've taken a break :-)

-km