[EM] What happens when Approval doesn't let you vote Favorite>Dem>Repub?

[Juho Laatu]

On 29.5.2012, at 22.48, Michael Ossipoff wrote:

>> On 29.5.2012, at 3.05, Michael Ossipoff wrote:
>> 
>>> And Approval doesn't share Condorcet's favorite-burial incentive
> problem.
>> 
>> All Condorcet methods fail the FBC criterion. But in practice situations
> where the
>> related burial strategy would be easy to use successfully or defensive
> burial
>> would be really needed are rare
> 
> Irrelevant, even if true. I've repeated many times for you my observations
> and conversations with Democrat lesser-evil voters. No, I'm not going to
> repeat it again for you. So since you've forgotten about it, then just take
> my word for it that the available evidence indicates that favorite-burial
> will be rampant in Condorcet.

We seem to disagree on the conclusion side.

> 
>> (possibly not even relevant in typical real life)
> 
> Causing people to favorite-bury makes it relevant.
> 
>> elections). In Approval it is quite common that you have to "betray" your
>> favourite to the level of its competitors (but not below). 
> 
> That's your new definition of "burial" :-)

I gave a definition of "betrayal". The alternative definition is as valid as the other one.

> 
>> Approval may thus well
>> involve more "favourite betraying" (in a wider meaning than what FBC uses)
> 
> In a modified meaning that you've come up with ad hoc, to try to make
> Approval sound as bad as Condorcet.
> 
> 
>> than
>> Condorcet methods.
> 
> In Approval, no one has any reason to not top-rate their favorite(s). That
> isn't true of Condorcet.

True. And in Condorcet you can usually top-rate your top-favourite above the less liked candidates. That isn't true of Approval.

> 
>> 
>>> And Condorcet shares Approval's C/D problem.
>> 
>> Approval is quite clueless when there are more than two potential winners.
> 
> I've just finished explaining to you that the goal of single-winner reform
> assumes that there will be more than two potential winners.

Too bad for Approval.

> 
> Evidently you're too clueless to be aware that I just finished explaining
> that to you.
> 
> You'd need to define "cluelessness" as an attribute of a voting system. 

"Inability of voters to find a reasonable way to vote".

> 
> Ok, that does it. This is my last reply to you, at least on this issue. I'm
> not wasting any more time on you.
> 
> I'll finish this reply, and then that's it.

I'll write the mail that I promised and a summary of my findings.

> 
> To anyone else:
> 
> My subsequent non-reply to Juho doesn't mean that he's said something
> irrefutable. It just means that I finally decided to stop wasting my time by
> repeating answers to his repeated arguments.
> 
>> Voters have to guess which important preference they should indicate and
> which
>> one they should falsify and claim to be a tie. 
> 
> Juho is all confused about what "falsify" means. To falsify a preference is
> to vote a preference that isn't really a preference of yours. Condorcet, but
> not Approval, gives incentive to falsify  one or more preferences.

Approval of course falsifies A>B to A=B. Also here one English word ("falsify") can refer to many kind of falsifications.

> 
> I've repeated for you, many time, some suggestions for voting in Approval.
> Those suggestions included expectation-maximizing strategies. No, I won't
> repeat them again for you.
> 
> 
> In Condorcet we are again talking
>> about cases that may be very marginal. 
> 
> Juho neglected to define marginal. "Marginal" might not mean much, when
> describing a strategy that will be widely used.

"Occurs seldom and/or has only minor impact". My dictionary says "of secondary or minor importance; not central".

> 
> 
> Juho says:
> 
>> In Condorcet sincere ranking (indication
>> of all preferences) is a good strategy.
> 
> Only in 0-info, non-u/a elections. 
> 
> We've already discussed this. If you've forgotten, it isn't my
> responsibility to discuss it again with you.
> 
> 
>>> Therefore your assumption that the Republocrats are all that's viable
>>> won't hold up long in Approval.
>> 
>> I don't assume that. Sooner or later some of the former minor
>> parties/candidates are likely to grow, and there will be more than two
> potential
>> winners. There may however be some tendency to stick to the old major
> parties
>> in the system that you proposed (where Approval is used in single-member
>> districts to elect representative bodies) since those old parties may
> still be the
>> most influential ones in the political system of the U.S. for a long time.
> 
> Dream on, Juho.
> 
> Juho is saying that the Republicans and Democrats will remain influential
> because they'll remain influential.

No. They may quite well fall.

> 
> Juho neglects to explain why or how they'll remain influential when (for
> reasons that I've amply described) more liked platforms and proposals are
> well-known, and the high support numbers for those platforms are well-known.
> 
> 
> Juho is engaging in wishful thinking.
> 
>>> The "problem" that you speak of, in
>>> which people have the preferences Green>Democrat>Republican will
>>> vanish when it becomes obvious that he Republocrats aren't as popular
>>> as the media have been claiming. The Republican threat will no longer
>>> be taken seriously, and the idea of a need to support  the
>>> nearly-identical Democrat, to protect from the Republican, will be seen
> as
>>> hilarious.
>> 
>> In many political systems that use Plurality, IRV or TTR voters are quite
> unaware
>> of the problems of the methods and better alternative methods, and
> therefore
>> they do not care much about the problems. 
> 
> On the contrary, voters in Plurality countries know and care about
> Plurality's favorite-burial need. The care enough about it to always bury
> their favorite.

They may think that this is normal. They may be unaware that there are methods that would be better from their point of view.

Juho


> 
> 
>> I however assume that in Approval
>> the strategic problems with more than two serious potential winners would
> raise
>> some concern. People might be "hilarious" after the Republican is no
> longer a
>> threat
> 
> No no what I said was that the notion of the Republican being a threat will
> be seen as hilarious.
> 
> 
>> , but maybe confused while any of the three might still win.
> 
> Typically, it isn't known who will win. But I've provided suggestions,
> including strategic suggestions, for voting in Approval.
> 
> And, more recently I've specifically looked at strategy considerations with
> Progressives, Dems and Repubs.
> 
> As I said, this is my last reply to Juho on this issue.
> 
> My non-reply won't mean that Juho has said something irrefutable. It will
> merely mean that I'm not longer wasting my time repeating things for him.
> 
> Mike Ossipoff
> 
>