Re: Opening problems for AI: Problem 32
Posted: Sat Jul 18, 2020 12:10 pm
What about 
Within 1% of the 2d line extension. I figure the difference is very likely to be noise.Gomoto wrote:What about
Life In 19x19
Life in 19x19. Go, Weiqi, Baduk... Thats the life.
https://lifein19x19.com/
Opening problems for AI: Problem 32
Page 2 of 2
Re: Opening problems for AI: Problem 32
Posted: Sat Jul 18, 2020 12:19 pm
Re: Opening problems for AI: Problem 32
Posted: Sat Jul 18, 2020 12:25 pm
This time I went looking for a modern example later in the game, but still no tenuki. Here it is.
GoGoD 1939-04-19, Inoue Ichiro, 4 dan (W) vs. Takagawa Kaku, 4 dan. Oteai.
Mistake. Loses 16% to par. Enclose the top left corner at a.
GoGoD 1939-04-19, Inoue Ichiro, 4 dan (W) vs. Takagawa Kaku, 4 dan. Oteai.
Mistake. Loses 16% to par. Enclose the top left corner at a.Re: Opening problems for AI: Problem 32
Posted: Sat Jul 18, 2020 3:36 pm
How early is early? Earliest decision point with this joseki seems to be move 15. Four contrasting examples in roughly last decade: twoopting for tenuki and two for hanging connection.Still being played? Early in the game?
Hane: in recent times the earliest this was played was move 18 (2005). Cases since then have been rather late in the game (beyond move 60).
If you go back another ten years or so, early decision point combined with early hane or hanging connection seems to be the norm (moves 15 to 23).
I could imagine pro views on each choice haven't really changed, but we are seeing far fewer examples now simply because takamoku has fallen right out of fashion in the AI age: a sharp decline culminating in just 7 examples in 2019 and none so far in 2020.
Re: Opening problems for AI: Problem 32
Posted: Sat Jul 18, 2020 5:32 pm
Early enough so that a normal option, such as playing in an empty corner, approaching a 3-4, 5-4, or 5-3, or making a shimari with one of them, or playing a press against a 3-4 is available. There may be positions where the hane or the hanging connection is preferable to one of those, but I would like to see such a position.John Fairbairn wrote:How early is early?Still being played? Early in the game?
Good point.John Fairbairn wrote:I could imagine pro views on each choice haven't really changed, but we are seeing far fewer examples now simply because takamoku has fallen right out of fashion in the AI age: a sharp decline culminating in just 7 examples in 2019 and none so far in 2020.
I think that there are positions where the 5-4 is playable, but not many where it is preferable to a 4-4 or 3-4.Re: Opening problems for AI: Problem 32
Posted: Sun Jul 19, 2020 12:28 am
Applying this criterion strictly, I see only three recent examples, and it may be significant that the "perpetrator" in each case is Komatsu Hideki (games on 2006-07-06, 2007-08-09 and 2011-11-06).Early enough so that a normal option, such as playing in an empty corner, approaching a 3-4, 5-4, or 5-3, or making a shimari with one of them, or playing a press against a 3-4 is available.
And who does Hideki think he is to be so different? It may also be significant that the first recorded example of this line was by the famous joseki man Komatsu Kaizen!
Re: Opening problems for AI: Problem 32
Posted: Sun Jul 19, 2020 1:30 am
Thanks.John Fairbairn wrote:Applying this criterion strictly, I see only three recent examples, and it may be significant that the "perpetrator" in each case is Komatsu Hideki (games on 2006-07-06, 2007-08-09 and 2011-11-06).Early enough so that a normal option, such as playing in an empty corner, approaching a 3-4, 5-4, or 5-3, or making a shimari with one of them, or playing a press against a 3-4 is available.
And who does Hideki think he is to be so different? It may also be significant that the first recorded example of this line was by the famous joseki man Komatsu Kaizen!
I was aware of the 2007-08-09 game, where Komatsu, vs. Iyama, played tenuki, but it was to approach a 4-4. But then he kept sente and came back and played the hanging connection instead of enclosing a corner. That last decision may be justified by keeping Iyama from expanding his moyo. Let me take a quick look at those games in the Elf GoGoD commentaries. Edit: I'm not finding the 2011/11/06 game in the commentaries.
But that's OK. The other two games were worth the look. 
Re: Opening problems for AI: Problem 32
Posted: Sun Jul 19, 2020 2:48 am
The Komatsu-Iyama game on 2007/08/09
The position after
could be an Only Move problem.
Game continuation.
and 
Within the margin of error.
Minor error by comparison with approaching the top left corner
Within the margin of error.
Minor error. Better to slide at a, expecting to slide to b later. That makes a good formation within Black's sphere of influence.
anticipates AlphaGo.
Mistake. Loses 16½% to the enclosure at a.
replied at b, another mistake, indicating a shared blind spot.
After encloses the top left corner, if expands Black's territorial framework, the slide to 
is fine. This is a matter of judgement, OC.
After, encloses the top right corner. Then starts an interesting reduction of Black's moyo. 
makes a dogleg keima with and 
. 
and 
embrace , and then 
plays a jump attachment. Very interesting reduction, IMO. (Not that it is finished, OC. Play continues.)
The high approach to the top left corner, followed by the typical joseki, is good against White's thickness in the bottom left.
Elf's verdict: Tenuki OK, hanging connection a mistake.
The position after
could be an Only Move problem. and Within the margin of error. Minor error by comparison with approaching the top left corner Within the margin of error. Minor error. Better to slide at a, expecting to slide to b later. That makes a good formation within Black's sphere of influence. anticipates AlphaGo. Mistake. Loses 16½% to the enclosure at a. replied at b, another mistake, indicating a shared blind spot.After
encloses the top left corner, if expands Black's territorial framework, the slide to is fine. This is a matter of judgement, OC. After
, encloses the top right corner. Then starts an interesting reduction of Black's moyo. makes a dogleg keima with and . and embrace , and then plays a jump attachment. Very interesting reduction, IMO. (Not that it is finished, OC. Play continues.)The high approach to the top left corner, followed by the typical joseki, is good against White's thickness in the bottom left.
Elf's verdict: Tenuki OK, hanging connection a mistake.
Re: Opening problems for AI: Problem 32
Posted: Sun Jul 19, 2020 3:34 am
2006/07/06 game. Komatsu Hideki (W) vs. Mimura Tomoyasu
The fuseki is the same as the 2007 game through, the tenuki to approach the bottom right corner.
is within 1% of Elf's top choice at a.
Within the margin of error.
Within the margin of error. After 
Elf's top choice for Black shifts to b, enclosing the top right corner. That may be explained by the fact that limits Black's potential development from a.
Within the margin of error. Elf's top choice for White has shifted to the invasion of the top right corner at c.
at d. Minor error by comparison with a. Elf's top choice has shifted again.
Verdict: Tenuki still OK.
The fuseki is the same as the 2007 game through
, the tenuki to approach the bottom right corner. is within 1% of Elf's top choice at a. Within the margin of error. Within the margin of error. After Elf's top choice for Black shifts to b, enclosing the top right corner. That may be explained by the fact that limits Black's potential development from a. Within the margin of error. Elf's top choice for White has shifted to the invasion of the top right corner at c. at d. Minor error by comparison with a. Elf's top choice has shifted again. Verdict: Tenuki still OK.
Re: Opening problems for AI: Problem 32
Posted: Sat Jan 01, 2022 4:04 pm