Bki's study journal

[hl782]

I've finally decided on a way to go through the Segoe tesuji dictionary that satisfy me. After my first pass through all the C problems, I'm currently going over it again skipping only the A level ones. I spend 3 minutes on each problem, doing around 10-20 problems per session. I will note some of the problems that I failed so that I can play it over my board later. I do so when :
1 - I completely missed the idea of the problem.
2 - I completely missed one of the possible defense given in the solution.
3 - I was bothered by a possible defense that isn't in the solution and so need more time to figure out.
4 - The problem was just that interesting.

Isn't the Segoe tesuji dictionary too difficult for your current level, even if it is the C-Level problems? I still find them pretty challenging even now, so I've pushed it aside until I become a bit better at reading.

[Bki]

I've finally decided on a way to go through the Segoe tesuji dictionary that satisfy me. After my first pass through all the C problems, I'm currently going over it again skipping only the A level ones. I spend 3 minutes on each problem, doing around 10-20 problems per session. I will note some of the problems that I failed so that I can play it over my board later. I do so when :
1 - I completely missed the idea of the problem.
2 - I completely missed one of the possible defense given in the solution.
3 - I was bothered by a possible defense that isn't in the solution and so need more time to figure out.
4 - The problem was just that interesting.

Isn't the Segoe tesuji dictionary too difficult for your current level, even if it is the C-Level problems? I still find them pretty challenging even now, so I've pushed it aside until I become a bit better at reading.

There are outliers, but generally I think that the C and B problems are right for my level. They're not easy, but interesting and fulfilling and when you fail to solve the problem you still learn something.

[hl782]

Obviously I don't have the right to tell you how you should study, but I feel like looking back at your overall journal history, you've been stuck around the 5-3kyu mark for quite some time - despite having stated that you solved many problems. Perhaps you are overestimating your reading ability or you aren't studying the tsumego properly?

The Segoe dictionary is recommended to start studying around the Dan mark by many for a reason. If you are getting the C problems wrong, then it's likely an indicator that the book may be too difficult for you. Just my 2 cents.

[Bki]

Obviously I don't have the right to tell you how you should study, but I feel like looking back at your overall journal history, you've been stuck around the 5-3kyu mark for quite some time - despite having stated that you solved many problems. Perhaps you are overestimating your reading ability or you aren't studying the tsumego properly?

The Segoe dictionary is recommended to start studying around the Dan mark by many for a reason. If you are getting the C problems wrong, then it's likely an indicator that the book may be too difficult for you. Just my 2 cents.

I mean, I've stopped playing again a few months ago after 4 years of barely playing at all. It's not surprising that I wouldn't have improved during that time, and I feel I've been making consistent improvement since then, so I would say that my current training program seems to work satisfactorily for now...

Edit : also I've seen the Segoe dictionary recommended as a way to get to dan level, rather than to start studying when you're already there.

At this point I usually get the C problem right, and even with the B problems my failure rate isn't discouraging yet.

Anyway I haven't posted a game in the last few days, so here one I just finished :



Would anyone has suggestion for other answers to the far approach? Also I don't like how I was forced by , , but I don't feel like I could afford to ignore them. I think him ignoring white 130 was the reason I won. It destroyed his central territory. Him backing off from the ko at white 180 also certainly helped but I think I was ahead even then (and would probably get compensation from the ko fight).

[Knotwilg]

30 - The strategy based on 3-3 invasions has been popular and successful lately. I’m not sure about the line of invasion. The 4th line is traditionally the line of development, not stability, so I would have played 3rd line here.
46 – forcing with an Atari on Black’s 3rd line stone before extending up would be the common sabaki technique
53 – I didn’t check with Lizzie but my feeling is that Black gets a natural attack on a heavy group, while making territory. Are White’s 4 corners enough? Maybe.
63 – Black should have extended towards the centre stone
71 – and then plays an almost pass move. Whenever you make an empty triangle, think again!
72 – disconnecting Black on a large scale is huge!
74 – simply capture and make your 2 eyes already
90 – your dragon will find a 2nd eye, so play elsewhere: good!
140 – with this sequence, you destroy his major other potential. The endgame must still be played, but you take the lead here.
152 – last comment: beware of disconnecting strong groups. The opponent may ignore that “threat”.

[Uberdude]

Would anyone has suggestion for other answers to the far approach?

Your corner kosumi is fine as a patient and territorial move, but gives him the chance to play a speedy opening so you will likely need some good invasion skills later. Also remember that if he doesn't extend from his approach then you deserve to pincer and attack it later. Other choices would be the knight's move f16 to keep influence/centre access (e16 is a good follow up for him in game), c15, or simply tenuki would display good fighting spirit e.g. make your own big shimari at f4.

Also I don't like how I was forced by , , but I don't feel like I could afford to ignore them.

Answering 31 is acceptable, and something you have to endure for getting the sente invasion and not giving him the eyeshape with playing hane-connect there (you'd already crawled the extra time at n18, if you hadn't should do so before blocking). 34 is ok, but 35 is too kind (not that it means you are loosing, but you fail to punish his premature move). Better is tenuki to help your side stone, if he cuts you r18 atari and then m18 crawl: if he extends then you live with r19 descent, if he ataris the 4 stones you tenuki and hane out at l17 and let him capture them if he wants. The purpose of m18 over immediate r19 live is to avoid his turn at m18 being sente. If you don't like the trade of losing the 4 stones you can crawl at m18 before the tenuki to the side stone, but maybe he will display fighting spirit and attack the side stone and then you wonder do you help it or display your own fighting spirit and follow up at l17.

This 'take both 3-3 and then side hoshi' actually happened in a game between Andrew Kay 4d and Boris Mitrovic 2d at the British championship league a few months ago. Serendipitously it was the game which Antti Tormanen 1p did his recorded lecture for the London Go Centre on and he explains various of these ideas and how to deal with the attachment under the hoshi too:

https://youtu.be/eN-T3keIe4I?t=899

[Bki]

30 - The strategy based on 3-3 invasions has been popular and successful lately. I’m not sure about the line of invasion. The 4th line is traditionally the line of development, not stability, so I would have played 3rd line here.
46 – forcing with an Atari on Black’s 3rd line stone before extending up would be the common sabaki technique
53 – I didn’t check with Lizzie but my feeling is that Black gets a natural attack on a heavy group, while making territory. Are White’s 4 corners enough? Maybe.
63 – Black should have extended towards the centre stone
71 – and then plays an almost pass move. Whenever you make an empty triangle, think again!
72 – disconnecting Black on a large scale is huge!
74 – simply capture and make your 2 eyes already
90 – your dragon will find a 2nd eye, so play elsewhere: good!
140 – with this sequence, you destroy his major other potential. The endgame must still be played, but you take the lead here.
152 – last comment: beware of disconnecting strong groups. The opponent may ignore that “threat”.

30 - My feeling was that the 3rd line was too deep, and it seems too easy for black to get solid territory and influence while forcing white to live small.
46 - You're definitely right, I should have ataried and sacrificed a few stones for some free forcing moves. Some possibilities appear for black to be split up on this side (which reduce his territory by a lot, something that my invasion kind of failed in the game), or for white to get a stronger shape than I ended up having.
53 - He definitely has more solid territory at the moment, but I do have a lots of way to get more on the side, so at this point I think the game is about even.
63 - Certainly.
72 - Actually I think it was a bit thin. If he wedges in I can only give up a stone.
74 - I'm ashamed, I'm not sure what I was thinking in the game.
152 - With H-J13 though I do get some profit if my opponent let me cut. Maybe not enough for now though.

Would anyone has suggestion for other answers to the far approach?

Your corner kosumi is fine as a patient and territorial move, but gives him the chance to play a speedy opening so you will likely need some good invasion skills later. Also remember that if he doesn't extend from his approach then you deserve to pincer and attack it later. Other choices would be the knight's move f16 to keep influence/centre access (e16 is a good follow up for him in game), c15, or simply tenuki would display good fighting spirit e.g. make your own big shimari at f4.

Also I don't like how I was forced by , , but I don't feel like I could afford to ignore them.

Answering 31 is acceptable, and something you have to endure for getting the sente invasion and not giving him the eyeshape with playing hane-connect there (you'd already crawled the extra time at n18, if you hadn't should do so before blocking). 34 is ok, but 35 is too kind (not that it means you are loosing, but you fail to punish his premature move). Better is tenuki to help your side stone, if he cuts you r18 atari and then m18 crawl: if he extends then you live with r19 descent, if he ataris the 4 stones you tenuki and hane out at l17 and let him capture them if he wants. The purpose of m18 over immediate r19 live is to avoid his turn at m18 being sente. If you don't like the trade of losing the 4 stones you can crawl at m18 before the tenuki to the side stone, but maybe he will display fighting spirit and attack the side stone and then you wonder do you help it or display your own fighting spirit and follow up at l17.

This 'take both 3-3 and then side hoshi' actually happened in a game between Andrew Kay 4d and Boris Mitrovic 2d at the British championship league a few months ago. Serendipitously it was the game which Antti Tormanen 1p did his recorded lecture for the London Go Centre on and he explains various of these ideas and how to deal with the attachment under the hoshi too:

https://youtu.be/eN-T3keIe4I?t=899

I had missed that I could get away with ignoring 35 since he has to capture before he can play R19.

That was a nice lecture. Of course the side hoshi invasion is kind of reliant on the opponent letting get away with a sente 3-3 invasion which really shouldn't happen. Though given this maybe I should have played something else instead of the second 3-3? Though attaching to his extension if he get sente to play the side hoshi is another cliche technique in such situation.

[Bill Spight]

Several comments.

[Bki]

Bill : Thanks for the comment. About the wedge, I don't think I thought about it during the game, it did cross my mind while reviewing, but I dismissed it because corner > side. Still, I am not an AI and even if they might prefer the 3-3, the wedge should give me a comfortable game. So maybe I shouldn't strive for perfection when I can't properly do the followup to make it works well.

And actually, now that I look at that position again in Lizzie, while it prefers the 3-3, your wedge is the second most visited move. So clearly I should be more careful about blindly applying dogma, and there's more to the AI era play than invading the 3-3.

Also I like your variation where white let the corner die but steal most of the side and can look forward to an attack on the other wall next.

[Bki]

So for this game I did do the "invade only one 3-3 then wedge" thing and I had a good game going, though some unreasonable fighting spirit on my part led to it being a lot more difficult than it should have been. Oh, and my opponent blundered 35 points for a half point ko at the end, but I was winning anyway so it doesn't matter.

[Bki]

I just played this game that I won on time. Somewhat my opponent managed to lose on time despite the slow standard time control of IGS.



I was in unfamiliar territory in the lower right corner after , but I think I handled it well. I don't play the high approach more but the knight move approach felt hard to play. Attaching to Q3 didn't appeal to me after getting pincered so I played the natural O4. I don't really know the usual joseki, but I feel like there's no way followed by is correct. in particular feels cowardice. After that I felt happy with the result, though I decided to settle my group before enclosing the top right. The crosscut tesuji at (well, if I didn't miss anything and it is actually a tesuji) felt good and let me get sente.

I'm never sure which enclosure is best. I've played the large knight move a lot, and the small knight move enclosure sometime even if the ai don't like it. I've yet to try the two space jump since I've no idea how to handle it well. Anyway, high felt right here since black had a high position overall so I felt like contesting him a bit. Once he approached kick and pincer were natural moves. I was happy to see since I was a bit worried about the inner peep at C16 and this removed it as a concern while still leaving the upper side open. I felt that was irrelevant.

There's lots of point to invade, and felt a good one since I can get free moves against the corner. In the game the move didn't enter my mind, but now that I look at this position might have been better at S13... Yes, looking in more depth it's certainly better since it looks like black corner isn't locally alive anymore. (black S14, white defend at R11, black S18, white hane at S19, black T16, white T18).

Giving up the three stones felt like a good enough trade for the entire right side and not having to worry about the invading group's life anymore, especially since the three stones aren't dead yet. And given that he decided to protect the upper side, I decided to pull them out with . Given his big lump of stone, my instinct was that I could live by threatening it.

White 130 and 132 was because I had misread the "tesuji" of white 134. But then it turned out all right because I realised I could play white 150 and win the capturing race.

Then my opponent tried something against the corner and ran out of time on the last stone before his next 10 minutes of time. I do think I'm winning regardless though.

[Knotwilg]

Instead of giving answers, here are some hints:

40 - I'm not thinking about which enclosure to play. Why?

58 - yeah you solved that one;

74-75 - I agree connecting two groups is better than trying to save those 3 stones. This is already a hint at what White should do next at 76. It's a very interesting position, which I'm going to run through LZ.

[Knotwilg]

Analysis with LZ, focusing on 40 and 78

[/Hide]

[Bki]

: (Before looking at your lz analysis) : It feels like the bottom right might need a move. Black S15 and the peep at O6 are quite annoying. Doing R13 and killing his top right corner (or profiting in sente from him living) is another option.

: Maybe I could cut black's apart. But black can sacrifice three now unimportant stones... So I guess playing elsewhere might be better? The top is open, invade the top?


After looking at the variations : Interesting. In retrospect, yes, I didn't really had sente at the bottom right and simply didn't get punished there. I'm not too worried about black growing the bottom to ridiculous amount because it's still open in several places and white can reduce from the top or from the side, and you can probably exploit the three sacrificed stones to reduce deeper. So the top feel more important.

Also seeing white M17 make my own play for at N15 feel wanting.

[Bki]

An exciting game... Which I could have afforded to make less exciting.



After white 148 I'm easily winning, and I could have let him have the six stones on the left side. But doing so would be too easy on him, there's no way that should die. And well yes, after 162 I can be completely alive in one move. But no. Then after black 165, B10 set up a ko and I can either win it or make an eye and connect. Easy. But no. Then 170 at G10 and the group still live easily (and I can still play B10). But no, sloppiness in l&d (to be fair it's getting somewhat late and I'm tired) means that I play the terrible A10 and I give him sente to enclose me. I was lucky to find the nose play at 204.

Anyway let's talk a bit about the start of the game :

low because I didn't like what followup I had with the high play at D7. Though with him tenukiing I would have felt better having it high... Ah well, it's still a reasonable shape of a double approach in answer to the pincer.

Then in the following fight that spilled over three quarter of the board I didn't feel like letting him get away with anything that felt like ambition on his part, so the guiding principle seems to have been cut first think later. Given that I avoided dying anywhere and then seized the entire right side I feel it went well for me. Well his double hane at was bad after all.