Life In 19x19 http://lifein19x19.com/ |
|
Japonese counting http://lifein19x19.com/viewtopic.php?f=45&t=18300 |
Page 1 of 20 |
Author: | Gérard TAILLE [ Fri Aug 06, 2021 3:54 am ] |
Post subject: | Japonese counting |
Black to play in japonese rules. How do you finish the game to get the best result? |
Author: | Mike Novack [ Mon Aug 09, 2021 5:12 am ] |
Post subject: | Re: Japonese counting |
I suspect the lack of response is because not understanding what is meant by "best result" in a situation like this. Maybe we just aren't seeing why all the outside black stones won;t die. |
Author: | RobertJasiek [ Mon Aug 09, 2021 5:50 am ] |
Post subject: | Re: Japonese counting |
Speaking for myself, I have not responded because, by experience, writing down a careful rules application for this particular position requires ca. 3 to 4 hours. (The term "counting" means the mechanical rearrangement or other physical procedure to count the score. The term "scoring" means the system or method by which the rules define what the score is for the game end situation determining the winner. The thread title means "Japanese scoring" while the counting procedure is immaterial for the OP's question. In ordinary, non-go English, the words counting and scoring are to some extent interchangeable. In go terminology since 1993, however, they have become terms with different meanings.) |
Author: | Gérard TAILLE [ Mon Aug 09, 2021 6:18 am ] |
Post subject: | Re: Japonese counting |
Mike Novack wrote: I suspect the lack of response is because not understanding what is meant by "best result" in a situation like this. Maybe we just aren't seeing why all the outside black stones won;t die. You are right Mike. I think I failed to explain what was the problem in realtion with the rules. Surely Robert is right concerning the wordings "counting" or "scoring". I am not quite fluent in english but I do my best. Let me propose a far simplier (and more efficient?) position: Black to play. Obviously cannot save her stones on the top. My feeling is that black must play ![]() ![]() and white wins by 11 points. But I suspect it is only the feeling of an old player using J49 rule. What about J89 rule? I am not sure but now the best sequence seems: ![]() ![]() ![]() and white wins by 10 points. I am not sure of my rule interpretation. Thank you for an expert able to clarify the situation in J49 and in J89 contexts? What is the point? After ![]() ![]() |
Author: | RobertJasiek [ Mon Aug 09, 2021 8:16 am ] |
Post subject: | Re: Japonese counting |
Black 1 at A7 is another start to be considered. J49? Serious? |
Author: | Gérard TAILLE [ Mon Aug 09, 2021 9:40 am ] |
Post subject: | Re: Japonese counting |
RobertJasiek wrote: Black 1 at A7 is another start to be considered. J49? Serious? OK Robert, let's begin by considering only J89. You propose black 1 at A7. What score do you expect to reach with this move? A white win by 10 points or 11 points or ? It looks to me that this move loses one point but I may be wrong OC. Can you clarify how you finish the game after this move? IOW when will begin the confirmation phase? Surely white expects |
Author: | RobertJasiek [ Mon Aug 09, 2021 10:00 am ] |
Post subject: | Re: Japonese counting |
I have not studied the sequences but you just may not ignore A7 in your study. If it should be futile, you must prove it. Making assumptions is not applying the rules. |
Author: | Gérard TAILLE [ Mon Aug 09, 2021 10:31 am ] |
Post subject: | Re: Japonese counting |
RobertJasiek wrote: I have not studied the sequences but you just may not ignore A7 in your study. If it should be futile, you must prove it. Making assumptions is not applying the rules. Yes Robert. In my post https://lifein19x19.com/viewtopic.php?p=266383#p266383 I try to explain why this move black in A7 loses one point due to a cycle with one more black prisoner. OC we must apply the rules and it is really my question with my example. I suggest finishing the game by: ![]() ![]() ![]() and white wins by 10 points. but because I am not completly sure of my understanding of the rule I ask to the experts if this sequence and this count is really the best result for both. You see the curiosity here? In normal play white do not need to add a move to kill black stones, but, due to the confirmation phase, white must play three moves before the end of the game in order to declare the black stones really dead ![]() |
Author: | Mike Novack [ Mon Aug 09, 2021 11:59 am ] |
Post subject: | Re: Japonese counting |
First of all, I object to the "simplification". Totally different problem. In the first example, under consideration is whether black can prevent the large white group from becoming unconditionally alive, because if so, it's all over(white can fill the upper left ko(s) and black has no ko threats after the one to live on the top is used. |
Author: | Gérard TAILLE [ Mon Aug 09, 2021 2:51 pm ] |
Post subject: | Re: Japonese counting |
Mike Novack wrote: First of all, I object to the "simplification". Totally different problem. In the first example, under consideration is whether black can prevent the large white group from becoming unconditionally alive, because if so, it's all over(white can fill the upper left ko(s) and black has no ko threats after the one to live on the top is used. It is not clear to me what you mean here by "unconditionally alive". this wording make sense if it exists a rather unknown environment typically with potential ko threats. Here all the board is known and we know it does not exist any ko threat. In my first example the black group at the top is dead. The problem is not to prove black stones are dead but rather to find the sequence which gives the best result. As I said before my example is not that efficient because after: I am worried by the dame at "a". Because of that I have to improve the position. The following one is far better: What is the best sequence for both? If I am not wrong there are a strong relationship with the following position I proposed: |
Author: | jann [ Mon Aug 09, 2021 9:46 pm ] |
Post subject: | Re: Japonese counting |
I think in the simpler position B don't want to pass since then W can pass too and have no problems (B dead). B needs to throw in once to force W to capture in the main game (costing him points). |
Author: | kvasir [ Mon Aug 09, 2021 9:57 pm ] |
Post subject: | Re: Japonese counting |
I thought this was a question if white needs to play zero, one or two moves on the inside. For example something like this: Or if white just passes if black can force white to play the last move (and possibly force white to play the last move twice in exchange for black passes). For example: ...and this position is starting to get similar to the second position by Gérard. Possibly the players fills the liberties directly or cycle the ko while filling the liberties, and white always refuses to play extra moves on the inside. As far as I can see this is an interesting ko study that doesn't necessarily have anything to do with specific rule differences, in the endgame white may get off easily if there is no penalty for finishing the ko but similar situations happen in middle game kos were white may find it difficult to finish the ko immediately. Basically, it appears to be a ko situation in which white can clearly win a capturing race by making an eye, but if white refuses to finish the ko he can lose the advantage because the ko can potentially become direct for black to capture. |
Author: | Cassandra [ Mon Aug 09, 2021 10:46 pm ] |
Post subject: | Re: Japonese counting |
Just returned from walking the labrador. His name is Honte ![]() So ... ![]() ![]() ![]() @ Gérard: White should NOT pass her first move in the Nihon Kiin 1989 Rules status confirmation! ![]() ![]() ![]() |
Author: | Gérard TAILLE [ Tue Aug 10, 2021 2:50 am ] |
Post subject: | Re: Japonese counting |
Kvasir, your analysis is exactly what I had in mind ![]() kvasir wrote: I thought this was a question if white needs to play zero, one or two moves on the inside. For example something like this: That's true. When I built the problem I just forget to take into account the dame ![]() kvasir wrote: That is the sequence I expected ![]() In order to avoid the unexpected sequence mentionned at the beginning of this post I proposed a slight change in the initial position. My modified position is the following: Now the expected sequence is and we reach a position identical to the simplication I proposed. I will try to analyse this simpler position in an other post, and please try to take into account only my modified position. |
Author: | Gérard TAILLE [ Tue Aug 10, 2021 3:39 am ] |
Post subject: | Re: Japonese counting |
jann wrote: I think in the simpler position B don't want to pass since then W can pass too and have no problems (B dead). B needs to throw in once to force W to capture in the main game (costing him points). OK Jann, let's see what happens if both players pass in the position above and here is really the rule interpretation issue ![]() I imagine the following dialog: White : your black stones at the top are dead black : no, if I you do not add at least one move, they are living. IOW if you want to kill them you have to resume the game and add at least one move. white : no I do not want to add moves (because I will lose points). black : OK in that case we enter the confirmation phase. Because you refuse to add a move that means you consider black stones are dead even if it is black to play right? White : yes Black : in that case I start the confirmation phase and you will try to kill me, correct? White : yes, let's go and the cycle will go for ever. Does that prove that white cannot kill black stones? That is the point. |
Author: | RobertJasiek [ Tue Aug 10, 2021 3:42 am ] |
Post subject: | Re: Japonese counting |
Analysis so far misses the following: - One sequence per status might not be enough. - Play before successive passes can result in different scoring positions, of which each must be analysed by hypothetical play to then evaluate strategy before successive passes. - Different Japanese rulesets require different applications. 3-4 hours analysis time is the lower bound. Spending a few minutes on random thoughts does not do it. |
Author: | Cassandra [ Tue Aug 10, 2021 4:33 am ] |
Post subject: | Re: Japonese counting |
Gérard TAILLE wrote: black : OK in that case we enter the confirmation phase. Because you refuse to add a move that means you consider black stones are dead even if it is black to play right? White : yes Black : in that case I start the confirmation phase and you will try to kill me, correct? White : yes, let's go As I already stated above, the text in red is mistaken, at least if you want to apply Nihon Kiin 1989 Rules. The STATUS CONFIRMATION of each and every single group of stones on the board starts with the FINAL POSITION of "PLAY". It does NOT matter at all which side put the last stone onto the board. There is NO conjunction between PLAY and STATUS CONFIRMATION. "... cannot be captured by the opponent, ..." implies that White starts the status confirmation for a Black group. If Black ever wanted to add a stone before, he would have had to do it during "PLAY". + + + + + + + + + + + For comparison: If STATUS CONFIRMATION is entered with the central point of a genuine Nakade shape unoccupied (both players had a blind spot here), the status of the relevant group is "dead", NOT "alive". |
Author: | jann [ Tue Aug 10, 2021 4:51 am ] |
Post subject: | Re: Japonese counting |
I also think J89 analysis starts with attacker. |
Author: | Cassandra [ Tue Aug 10, 2021 5:22 am ] |
Post subject: | Re: Japonese counting |
Gérard TAILLE wrote: I will try to analyse this simpler position in an other post, and please try to take into account only my modified position. + + + + + + + + + + ![]() + + + + + + + + + + + + + + + + + + + + All of Black's stones in question are dead. _________________________________ EDITED |
Author: | jann [ Tue Aug 10, 2021 5:49 am ] |
Post subject: | Re: Japonese counting |
In confirmation cannot retake ko without passing for it. |
Page 1 of 20 | All times are UTC - 8 hours [ DST ] |
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group http://www.phpbb.com/ |