It is currently Tue Mar 19, 2024 3:59 am

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 117 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next
Author Message
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #21 Posted: Sat Jul 17, 2021 2:54 pm 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Cassandra wrote:
Even assuming that White did NEVER intend playing the double atari of a (to win), it is ABSOLUTELY IMPOSSIBLE to reach that position by alternating play (NO passes in between).


yes I agree but with a good number of white passes it is easy to reach the position with legal moves.

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #22 Posted: Sat Jul 17, 2021 7:31 pm 
Lives in sente

Posts: 757
Liked others: 114
Was liked: 916
Rank: maybe 2d
Cassandra wrote:
You cannot "prove" a supposed weakness of a rule set by using an example that can be reached only by very young kids putting Black and White Smarties onto the go board.


Well... most rulesets for other games have the property that upon reaching a terminal position, the result can be unambiguously and correctly determined by players of even moderate experience (or determined that the game is not legally over yet), for any possible position that could ever be constructed, regardless of how artificial such a position is. ;-)

And this even sometimes has practical consequences, not just theoretical ones. There's a reason there are almost no computer programs that correctly implement Japanese scoring in Go without getting some real-life corner cases wrong, even if they are still very rare. Whereas there are tons of independent and correct implementations of Chess, or Hex, or Xiangqi, or Shogi, or Pente, or Reversi, or Havannah, or any other number of games, such that those implementations don't have any corner cases whatsoever where they fail to apply the rules correctly, not even absurd artificial positions, so long as the positions are still legal.

If a ruleset becomes ill-defined, or defined but so complex as to be impossible to apply, even in a position that only "very young kids" could reach... that actually is a weakness of those rules relative to almost every other abstract strategy game, which do not have such a problem.


This post by lightvector was liked by: Gérard TAILLE
Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #23 Posted: Sat Jul 17, 2021 7:59 pm 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
Dear "legalists",

I am very afraid that you did not really get the point.

The examples given here in this thread lack a decisive amount of internal logic:
-- During their creation process, both players play extremely kiddish. If you don't like this reference to an infantile behaviour, utilise "irreal" for replacement.
-- After that creation process, you ask a professional question about the application of a certain rule set.
-- Before you try to answer this kind of questions, you have thrown your practical knowledge of the game overboard.

Let's forget for a while that Go is a game for two and assume that is was really possible to capture a surplus of 582 Black stones in the last example.
Then we can also assume for sure that BOTH sides played with an attitude to LOSE the game, NOT to win it. But then -- like lightning out of the blue -- that attitude turned into its opposite, and suddenly WINning the game became the issue???

Even if we accept that irreal change of attitude, we must not throw away our practical knowledge of the game!
If the game really was close at the position shown, it would not be finished. Black would simply capture White's stones. As you certainly know, the smallest living group in the centre of the board contains ten stones. We can expect for sure that Black will be able to create more that only two points of additional territory with his ten tenuki stones.
In practice, White will have to sacrifice a lot of stones, in order to get two eyes for her remaining group on the board.

Black will simply win the game by capturing White's eyeless centre group in that examplary position. No matter which ruleset you apply.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #24 Posted: Sat Jul 17, 2021 8:37 pm 
Lives in sente

Posts: 757
Liked others: 114
Was liked: 916
Rank: maybe 2d
For the position given, everyone agrees that the position itself has zero real-world consequence, and it is of course "decisively internally illogical" to care about the ruling of that position, with the given absurd assumptions of how it came to be. You don't need to spend so much effort arguing that, it's already obvious. People already did get the point, they're making a *different* point that you're not addressing. :)

Which is that the mere fact that it even exists at all, is one of the many hints of the broader nature of the ruleset, and it is this broader nature, not that position itself, that does occasionally have some minor consequences - such as the fact that the Japanese rules are only computer-implementable with enormous difficulty, whereas many versions of the Chinese rules are relatively easy to implement flawlessly.

Sometimes, a thing can be itself not relevant or consequential, not even in the slightest, and yet still be one of many small indicators of something *else* that is consequential.


This post by lightvector was liked by: dfan
Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #25 Posted: Sat Jul 17, 2021 9:26 pm 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
Dear lightvector,

I understand your point of view.

But Gérard provided an example for a "real world" discussion.

The fathers of the rules did not have their applicability in the age of AI in mind.
Indeed, the application of territory-style rules (for the scoring process) presupposes a certain amount of knowledge of the game, which naturally is very difficult to mirror with AI.
Area-style rules do not have such prerequisites, and so are naturally easier to implement.

Therefore, if you wanted a discussion concerning the (probably problematic) application of a ruleset in the "real" (i.e. human) world, it would be indispensable to have the example created with Go stones, not with Smarties.

But if you wanted to discuss the (probably problematic) application of a ruleset in the world of AI, the opposite might be more constructive, yielding faster results.


+ + + + + + + + + +

The starting position of Igo Hatsuyôron 120 can be reached by LEGAL moves (NO pass included).

I think that there is a certain probability that I (a poor Kyu player holding Black) would win a game against any AI (a professional Dan player holding White) that has not been specifically trained on that problem.

==> Don't try to apply a ruleset (in the "real" world) on positions that it has not been designed for.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #26 Posted: Sun Jul 18, 2021 12:13 am 
Judan

Posts: 6082
Liked others: 0
Was liked: 786
Rules are supposed to apply to each legal position and it is legal that players make strategic mistakes. Therefore, rules must apply very well to each legal position.


This post by RobertJasiek was liked by: Gérard TAILLE
Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #27 Posted: Sun Jul 18, 2021 12:44 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
RobertJasiek wrote:
Rules are supposed to apply to each legal position and it is legal that players make strategic mistakes. Therefore, rules must apply very well to each legal position.

Robert Jasiek:
"According to my (Western) understanding of the matter, rules are supposed to apply to each legal position and it is legal that players make strategic mistakes. Therefore, rules must apply very well to each legal position."

Nihon Kiin:
"Our rules are supposed to apply to professional play under the responsibility of the Nihon Kiin and the Kansai Kiin. The players are supposed to operate these rules in the spitit of good sense and mutual trust."

In these rules, you will not find anything about "each legal position". "Legality" is assigned to moves only.

+ + + + + + + + + + + + +

In addition, the result of a mutual agreement to use Smarties, instead of Go stones, cannot be called a "series of strategic mistakes".

That line of play -- created with Go stones -- would indeed contain a series of mistakes.
However, if created with Smarties, every single move is played on purpose, and therefore cannot be mistaken. As a matter of course, the aim of the Smarties agreement is not compatible with the aim of the GAME of Go.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #28 Posted: Sun Jul 18, 2021 2:43 am 
Judan

Posts: 6082
Liked others: 0
Was liked: 786
Cassandra wrote:
Robert Jasiek:
"According to my (Western) understanding of the matter, rules are supposed to apply to each legal position and it is legal that players make strategic mistakes. Therefore, rules must apply very well to each legal position."


While I understand what you are trying to say, please cite properly! I have not written "According to my (Western) understanding of the matter". To say that what I have written was according to my (Western) understanding of the matter, you might write that [citation or indirect speech of my writing] was, in your opinion, according to my (Western) understanding of the matter.

Quote:
Nihon Kiin:
"Our rules are supposed to apply to professional play under the responsibility of the Nihon Kiin and the Kansai Kiin. The players are supposed to operate these rules in the spitit of good sense and mutual trust."

In these rules, you will not find anything about "each legal position".


While tournament rules can be written specifically for a particular organisation, such as the Nihon Kiin, rules of play should be written for everybody (in this case, wishing to apply Japanese-style rules) because the game is for everybody.

A central purpose of rules of play is to always specify what is legal play. Failing to do so is a major flaw of a ruleset. Especially one for tournament play.

Thereby, plain application of the rules of play always enables a spirit of good sense and mutual trust without any problem of having to interpret what the latter might mean other than application of the rules of play.

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #29 Posted: Sun Jul 18, 2021 3:16 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
RobertJasiek wrote:
Cassandra wrote:
Robert Jasiek:
"According to my (Western) understanding of the matter, rules are supposed to apply to each legal position and it is legal that players make strategic mistakes. Therefore, rules must apply very well to each legal position."

While I understand what you are trying to say, please cite properly! I have not written "According to my (Western) understanding of the matter". To say that what I have written was according to my (Western) understanding of the matter, you might write that [citation or indirect speech of my writing] was, in your opinion, according to my (Western) understanding of the matter.

You might have overlooked the "proper" citation of yours that started my posting ...

The text in quotation marks is my best guess of what you should have written.

Quote:
Quote:
Nihon Kiin:
"Our rules are supposed to apply to professional play under the responsibility of the Nihon Kiin and the Kansai Kiin. The players are supposed to operate these rules in the spitit of good sense and mutual trust."

In these rules, you will not find anything about "each legal position".

While tournament rules can be written specifically for a particular organisation, such as the Nihon Kiin, rules of play should be written for everybody (in this case, wishing to apply Japanese-style rules) because the game is for everybody.

The text in quotation marks is my best guess of the Nihon Kiin's intention of creating their 1989 Rules.

I am very sure that you already know that the Nihon Kiin 1989 Rules are NO tournament rules. These do not mention anything about "tournaments".

Quote:
A central purpose of rules of play is to always specify what is legal play. Failing to do so is a major flaw of a ruleset. Especially one for tournament play.

I am very sure that you already know which tyes of moves are NOT allowed / forbidden by the Nihon Kiin 1989 Rules. This implies that every other type of move is legal.

As I already stated above, these rules do NOT specify anything about "tournaments".

Quote:
Thereby, plain application of the rules of play always enables a spirit of good sense and mutual trust without any problem of having to interpret what the latter might mean other than application of the rules of play.

Again, this is YOUR opinion.

But it does not match the Nihon Kiin 1989 Rules.
These rules have it the other way round. "A spirit of good sense and mutual trust" is considered an absolutly necessary precondition for the desired application of the rules.

Nowhere is stated that easing that application was one of the aims for creating that ruleset.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #30 Posted: Sun Jul 18, 2021 3:50 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Cassandra wrote:
The starting position of Igo Hatsuyôron 120 can be reached by LEGAL moves (NO pass included).

Cassandra, can you show us how you can reach this position by using Go stones instead of Smarties?

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #31 Posted: Sun Jul 18, 2021 3:51 am 
Judan

Posts: 6082
Liked others: 0
Was liked: 786
As to tournament rules in J1989, see my commentary on them.

You suggest that a spirit of good sense and mutual trust was considered an absolutly necessary precondition for the desired application of the rules.

If that was so, would you say that everybody would know what spirit of good sense and mutual trust should be?

E.g., consider territory surrounded by one player's live stones, of which some are adjacent to dame. You might suggest that a spirit of good sense and mutual trust lets territory be territory. However, J1989, official commentary and application in professional Nihon Kiin games void such territory until the dame are filled. The J1989 spirit of good sense and mutual trust is not absolute but highly ambiguous.

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #32 Posted: Sun Jul 18, 2021 5:21 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
Gérard TAILLE wrote:
Cassandra wrote:
The starting position of Igo Hatsuyôron 120 can be reached by LEGAL moves (NO pass included).

Cassandra, can you show us how you can reach this position by using Go stones instead of Smarties?

I do not remember that I claimed this would be possible.

You simply did not get the point!
Neither standard AI nor standard ruleset are designed to solve ANY legal POSITION that is possible on the Go board.

Has a whole new can of worms (of about the same size as with the ruleset discussions) been opened discussing the problem that standard AI does not play optimal (with the meaning of maximising the final score) endgame sequences? This is just another matter that it has not been designed for.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #33 Posted: Sun Jul 18, 2021 5:41 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
RobertJasiek wrote:
E.g., consider territory surrounded by one player's live stones, of which some are adjacent to dame. You might suggest that a spirit of good sense and mutual trust lets territory be territory. However, J1989, official commentary and application in professional Nihon Kiin games void such territory until the dame are filled. The J1989 spirit of good sense and mutual trust is not absolute but highly ambiguous.

I do not remember that I suggested anything like that.
However, if both players are fine with not counting parts of the board as territory, there is no way to stop them from doing so.

It's the same matter as with the infantile style that is necessary to create a lot of examples that are utilised as examples in the ruleset discussion here:
If one or both players intend to lose the game, there is no way to stop them from doing so.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #34 Posted: Sun Jul 18, 2021 5:47 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
ElomKW wrote:
What all your statements seem to to be implying is that the Japanese rules aren't supposed to be a proper ruleset per say but a gentlehominins agreement.

-- How is "proper ruleset" defined?

-- Who has made that definition?

-- Who is the target audience for the application of this definition?

-- Who decides whether this definition is binding?

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #35 Posted: Sun Jul 18, 2021 6:02 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Cassandra wrote:
Gérard TAILLE wrote:
Cassandra wrote:
The starting position of Igo Hatsuyôron 120 can be reached by LEGAL moves (NO pass included).

Cassandra, can you show us how you can reach this position by using Go stones instead of Smarties?

I do not remember that I claimed this would be possible.

You simply did not get the point!
Neither standard AI nor standard ruleset are designed to solve ANY legal POSITION that is possible on the Go board.

Has a whole new can of worms (of about the same size as with the ruleset discussions) been opened discussing the problem that standard AI does not play optimal (with the meaning of maximising the final score) endgame sequences? This is just another matter that it has not been designed for.


You say that standard ruleset is not designed to solve ANY legal POSITION that is possible on the Go board.
Obviously I am stupid but I do not understand how do you recognise a position for which the standard ruleset is really designed.



This game begins by two "smarties" moves. Do you conclude that the standard ruleset cannot apply to this game (without knowing if the following moves seem correct or not)?

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #36 Posted: Sun Jul 18, 2021 6:04 am 
Judan

Posts: 6082
Liked others: 0
Was liked: 786
Cassandra wrote:
there is no way to stop them from doing so.


A sportsmanship-like tournament rule prevents such collusion of creating an arbitrary result in the interest of, in particular, all other participants of the tournament.

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #37 Posted: Sun Jul 18, 2021 6:24 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
Gérard TAILLE wrote:
This game begins by two "smarties" moves. Do you conclude that the standard ruleset cannot apply to this game (without knowing if the following moves seem correct or not)?

Quite obviously, the matter gets more and more infantile.

So far, your examples had roughly as many empty points on the board as are filled in your newest one.
Run out of arguments???

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #38 Posted: Sun Jul 18, 2021 6:28 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
RobertJasiek wrote:
Cassandra wrote:
there is no way to stop them from doing so.

A sportsmanship-like tournament rule prevents such collusion of creating an arbitrary result in the interest of, in particular, all other participants of the tournament.

Quite obviously, the matter gets more and more infantile.

If I remember correct, the matter so far was related to a ruleset for the GAME.
A regulation in the ruleset for a TOURNAMENT shall be able to do a job that a regulation in the ruleset for a GAME cannot???
Run out of arguments???

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #39 Posted: Sun Jul 18, 2021 7:30 am 
Judan

Posts: 6082
Liked others: 0
Was liked: 786
Cassandra, when you introduce a tournament-rules-related issue, then do not blame discussion for also proceeding with tournament rules.

Top
 Profile  
 
Offline
 Post subject: Re: No result game without loop (in japonese rule) ?
Post #40 Posted: Sun Jul 18, 2021 7:42 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Cassandra wrote:
Gérard TAILLE wrote:
This game begins by two "smarties" moves. Do you conclude that the standard ruleset cannot apply to this game (without knowing if the following moves seem correct or not)?

Quite obviously, the matter gets more and more infantile.

So far, your examples had roughly as many empty points on the board as are filled in your newest one.
Run out of arguments???


Cassandra, I am not arguing, because simply I still do not understand your point. Sorry for that.
What do you mean by saying that standard ruleset is not designed to solve ANY legal POSITION that is possible on the Go board?
IOW, how do you recognize that a position is relevant or not to the standard ruleset? It may be obvious for some positions but what about any other positions like Igo Hatsuyôron 120 and many others?

Be sure I will not try to convince you about the interest of studying irreal positions!
I just try to understand what you consider being a "real" (?) go positions, relevant to the standard ruleset. Nothing else.

Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 117 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group