Life In 19x19
http://www.lifein19x19.com/

Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
http://www.lifein19x19.com/viewtopic.php?f=45&t=9469		 Page 1 of 1
Author:  Cassandra [ Tue Dec 03, 2013 8:50 am ]
Post subject:  Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
Taken from Sekiguchi Harutoshi’s book “囲碁ルールの研究―理論と歴史”

“Igo rūru no kenkyū ― riron to rekishi“ / “Study of Go rules – Theory and History”
ISBN 978-4-286-03142-2
Japan, 2007

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

In the book mentioned above, I found an interesting concept for “Japanese Rules” of the game of Go, which might not be known in the Western World so far.

Perhaps the one or the other of our "rules' experts" might be interested in.
One major issue of this model might be that "permanent stone(s) after capture" are possible only at points that were occupied by (later) captured stone(s) in the starting position.

In the attached PDF, I concentrated on “status confirmation”, because this phase involves the main difference to the “Nihon Kiin Rules” from 1989. Please be so kind to note the paper's disclaimer.

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

If there are any questions, I will try to answer. But please be aware that my knowledge of Japanese is a very rudimental one only.

Attachments:
L&DExamplesNew.pdf [737.59 KiB]
Downloaded 625 times
Author:  RobertJasiek [ Tue Dec 03, 2013 11:07 am ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
Can, according to the book, there be F3 stones that are also S1 or can there be F3 stones that are also S2?

Does Sekiguchi assume that each status confirmation starts with Black?

He seems to have overlooked capturable-2 and the related equivalence to WAGC-alive, i.e., he lacks behind the state of the art a few years. Not to mention Simplified Japanese Rules.
Author:  Cassandra [ Tue Dec 03, 2013 11:43 am ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
RobertJasiek wrote:
Can, according to the book, there be F3 stones that are also S1 or can there be F3 stones that are also S2?

If I understand the Japanese text right, "S" is considered only for stones that can neither reach F1 nor F2.

I think that it is natural to first ask whether a group will be "alive". If it is not, second comes the question whether it might be part of a Seki. I do not see any sense in asking whether "alive" stones might also be "Seki stones".

Quote:
Does Sekiguchi assume that each status confirmation starts with Black?

As far as I have experienced, in Japanese texts on rules the pair "Black" & "White" is widely used as what you will know as pair "player" & "opponent". Just to simplify the text.

Status confirmation works per single bunch of stones at a time. And each confirmation is independent from each other.
As a matter of course, if you took the text literally, and wanted to consider the status of a White bunch of stones, you would have to inter-change the colours in the text.

Quote:
He seems to have overlooked capturable-2 and the related equivalence to WAGC-alive, i.e., he lacks behind the state of the art a few years. Not to mention Simplified Japanese Rules.

If you provided me with a sample position, I could try to apply his method on it, and let you know the result.
Author:  RobertJasiek [ Tue Dec 03, 2013 12:12 pm ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
Example 0000 in

http://home.snafu.de/jasiek/j1989c.html
Author:  Cassandra [ Tue Dec 03, 2013 12:52 pm ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
It seems that your example 0000 is a bit esoteric, isn't it ?

Just add "... by Black stones or a forbidden point." at the and of F3, and everything is fine ;-))
Author:  Bill Spight [ Tue Dec 03, 2013 1:28 pm ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
Cassandra wrote:
It seems that your example 0000 is a bit esoteric, isn't it ?

Just add "... by Black stones or a forbidden point." at the and of F3, and everything is fine ;-))



Sometimes you make things better by making them worse, but this is ridiculous! :shock:
Author:  Cassandra [ Tue Dec 03, 2013 2:12 pm ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
Bill Spight wrote:
Sometimes you make things better by making them worse, but this is ridiculous! :shock:

In principle "to set up a permanent stone after capture" means that you have to get the referenced point under your control.

This can be done either by occupying this point by one of your stones that cannot be captured thereafter.
Or by transferring this point into an eye of a two-eyed group.

The latter is the exceptional case.

You can easily be overdoing things in order to extend rules for every exceptional case. If you restricted yourself on the regular case ("occupy permanently with one of your stones"), it would not be necessary to mention this "with one of your stones" explicitely in the rules.

If you wanted to have the exceptional case included, because otherwise you were convinced that the rules were full of flaws, it would become necessary to include the regular case, too.
Author:  RobertJasiek [ Tue Dec 03, 2013 4:25 pm ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
The single point eyes of a two-eye-formation are NOT an exception, but the whole two-eye-formation (or n-eye-formation) is the core of independent life. The concept of permanent-stone is caused by insufficient reflection of what a two-eye-formation is: it consists of both the two single point eyes and the stones of the adjacent strings. Always. For example 0000 and for all other independent life examples. That J1989 did it wrongly is no justification for repeating the mistake. I needed 10 or 11 years to recognise and avoid the mistake, so I can imagine that others having worked presumably independently can have similar difficulties of avoiding the mistake.
Author:  asura [ Wed Dec 04, 2013 7:15 am ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
Example 0000 [from J2003] may be rare or not so important for reality.
But for creating or testing rule models it is very valuable, because that kind of position is easy to overlook and making a mistake there will be quite bad for a model (from a theoretical point of view).
For me I'd put it in the top ten of positions for rule models :)
Author:  Cassandra [ Wed Dec 04, 2013 7:31 am ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
asura wrote:
Example 0000 [from J2003] may be rare or not so important for reality.
But for creating or testing rule models it is very valuable, because that kind of position is easy to overlook and making a mistake there will be quite bad for a model (from a theoretical point of view).
For me I'd put it in the top ten positions for rule models :)

Yes, indeed.

However, its main (and single relevant) feature is to make evident that there might be cases (may be this single case only), wherein a "forbidden point" must be set equivalent to a "permanent stone", when discussing "status confirmation" of a single stone that has become captured.

After capturing, a single stone cannot become replaced "partially" (for the "permanent" case). Thus, it we wanted this stone to be "alive", the point that has been occupied by this single stone must become part of a group that contains two "forbidden points". That's all.

It is understandable that a single position that deserves at least 10 years of search -- according to Robert -- did not find its way into "commonly" written rules.

As I have already written, the consequence is not necessarily a statement that the ruleset is mistaken, but may be a simple adjustment of the rules' text. Solely a cosmetic issue.
Author:  RobertJasiek [ Wed Dec 04, 2013 7:45 am ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
It was not a matter of 10+ years searching a position, but of 10+ years progressing understanding of Japanese rules. I understood bent-5 before searching a plain ko-less example for capurable-2.
Author:  asura [ Wed Dec 04, 2013 8:06 am ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
Sorry for getting more off topic, but I don't want to make a new thread only for that idea. Btw. I haven't read this model yet but I will soon :)

What about slightly changing the definition for capturable-1 so that the defender needs a permanent stone on local-1 OR an eye on local-1 ?

This would cause that some stones that are original capturable-2 becomes capturable-1. As far I can see this would also mean that all stones which surround territory are either uncapturable or capturable-1 now.
Author:  RobertJasiek [ Wed Dec 04, 2013 11:01 am ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
(Eye = one of (at least) two single empty intersections.)

Your idea is new and deserves testing / study.
Author:  asura [ Thu Dec 05, 2013 9:16 am ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
Cassandra, a gereral question to the model, are the comments inside the text from Thomas Redecker or your's?

My first impression is that the rules look relative close to the verbal japanes rules (or a variation of that). The example_4 demonstrates the / a difficulty to reach always the J1989 result with that rules. But if this result is not (always) required I think this result could be tolerated, because it's a quite strange position. (Although it seems a bit unfair to give everything to black there.)


For me the definitions at the beginning could be a bit clearer (for F3 in relation to S1/2).
In example_10 one variation is not mentioned: What if white captures the single b stone and connects there?





Overall the rules are quite intuitive but in example_24 the scoring feels a bit strange. I have added unlimited ko-threats (for the playing phase) and now white gets an undeserved (imo) point for not connecting the ko. Whether 'A' is a dame or occupied by a stone doesn't matter for it.
Author:  Cassandra [ Thu Dec 05, 2013 9:53 am ]
Post subject:  Re: Taken from Sekiguchi Harutoshi’s “Study of Go rules ..."
asura wrote:
Cassandra, a gereral question to the model, are the comments inside the text from Thomas Redecker or your's?
The comments are from both of us ;-)

Quote:
My first impression is that the rules look relative close to the verbal japanes rules (or a variation of that). The example_4 demonstrates the / a difficulty to reach always the J1989 result with that rules. But if this result is not (always) required I think this result could be tolerated, because it's a quite strange position. (Although it seems a bit unfair to give everything to black there.)

As stated in my paper, the original does not include a special Ko-ban rule for "status confirmation".
This leads to several results that (at first sight -- with the "common" background) look very odd. But they are -- as the results for examples 16 to 18 -- coherent within the given environment. In my opinion, this also gives a valid stimulation for thinking about the pros and cons of a status confirmation that claims stones to be dead that cannot be captured during actual play.

The result for example 4 is a matter of taste, in my opinion.
For me, it has some charme to declare all white stones to be dead, despite the fact that not all of them could be really captured with actual play.
However, I think that this example also highlights the philosophy used for "permanent stones after capture". In principle, it restricts the valid area to stones that the opponent has to really capture in the process of taking another group off the board. Placing any stone "elsewhere" on the board does not affect "status confirmation".


Quote:
For me the definitions at the beginning could be a bit clearer (for F3 in relation to S1/2).

My fault with "translating" / shortening the text.

Quote:
In example_10 one variation is not mentioned: What if white captures the single b stone and connects there?

To be honest, I did not intend to create diagrams for every variation.
But you are right, this is another "capturable-2"-case.
In the book, only both "property / status" diagrams are given, no sequence. So perhaps the author had this implication already in mind ;-)

Quote:
Overall the rules are quite intuitive but in example_24 the scoring feels a bit strange. I have added unlimited ko-threats (for the playing phase) and now white gets an undeserved (imo) point for not connecting the ko. Whether 'A' is a dame or occupied by a stone doesn't matter for it.

"status confirmation" may always be overridden by earlier sequences within "play".

Due to the non-existent special Ko-ban rule for "status confirmation", there are also other positions with the original, where you can find a point of territory at a spot you would usually call "false eye", but with the single stone boardering it never touched during the sequence.
Page 1 of 1 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/