Life In 19x19 http://lifein19x19.com/ |
|
This 'n' that http://lifein19x19.com/viewtopic.php?f=12&t=12327 |
Page 10 of 53 |
Author: | Bill Spight [ Fri Feb 12, 2016 8:57 am ] |
Post subject: | Re: This 'n' that |
Dear Friends, I am going through a difficult time now, which is why I have not been posting much. Thanks to people for their interesting comments. Let me briefly say something about Dieter's suggestion now, and comment more fully later. Dieter is right that is better than "a", Sakata's play, on average. That is: This connection is better than the one in the book. On average it is 2 points better, which means that the book play is a serious flaw. A kyu level mistake, if you will. Kudos to Dieter! However, let us compare the two plays with a difference game. Who wins the difference game, playing first? , capturing , clearly wins the difference game. If White tries to capture the stone, Black throws in at 5 and connects by capturing two White stones. at would be connect-and-die. makes a large capture, and wins. Does this mean that sometimes is correct? |
Author: | Shaddy [ Fri Feb 12, 2016 11:29 pm ] |
Post subject: | Re: This 'n' that |
The result is strange because that's the wrong difference game. This should be the right one: |
Author: | Bill Spight [ Sat Feb 13, 2016 12:49 am ] |
Post subject: | Re: This 'n' that |
Shaddy wrote: The result is strange because that's the wrong difference game. This should be the right one: This is a transposition of the first Black first variation in the difference game above. Yes, Black to play wins the difference game. The question is whether White first also wins. But you are on the right track about whether the above difference game is the right one. |
Author: | Jhyn [ Sat Feb 13, 2016 3:49 am ] |
Post subject: | Re: This 'n' that |
Bill Spight wrote: makes a large capture, and wins. Does this mean that sometimes is correct? I might be wrong about the purpose of a difference game, and this seems like the perfect example to clarify it. There is a time where can be obviously better, and that is as a ko threat. Furthermore your above diagram seems to illustrate this, i.e. has the biggest follow-up after a tenuki. Do you think this could be the explanation? |
Author: | Bill Spight [ Sat Feb 13, 2016 8:53 am ] |
Post subject: | Re: This 'n' that |
Jhyn wrote: Bill Spight wrote: makes a large capture, and wins. Does this mean that sometimes is correct? I might be wrong about the purpose of a difference game, and this seems like the perfect example to clarify it. There is a time where can be obviously better, and that is as a ko threat. Furthermore your above diagram seems to illustrate this, i.e. has the biggest follow-up after a tenuki. Do you think this could be the explanation? Difference games rely upon the independence of the two regions of the board being compared. Kos may destroy that independence. You are right that one line of play may be better than the other because of ko threats, but that is not something that difference games address. If this is the correct difference game to compare the two plays, then sometimes (the book play) will be better than the first line cut, even if there are no kos or ko threats involved. The large follow-up after tenuki is the reason. The caveat about kos means that difference games do not say absolutely that one play is better than another, but they give stronger evidence than the size of plays. |
Author: | Jhyn [ Sat Feb 13, 2016 9:23 am ] |
Post subject: | Re: This 'n' that |
Bill Spight wrote: Difference games rely upon the independence of the two regions of the board being compared. Kos may destroy that independence. You are right that one line of play may be better than the other because of ko threats, but that is not something that difference games address. If this is the correct difference game to compare the two plays, then sometimes (the book play) will be better than the first line cut, even if there are no kos or ko threats involved. The large follow-up after tenuki is the reason. The caveat about kos means that difference games do not say absolutely that one play is better than another, but they give stronger evidence than the size of plays. I understand that ko cannot be taken fully into consideration because of the possible dependence between the two regions. Still, consider that would be sente in normal circumstances: if Black cuts, then in the absence of follow-up, this is a beneficial exchange for Black. If White goes first in the difference game, it means that the move was played as a gote, which I think doesn't make sense outside of an ongoing ko somewhere else on the board. To reiterate my point: I believe that if Sakata's move is better, then this can only happen in the case that some ko is involved. I would love to be proven wrong. |
Author: | Bill Spight [ Sat Feb 13, 2016 11:25 am ] |
Post subject: | Re: This 'n' that |
Jhyn wrote: To reiterate my point: I believe that if Sakata's move is better, then this can only happen in the case that some ko is involved. I would love to be proven wrong. IMO, you are right. But that means that the above difference game is the wrong one to compare the two moves. |
Author: | Bill Spight [ Sun Feb 14, 2016 5:02 pm ] |
Post subject: | Re: This 'n' that |
Shaddy is right that we are comparing the wrong difference game, and jhyn is right that the 3d line cut could be used as a ko threat. Remember that difference games and kos don't mix. So what is the right difference game and why? How can we tell? This is the correct difference game because - is a reverse which we can, and should for the purpose of the difference game, consider as a unit. OC, prevents White from playing there to win the difference game. Let's play the game out. Obviously, Black wins. Black makes mirror go, and gets jigo. This way Black gets jigo, as well. Black wins the difference game, and never loses, so Dieter's play is superior. |
Author: | Bill Spight [ Sun Feb 14, 2016 6:28 pm ] |
Post subject: | Re: This 'n' that |
So how do we know that that - sequence reverses? I confess that I overlooked the reverse at first. I was surprised that White to play could win the difference game, but I am used to being surprised by difference games. Still, it just seemed wrong. What conceivable position of the whole board could there be (with no kos) such that is better than ? This is a reverse because is so good. It is good enough so that the position after is equal to or worse than the original position for White. How can we tell that? With a difference game, of course. In this difference game we are comparing the original position (top) with the position after (bottom). Now all we have to check is whether White to play can win the difference game. If not, then reverses. prevents the underneath connection, but then so does , for jigo. If connects, then so does . We already know that White cannot win this game. So White to play cannot win the difference game, and the - sequence is a reverse. What is the other way to detect the reverse? Look at the first variation of the difference game. was Black's best play if it was her turn in the original position. The exchange - has not eliminated it or made it worse. Black has not lost ground. Let's check another sequence. Conceivably this is another reverse. If so, White to play should not be able to win the difference game below. If at "a", at 8. White wins, so that sequence does not reverse. We might have guessed as much since the first line cut by White eliminated Black's best play in the original position. |
Author: | Bill Spight [ Mon Feb 15, 2016 5:00 am ] |
Post subject: | Re: This 'n' that |
I wrote: Bill Spight wrote: I confess that I overlooked the reverse at first. I was surprised that White to play could win the difference game, but I am used to being surprised by difference games. Still, it just seemed wrong. What conceivable position of the whole board could there be (with no kos) such that is better than ? I misspoke. Here is such a position. Of course. is better than "a", achieving jigo. The point of the exercise, however, is that here is even better than "b". But wait! If is at "a", then the best that White can get is still just jigo. How can you say that is "even better"? Easy. Let's compare the two plays with a difference game. To do that we mirror the whole right side. Now we compare on the left side with on the right side. Well, we have made the left side strict miai, haven't we? So we can ignore it and just look at the right side. It looks familiar, doesn't it? White to play wins, Black to play does not. So is better than . |
Author: | Bill Spight [ Fri Feb 19, 2016 10:52 am ] |
Post subject: | Re: This 'n' that |
Okigo Jizai As some of you probably know, I am fond of Okigo Jizai, a 10 volume set of handicap go openings published in 1824 by Hattori Inshuku, founder of the Hattori School. I have posted material here from it several times. See, for instance, this thread. viewtopic.php?f=15&t=9409 Here are a couple of positions from the book for you to ponder. There is no plainly best play. But I like how Hattori treats them. invades the Mini-Chinese position. (The Mini-Chinese is over 200 years old. ) Where to play next? Hattori says that is good, but Black has to be able to weather White's attack in the top right corner. may be a bit of a surprise. Hattori says that is good, which it plainly is. White then plays inside Black's sphere of influence. How does Black respond? Enjoy! |
Author: | Kirby [ Fri Feb 19, 2016 11:00 am ] |
Post subject: | Re: This 'n' that |
Author: | Shaddy [ Sat Feb 20, 2016 7:44 pm ] |
Post subject: | Re: This 'n' that |
Consider this tewari? |
Author: | Bill Spight [ Sat Feb 20, 2016 9:00 pm ] |
Post subject: | Re: This 'n' that |
Interesting comments, guys. Thanks. I renewed my interest in Okigo Jizai when I ran across another book about openings, both even game and handicap, written in the early 20th century. What struck me was how similar the handicap openings were to most of those that I have seen in more modern books; that is, uninspired, and uninspiring. Black should play conservatively but not too defensively and avoid mistakes. That kind of thing. By contrast, I sense a spirit in Hattori's examples, as though he were asking himself, where would I play? There is a sharpness to the play that I find both refreshing and inspiring. (True, I sometimes think that he is setting up lessons, but that's OK, too. ) In the mini-Chinese example Hattori as Black tenukies. There is a hint of that with ( ), which gives Black a bit of elbow room. Black is outnumbered locally, 6 stones to 4, both White groups are fairly strong, and Black has no base. It is as though, despite taking a handicap, Black were taunting White. C'mon, do your worst. I dare you! White obliges with the jaw jab at 21. seems a bit lenient to me, but it is sente, as Black cannot now omit . Now the Black group has some eye shape. looks funny, as it invites the ideal pincer at 28. However, if White plays on the left side right away Black will stake out a large framework on the top side. prevents that with sente. Then takes the last big point of the opening. Both - and earlier - prevent the opponent from making an ideal position in sente, at the cost of leaving a weak group behind. There is a lesson there, but it involves judgement and the ability to handle the weak group if attacked. When you are giving a handicap, you usually can trust that you can weather any storm. What is surprising is that Hattori had Black play that way, instead of resting on the lead from the handicap. is a leaning attack against the White stones on the top side, but it also strengthens the Black group in the bottom right. Black does not want to get into a fight which puts that group in jeopardy. Black continues with a push through and cut on the top side. White pushes through into the center and then Black continues the leaning attack with . Instead of continuing the attack, Black shifts to the left side and center. This exchange strengthens White and yields territory on the side, but it also strengthens Black and makes a territorial framework in the center. The stones are weakened and have become difficult to save, if not impossible. This is another lesson from Hattori. The White stones in the top are weak, but Black does not just singlemindedly attack them. Black is flexible. You often see attackus interruptus in Hattori's examples. takes kikashi on the left side and then resumes the attack against the White stones on the top side. Note that the Black group in the bottom right, which was a liability, has become an asset. Note also the skillful use of sente, first by Black in the invasion of the mini-Chinese position, then by White on the top side, and then by Black in the exchange on the left side. |
Author: | Bill Spight [ Sun Feb 21, 2016 8:45 pm ] |
Post subject: | Re: This 'n' that |
The three stone game is questionable. True, White has a good play on that point, but makes the Black group on the left side somewhat overconcentrated. Players at that time accepted a bit more overconcentration than they do today, so Hattori probably thought that it was OK. Besides, he is setting up . Hattori praises , which sets up a huge framework. But what about the top right corner? Black cannot secure it in one move, and it is not yet urgent. Students of Hattori will certainly not be following White around the board. is normal vs. , but may be a surprise. However, if White hanes at "a", Black at 17 would allow Black to quickly settle his group. The shoulder blow, , is a standard reduction of the Black framework. and are normal. Reply to Kirby's peep: Black responds in New Fuseki style, with the keimas of and . Hattori was not wed to any dogma. prevents White from playing there, which would make good shape. Note that Black does not reply to , but blocks off the corner and prevents White from making a base on the side. Hattori says that the keima at 40 is good. It is as though Hattori were channeling Takagawa from the future. invades. I like how Hattori has Black handle the invasion. secures the corner and then is a normal tesuji. Black sacrifices to stake out the bottom right side. is the last big play of the fuseki. in quite a nice play, isn't it? In the resulting battle White sacrifices to cut off ( ). But then is a good peep. Hattori leaves things in media res. I think you can see why I like Okigo Jizai. |
Author: | Kirby [ Mon Feb 22, 2016 10:16 am ] |
Post subject: | Re: This 'n' that |
One thing that I have observed in the past year or so is that "rules" don't apply all of the time, especially when weak groups are involved. From an empty board, black's play might seem odd: Moves and , above, seem to be "floating". For example, it's not even clear to me that the territory is solid, as there seems to be aji in black's position: If black wanted to secure territory in this area, the "floating stones" would make more sense if they were lower: Now the position is much more clear. But in the game, there is a weak group involved: The prospects of black making territory in the marked area seem much higher, due to white's weak group. The same aji might exist as on the empty board, but white can't really think about exploiting it, since white's other group is weak. As white saves the group, black will naturally get more than if he played a "normal" move as what came to my mind (ignoring the peep, which may be questionable): |
Author: | skydyr [ Mon Feb 22, 2016 10:41 am ] |
Post subject: | Re: This 'n' that |
Well, even in a position like this but without the weak group, a move like has some merits because it builds on the black group up the left side and on the top. The tewari presented (5 white star point stones) is very different from this, to the point that regardless of what black played on the bottom, it would feel like black has already fallen behind. |
Author: | Kirby [ Mon Feb 22, 2016 10:54 am ] |
Post subject: | Re: This 'n' that |
skydyr wrote: The tewari presented (5 white star point stones) is very different from this, to the point that regardless of what black played on the bottom, it would feel like black has already fallen behind. Sure. The only thing I want to describe is the feeling I have gotten that the "rules" change when weak groups are involved. It's difficult for me to articulate this well, but nonetheless, interesting to me. |
Author: | ez4u [ Mon Feb 22, 2016 4:26 pm ] |
Post subject: | Re: This 'n' that |
Interestingly 18 below is not found in GoGoD as far as I can see. Invariably Black connects with the attachment at 'a' instead. That applies whether you are Onoda Chiyotaro playing Hashimoto Utaro in 1936 or Cho Han-seung playing Gu Li in 2013. |
Author: | Bill Spight [ Tue Feb 23, 2016 11:09 pm ] |
Post subject: | Re: This 'n' that |
Igo Tsumekata (How to play tsumego) is another ancient book I kind of like. I haven't worked through it all yet, though. It is by Kobayashi Tetsujiro, 6 dan, and his son, Kentaro, published in 1901. Link to online copy in the Japanese National Library: http://dl.ndl.go.jp/info:ndljp/pid/861034 Here are a couple of problems, one easy, one not. Enjoy! |
Page 10 of 53 | All times are UTC - 8 hours [ DST ] |
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group http://www.phpbb.com/ |