headers
Kerim Aydin | 25 Jul 2012 19:43

DIS: Re: BUS: self-reference is not the point


On Wed, 25 Jul 2012, Sean Hunt wrote:
> Proposal: No Klein Turtles (AI=1)
> {{{
> Amend the rule 'Win by Paradox' by replacing
>       actual or hypothetical, but not arising
>       from that case itself, and not occurring after the initiation of
>       that case
> with
>       actual or hypothetical, but not arising
>       from that case itself, not occurring after the initiation of
>       that case, and not involving self-reference or mutually recursive
>       references.
> }}}

1.  Doesn't everything paradoxical include some degree of self-reference?
Even the first one (was due to retroactivity, but was a retroactive
cancellation of itself)?

2.  I think it *is* the point.  The reason self-reference worked here is
because, in legislating conditional Promises, we didn't add "any condition
that is self-referential simply fails", like we do for conditional votes.
That was legislative error[*].  It's good to point out such legislative error 
through wins (and exploiting legislative mistakes is one whole point of
nomic, right?)

3.  I think instead we should get rid of "hypothetical" win conditions.
Basically, if you can set it up "for real" you should get it, but just
saying "If ABC were true, then it would be undecided" shouldn't be enough.
(Continue reading)

Permalink | Reply |
headers
Sean Hunt | 25 Jul 2012 19:51
Picon
Picon
Favicon

Re: DIS: Re: BUS: self-reference is not the point

On Wed, Jul 25, 2012 at 1:43 PM, Kerim Aydin <kerim <at> u.washington.edu> wrote:
> 1.  Doesn't everything paradoxical include some degree of self-reference?
> Even the first one (was due to retroactivity, but was a retroactive
> cancellation of itself)?

Perhaps explicit self-reference should be needed, but generally, this
isn't true. The intent of a Paradox win is to discover a logical
contradiction inherent in the rules, and I think that simply winning
by discovering a point where someone can introduce arbitrary
self-reference (and by extension, paradox) is not the intent.

> 2.  I think it *is* the point.  The reason self-reference worked here is
> because, in legislating conditional Promises, we didn't add "any condition
> that is self-referential simply fails", like we do for conditional votes.
> That was legislative error[*].  It's good to point out such legislative error
> through wins (and exploiting legislative mistakes is one whole point of
> nomic, right?)

Mentioned above.

> 3.  I think instead we should get rid of "hypothetical" win conditions.
> Basically, if you can set it up "for real" you should get it, but just
> saying "If ABC were true, then it would be undecided" shouldn't be enough.

In support of this as well.

-scshunt
Permalink | Reply |
headers
ais523 | 25 Jul 2012 22:10
Picon
Favicon

Re: DIS: Re: BUS: self-reference is not the point

On Wed, 2012-07-25 at 10:43 -0700, Kerim Aydin wrote:
> 3.  I think instead we should get rid of "hypothetical" win conditions.
> Basically, if you can set it up "for real" you should get it, but just
> saying "If ABC were true, then it would be undecided" shouldn't be enough.

I don't see what this gains. Allowing hypotheticals lets people test
their paradoxes without actually ruining the gamestate, and it should
come to the same thing either way. (There are also potential moral
issues around paradoxically criminal actions.)

--

-- 
ais523
Permalink | Reply |
headers
Kerim Aydin | 25 Jul 2012 22:55

Re: DIS: Re: BUS: self-reference is not the point


On Wed, 25 Jul 2012, ais523 wrote:
> On Wed, 2012-07-25 at 10:43 -0700, Kerim Aydin wrote:
> > 3.  I think instead we should get rid of "hypothetical" win conditions.
> > Basically, if you can set it up "for real" you should get it, but just
> > saying "If ABC were true, then it would be undecided" shouldn't be enough.
> 
> I don't see what this gains. Allowing hypotheticals lets people test
> their paradoxes without actually ruining the gamestate, and it should
> come to the same thing either way. (There are also potential moral
> issues around paradoxically criminal actions.)

Because it's far too easy.

For example, using the very first paradox as an example, it was realized
that by playing a sequence of cards, a paradox would be created.  It 
turned out that at least three people independently discovered it and were 
quietly waiting to have the right cards to set it up, which made one of
the cards desirable for otherwise unexplainable reasons.  It was Fun.

If it had just been: "Hey, if I play XYZ cards it's a paradox, therefore
I win", it's not really fun.

Also, in some cases, the point of illegality is to stop someone from
doing something for a win.  If the threat of illegality prevents 
someone from setting up a hypothetical; again, it's an effective
block.

-G.
(Continue reading)

Permalink | Reply |
headers
ais523 | 25 Jul 2012 23:05
Picon
Favicon

Re: DIS: Re: BUS: self-reference is not the point

On Wed, 2012-07-25 at 13:55 -0700, Kerim Aydin wrote:
> Because it's far too easy.
> 
> For example, using the very first paradox as an example, it was realized
> that by playing a sequence of cards, a paradox would be created.  It 
> turned out that at least three people independently discovered it and were 
> quietly waiting to have the right cards to set it up, which made one of
> the cards desirable for otherwise unexplainable reasons.  It was Fun.
> 
> If it had just been: "Hey, if I play XYZ cards it's a paradox, therefore
> I win", it's not really fun.
> 
> Also, in some cases, the point of illegality is to stop someone from
> doing something for a win.  If the threat of illegality prevents 
> someone from setting up a hypothetical; again, it's an effective
> block.

Oh, I see. I think the best fix would be to only allow hypothetical
actions that the initiator could perform in the current gamestate,
rather than hypothetical actions by anyone else or hypothetical
gamestate.

Perhaps you could word it in the form of promises, i.e. a win by paradox
would be if you had a promise (possibly/probably self-authored) that
would cause a paradox if cashed. That avoids gamestate damage, while
forcing the ability to actually set that damage up.

--

-- 
ais523
(Continue reading)

Permalink | Reply |
headers
comexk | 25 Jul 2012 23:40
Picon

Re: DIS: Re: BUS: self-reference is not the point

I don't think that would help much - as far as I recall, the recent liar's paradox wins all involved actions
anyone could have taken by announcement.

Sent from my iPhone

On Jul 25, 2012, at 2:05 PM, ais523 <callforjudgement <at> yahoo.co.uk> wrote:

> On Wed, 2012-07-25 at 13:55 -0700, Kerim Aydin wrote:
>> Because it's far too easy.
>> 
>> For example, using the very first paradox as an example, it was realized
>> that by playing a sequence of cards, a paradox would be created.  It 
>> turned out that at least three people independently discovered it and were 
>> quietly waiting to have the right cards to set it up, which made one of
>> the cards desirable for otherwise unexplainable reasons.  It was Fun.
>> 
>> If it had just been: "Hey, if I play XYZ cards it's a paradox, therefore
>> I win", it's not really fun.
>> 
>> Also, in some cases, the point of illegality is to stop someone from
>> doing something for a win.  If the threat of illegality prevents 
>> someone from setting up a hypothetical; again, it's an effective
>> block.
> 
> Oh, I see. I think the best fix would be to only allow hypothetical
> actions that the initiator could perform in the current gamestate,
> rather than hypothetical actions by anyone else or hypothetical
> gamestate.
> 
> Perhaps you could word it in the form of promises, i.e. a win by paradox
(Continue reading)

Permalink | Reply |
headers
Kerim Aydin | 26 Jul 2012 00:57

Re: DIS: Re: BUS: self-reference is not the point


omd wrote:
> I don't think that would help much - as far as I recall, the recent liar's 
> paradox wins all involved actions anyone could have taken by announcement.

Well, not *all* paradoxes need be interesting to set up, and it did point 
out the problem with introducing "knowledge" and "truth" into specific crime 
language for ratification.  (besides, I think that judgement was wrong though 
I wasn't paying enough attention that particular week to appeal).

ais523 wrote: 
> > Oh, I see. I think the best fix would be to only allow hypothetical
> > actions that the initiator could perform in the current gamestate,
> > rather than hypothetical actions by anyone else or hypothetical
> > gamestate.

One way might be to make it clear that, if a player cannot immediately and 
directly perform something e CFJs about, then it crosses the line to "overly" 
hypothetical (R591) and becomes IRRELEVANT.

-G.
Permalink | Reply |

Gmane