Life In 19x19 http://www.lifein19x19.com/ |
|
GT territory rule http://www.lifein19x19.com/viewtopic.php?f=45&t=18330 |
Page 2 of 8 |
Author: | jann [ Mon Aug 23, 2021 1:32 am ] |
Post subject: | Re: GT territory rule |
These "have a loop" and "have an advantageous loop" doesn't seem clear enough. Do sequences with double passes count as loop? If no, how to "have advantageous loop" if opponent passes (once) after my pass? If yes, would a sole double ko seki make a loop? Maybe an advantageous loop even, breaking it? |
Author: | Cassandra [ Mon Aug 23, 2021 2:03 am ] |
Post subject: | Re: GT territory rule |
https://lifein19x19.com/viewtopic.php?p=266763#p266763 Gérard TAILLE wrote: variation Needs some editing. is utilised twice. |
Author: | Cassandra [ Mon Aug 23, 2021 3:51 am ] |
Post subject: | Re: GT territory rule |
https://lifein19x19.com/viewtopic.php?p=266754#p266754 Gérard TAILLE wrote:
1)The inside border is made of only black stones Needs some editing. One Black stone is missing. |
Author: | Gérard TAILLE [ Mon Aug 23, 2021 6:57 am ] |
Post subject: | Re: GT territory rule |
jann wrote: These "have a loop" and "have an advantageous loop" doesn't seem clear enough. Do sequences with double passes count as loop? If no, how to "have advantageous loop" if opponent passes (once) after my pass? If yes, would a sole double ko seki make a loop? Maybe an advantageous loop even, breaking it? I like such questions Jann because they adresse directly my ideas rather then a formal text on which I need help to find the best english wording. So let's consider my ideas and I will see later how to improve my formal text, maybe with your help. The only basic question you have to answer is the following : can a player prevents her opponent to make infinite passes? If the answer is YES it looks an advantage for the player doesn't it? Surprisingly a go player is generally able to answer this question very quickly and may even consider such question as obvious Let's take the simple double ko Can white prevents black to make infinite passes? Obviously the answer is NO isn't it? Can black prevents white to make infinite passes? Obviously the answer is still NO? The situation appears symmetrical => no advantageous loop. Can white prevents black to make infinite passes? Obviously the answer is NO? Can black prevents white to make infinite passes? This time the answer is YES, isn't it? The situation is not symmetrical => black has an advantageous loop. Considering your questions the point is to avoid mixing the two questions above. It is true that if black plays a sequence showing infinite passes for black then white would have also infinite passes. But that was not the questions above. Can white prevents black to make infinite passes? Obviously the answer is YES? Can black prevents white to make infinite passes? the answer is still YES? The situation appears symmetrical => no advantageous loop. Additonnal comment on these three examples: In the first one the answers to the two basic questions are NO, NO => generally seki in common GO language or NO TERRITORY in my langage In the second example the answers to the two basic questions are NO, YES => generally alive stones against dead stones in common GO language or TERRITORY in my langage In the third example the answer to the two basic questions are YES, YES => generally seki in common GO language or NO TERRITORY in my langage. BTW in normal play this situation is a NO RESULT game. Let's try to apply this stuff to monshine life: Assume black claims for a black territory for the marked set of locations above then it follows: => it is black territory Now assume black is more greedy and claim a black territory covering all the board. What will happen? white, showing the loop above ask the two basic questions: Can white prevents black to make infinite passes? Obviously the answer is NO? Can black prevents white to make infinite passes? The answer is YES, isn't it? The situation is not symmetrical => black has an advantageous loop. => black is allowed to request "permenantly prohibited" ko. The game continue by and now pass and at the same time black requests the previous white ko capture at becomes a permanently prohibited ko. As a consequence you see white can capture all white stones => all the board is black territory. OC this result is wrong but where is the mistake? In fact white appears to be herself to greedy by trying to save all her stones. The mistaken move is (trying to save white stones at the bottom though they cannot be saved). Instead white should simply pass. That way the answer to the question can black prevents white to make infinite passes? is now NO and there are no advantageous loop! Surely black can capture white stone at the bottom but not the white stones in the double ko => the all board is not black territory. As usual everything must be made during confirmation phase to find the perfect play. |
Author: | Gérard TAILLE [ Mon Aug 23, 2021 7:02 am ] |
Post subject: | Re: GT territory rule |
Cassandra wrote: https://lifein19x19.com/viewtopic.php?p=266754#p266754 Gérard TAILLE wrote:
1)The inside border is made of only black stones Needs some editing. One Black stone is missing. yes Cassandra, thank you. It's done. |
Author: | Gérard TAILLE [ Mon Aug 23, 2021 7:04 am ] |
Post subject: | Re: GT territory rule |
Cassandra wrote: https://lifein19x19.com/viewtopic.php?p=266763#p266763 Gérard TAILLE wrote: variation Needs some editing. is utilised twice. Thank you again Cassandra. It is true. It's done now. |
Author: | Gérard TAILLE [ Mon Aug 23, 2021 7:13 am ] |
Post subject: | Re: GT territory rule |
Cassandra wrote: Gérard TAILLE wrote: Has black a advantageous loop? with black can now pass and whatever the answer of white you can see that black has still a loop and black can also make another pass etc. etc. for ever. OC you can try other variation for white but it is quite obvious that, in any case, black can make an infinite number of pass without losing the fact that black has a loop! As a consequence black is allowed to use permanent prohibited ko and it follows for example: and now with pass black request a permanent prohibited ko for the ko in . Now because white will not be allowed to capture this ko in it is clear that white will be quickly captured. =>all board is black territory. This two-step procedure after having identified a loop as "advantageous" looks too complicated in my eyes. There is a much easier option: BLACK's advantageous loop has a length of a 6, which fulfills the condition x = (2 + 2n) with n = 1, 2, 3, ... Therefore, simply prohibit the terminating move of an uninterrupted cycle (which would be a WHITE ko capture). Is it simply a wording problem or it is a problem on the procedure itself? BTW I did not use the wording "cycle" because for me a cycle is a sequence which is repeat indefinitly without any change. In that sense a cycle is only a particular case of loop. |
Author: | jmeinh [ Mon Aug 23, 2021 8:29 am ] |
Post subject: | Re: GT territory rule |
Interesting project. I'm not sure if "2) the outside border of this set is only made of stones of the opponent or is empty" really does what it's supposed to do (or if I don't quite understand it yet). Is the marked set of locations a black territory? if so, with what score? |
Author: | Cassandra [ Mon Aug 23, 2021 8:30 am ] |
Post subject: | Re: GT territory rule |
Gérard TAILLE wrote: Is it simply a wording problem or it is a problem on the procedure itself? BTW I did not use the wording "cycle" because for me a cycle is a sequence which is repeat indefinitly without any change. In that sense a cycle is only a particular case of loop. In my understanding (visualise a solidly connected triple-ko, for example): A loop, once started, has no end. It's like walking on the circumference of a circle. First walk on the complete circumference of a circle. We get infinitely close to the starting point without reaching it. A tiny additional jump brought us to the initial starting point. It follows the second walk on the complete cirumfence. Ad infinitum ... ############################ Just looked at your posting again. Probably I found the reason for our misunderstanding: Cassandra wrote: BLACK's advantageous loop has a cycle-length of a 6, which fulfills the condition x = (2 + 2n) with n = 1, 2, 3, ... Therefore, simply prohibit the terminating move of an uninterrupted cycle (which would be a WHITE ko capture). This is what I intended ( ) to write. Seems that it was too deep in the night... |
Author: | RobertJasiek [ Mon Aug 23, 2021 9:03 am ] |
Post subject: | Re: GT territory rule |
Gérard TAILLE wrote: Can white prevents black to make infinite passes? Obviously the answer is NO? Can black prevents white to make infinite passes? This time the answer is YES, isn't it? The situation is not symmetrical => black has an advantageous loop. Considering your questions the point is to avoid mixing the two questions above. It is true that if black plays a sequence showing infinite passes for black then white would have also infinite passes. But that was not the questions above. Can white prevents black to make infinite passes? Obviously the answer is YES? Can black prevents white to make infinite passes? the answer is still YES? The situation appears symmetrical => no advantageous loop. I do not understand what you are trying to say. Please explain! |
Author: | Gérard TAILLE [ Mon Aug 23, 2021 9:42 am ] |
Post subject: | Re: GT territory rule |
RobertJasiek wrote: Gérard TAILLE wrote: Can white prevents black to make infinite passes? Obviously the answer is NO? Can black prevents white to make infinite passes? This time the answer is YES, isn't it? The situation is not symmetrical => black has an advantageous loop. Considering your questions the point is to avoid mixing the two questions above. It is true that if black plays a sequence showing infinite passes for black then white would have also infinite passes. But that was not the questions above. Can white prevents black to make infinite passes? Obviously the answer is YES? Can black prevents white to make infinite passes? the answer is still YES? The situation appears symmetrical => no advantageous loop. I do not understand what you are trying to say. Please explain! Robert, I just tried to answer Jann questions (maybe I missed Jann point ). These "have a loop" and "have an advantageous loop" doesn't seem clear enough. Do sequences with double passes count as loop? If no, how to "have advantageous loop" if opponent passes (once) after my pass? If yes, would a sole double ko seki make a loop? Maybe an advantageous loop even, breaking it? I am not quite sure what additional information you want. Maybe I can clarify why I invented this "advantageous loop". When you take a set of positions with potential loop (typically with three kos) the results for these positions given by J89 or J2003 or any traditionnal japonese rules depend clearly of the specific position. The work I have done was to try to find a particularity which can tell me if the result will be territory or a seki. I hoped and I was alomost convinced that such particularity must exist to explain the logic of japonese rules and result expected. The result of my search is : yes this particularity is simply the answer to the two questions: Can white prevents black to make infinite passes? Can black prevents white to make infinite passes? If the answers to these two questions are not the same then, for all examples I studied, the result will be dead versus alive stones (using common GO language) and the answer to the two questions tell us which side is in the best position. If the answers to these two questions are the same then, for all examples I studied, the result will be seki, neither side having an advantage. In order to reach a result as close as possible to the expected result my view was simply to introduce these two questions within confirmation phase. OC the exact wording has to be studied carefully but this is the idea. |
Author: | Cassandra [ Mon Aug 23, 2021 9:43 am ] |
Post subject: | Re: GT territory rule |
Can White prevent Black from making infinite passes? Variation: see main line... ############################# ############################# Can Black prevent White from making infinite passes? are atari is atari. are atari. Ad infinitum... |
Author: | Gérard TAILLE [ Mon Aug 23, 2021 9:51 am ] |
Post subject: | Re: GT territory rule |
Cassandra wrote: Gérard TAILLE wrote: Is it simply a wording problem or it is a problem on the procedure itself? BTW I did not use the wording "cycle" because for me a cycle is a sequence which is repeat indefinitly without any change. In that sense a cycle is only a particular case of loop. In my understanding (visualise a solidly connected triple-ko, for example): A loop, once started, has no end. It's like walking on the circumference of a circle. First walk on the complete circumference of a circle. We get infinitely close to the starting point without reaching it. A tiny additional jump brought us to the initial starting point. It follows the second walk on the complete cirumfence. Ad infinitum ... ############################ Just looked at your posting again. Probably I found the reason for our misunderstanding: Cassandra wrote: BLACK's advantageous loop has a cycle-length of a 6, which fulfills the condition x = (2 + 2n) with n = 1, 2, 3, ... Therefore, simply prohibit the terminating move of an uninterrupted cycle (which would be a WHITE ko capture). This is what I intended ( ) to write. Seems that it was too deep in the night... I understand Cassandra and I agree with you for "simple" 3 kos positions. I just try to be more general, including position with 4 kos in which it is difficult to identify what the length of a cycle could be. In my GT rule I would like to take into all types of loops and not only 3 kos positions. |
Author: | Cassandra [ Mon Aug 23, 2021 9:53 am ] |
Post subject: | Re: GT territory rule |
Can White prevent Black from making infinite passes? Etc. ad infinitum. All moves from on are atari. ############################# ############################# Can Black prevent White from making infinite passes? Etc. ad infinitum. All moves from on are atari. |
Author: | Gérard TAILLE [ Mon Aug 23, 2021 9:56 am ] |
Post subject: | Re: GT territory rule |
Cassandra wrote: Can White prevent Black from making infinite passes? Variation: see main line... ############################# ############################# Can Black prevent White from making infinite passes? are atari is atari. are atari. Ad infinitum... That means that you agree with me? Can white prevents black to make infinite passes? NO Can black prevents white to make infinite passes? YES => black has an advantageous loop |
Author: | Cassandra [ Mon Aug 23, 2021 9:59 am ] |
Post subject: | Re: GT territory rule |
Gérard TAILLE wrote: I understand Cassandra and I agree with you for "simple" 3 kos positions. I just try to be more general, including position with 4 kos in which it is difficult to identify what the length of a cycle could be. In my GT rule I would like to take into all types of loops and not only 3 kos positions. If you had a concrete example with 4 ko available, we could discuss this issue deeper ... |
Author: | Cassandra [ Mon Aug 23, 2021 10:01 am ] |
Post subject: | Re: GT territory rule |
Gérard TAILLE wrote: That means that you agree with me? Can white prevents black to make infinite passes? NO Can black prevents white to make infinite passes? YES => black has an advantageous loop Oh, sorry. I thought the conclusion would be clear, given the pass stones below the diagrams. |
Author: | Gérard TAILLE [ Mon Aug 23, 2021 10:24 am ] |
Post subject: | Re: GT territory rule |
Cassandra wrote: Gérard TAILLE wrote: I understand Cassandra and I agree with you for "simple" 3 kos positions. I just try to be more general, including position with 4 kos in which it is difficult to identify what the length of a cycle could be. In my GT rule I would like to take into all types of loops and not only 3 kos positions. If you had a concrete example with 4 ko available, we could discuss this issue deeper ... Yes OC here is an example: Oops it does not seem to be the best example, I will try to find another. At least you can just try to identify what you mean by length of a cycle. |
Author: | RobertJasiek [ Mon Aug 23, 2021 10:48 am ] |
Post subject: | Re: GT territory rule |
"Can White prevent Black from making infinite passes?" What are infinite passes? After two successive passes, hypothetical play ends. Black always passes and does not care whether his stones are removed therefore White cannot prevent Black from passing. Apparently, you mean something else. Something like my "answer-force". Some dual strategy like "force local-2 permanent removal or an infinite cycle". Just guessing what you might mean. Write down what you mean! Next, you ask "Can Black prevent White from making infinite passes?" I suppose by posing two questions you try to invent status definitions similar to my basic ko type definitions with their typically four questions: http://home.snafu.de/jasiek/ko_types.pdf English spelling: Japanese. |
Author: | Gérard TAILLE [ Mon Aug 23, 2021 2:39 pm ] |
Post subject: | Re: GT territory rule |
RobertJasiek wrote: "Can White prevent Black from making infinite passes?" What are infinite passes? After two successive passes, hypothetical play ends. Black always passes and does not care whether his stones are removed therefore White cannot prevent Black from passing. Apparently, you mean something else. Something like my "answer-force". Some dual strategy like "force local-2 permanent removal or an infinite cycle". Just guessing what you might mean. Write down what you mean! Next, you ask "Can Black prevent White from making infinite passes?" I suppose by posing two questions you try to invent status definitions similar to my basic ko type definitions with their typically four questions: http://home.snafu.de/jasiek/ko_types.pdf English spelling: Japanese. Look at https://lifein19x19.com/viewtopic.php?p=266766#p266766 where I have analysed this position for the first time. At the very beginning of this post I clarified the objective of the corresponding hypothetical play (I am not sure it is good idea to introduce the wording "hypothetical play" in my rule but at least I understand what it means here in your post): Is all the board black territory? IOW, because there are no problem with the borders, the question becomes "is black able to build a two eye formation covering all the board?" Keeping this in mind you know if you can afford to pass: if by passing you cannot not reach your ojective then you simply cannot pass. Remember how works the confirmation phase. It is a loop with three points 1) A player claim that a set of location is her territory 2) Borders are verified 3) if borders are OK then the only purpose of the analyse is to know whether or not the player is able to build a two eye formation covering all the set of locations. When you say "after two successive passes, hypothetical play ends" it is OC a wording remark for my proposal text. I am sure you have understood my idea : because we have to care about ko ban, obviously I have to add that the "hypothetical play" ends after three passes (remember I have no ko-pass nor pass-for-ko in my rule) |
Page 2 of 8 | All times are UTC - 8 hours [ DST ] |
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group http://www.phpbb.com/ |