[EM] A few words about Approval & Condorcet strategy-need and voting experience
nkklrp at hotmail.com
Sat Apr 7 12:19:21 PDT 2012
A. Answer to the claim that Approval requires strategy and Condorcet doesn't:
1. People keep saying that Approval requires tactical voting, but that Condorcet doesn't. Gibbard and Satterthwaite
would be surprised to hear that.
As I've mentioned, Condorcet fully shares Approval's biggest strategy problem: The co-operation/defection problem.
No attempted improvement on Approval is worthwhile unless it is a method that is at least defection-resistant, as I've defined
that term. A defection-resistant method gets rid of the co-operation/defection problem as it exists in a similar nonresistant method.
I call that the primary C/D problem. There might (usually is) a _secondnary_ C/D problem, but it isn't as bad. Typically, defecting
and causing that problem requires a very counterintuitive and unlikely kind of voting.
Ordinary Approval, and Condorcet, are not defection-resistant.
That C/D problem can be dealt with in Approval, in the ways that I've described, and so I don't consider it a serious problem.
(Though I do claim that it's serious enough to be called Approval's main problem, and something that must be reduced in order
to meaningfully improve on Approval).
Surely the C/D problem can likewise be similarly dealt with in Condorcet. My point is that Condorcet retains Approval's worst
problem, and therefore is not a significant or meaningful improvement over Approval.
2. Condorcet has a serious strategy problem that Approval doesn't have: FBC failure. I've said much about why that's important
for our public political elections, and so I won't repeat that here. Condorcet is unsuitable and unsatisfactory for public elections, due to
3. Let me just add that (I know I've recently said this) ITC avoids both of those criticisms: It passes FBC, and it is defection-resistant.
B. Answer to the Condorcetists' agonizing about Approval stragtegy:
I can't say "I feel your pain", because I must admit that I don't.
Condorcetists and IRVists agonize about the dilemma of "Should I approve my 2nd choice???".
I've recently posted, several times, suggested Approval strategies. Different approaches to Approval strategy, depending on
1) Whether there are unacceptable candidates who could win; and 2) What facts you have information about, or have a feel for.
Apparently the Condorcetists and IRVists haven't read that, or heard about Approval strategies. So let me inform them now:
There are easy and natural ways to choose whether or not to approve your 2nd choice (and any and every other candidate).
Another thing that would help the agonizing Condorcetists and IRVists would be to actually try out Approval, and maybe Condorcet
and IRV too, in actual voting, in staw-polls. We've done that much on EM. It was for that purpose that I proposed a poll, a few months
ago, on EM. You didn't want to try using the methods in actual voting. Therefore, if you're a Condorcetist or an IRVist who wrings his hands
about what it's like to vote in Approval, then I would politely point out to you that you don't know what it's like to vote in Approval.
Or Condorcet or IRV.
If you want to find out what it's like to vote in those methods, you might want to try it.
If you don't want to, that's your choice. But then, don't tell us what it's like to vote in Approval.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Election-Methods