Class D, third week; insei-level opening theory

I’m writing this post partly as a test to see how big of a burden it is to write an update in the middle of an insei weekend, and partly because I feel inspired after what I learned today, and want to share the source of the inspiration to the world. Most of the game discussion in this blog post is rather higher-level, but most readers starting from strong kyu level players should find it useful. For those raring to know about my performance today, I scored four wins in four games, winning all by resignation.


I see this opening a lot in my insei games. For some reason, very often when the other insei plays white, he’ll answer with facing 3-4 stones to my 3-4 stones as shown above, and when I approach with black 5, they’ll most often answer with the pincer of white 6.

The game I’m discussing in this blog post continued in this fashion. Answering the attachment of black 7 with the second-line hane of white 8 is also fashionable nowadays, although I normally prefer to cut with white 8 at black 9 instead. White 12 is one way to play the corner out. Up to white 20 here, we’re still in the middle of the joseki.

Those readers who have reviewed recent pro games might answer black’s earlier corner attachment with the two-space extension of white 1. The sequence up to 11, shown here, occurred in another game of mine today. This seems quite reasonable for both.

Black might also answer the two-space extension of white 1 with the knight’s move of black 2. Then white might play the kosumi of 3, and up to 11, it seems like some fighting will ensue. I played something similar in one of my insei games a few weeks ago.

The sequence shown here is one of the more classic joseki variations. Black needs the ladder to play the knight’s move of black 4; otherwise, black wouldn’t be able to capture the stone of white 9 in a ladder. If the ladder didn’t work for black, white would play to the right of black 8 instead of white 11 here, and the two black stones would be lost. Up to white 15, the result is called joseki, though I play this very seldom as white. To me, black gets a reasonably strong centre group in sente, while white’s positions are low all around.

Returning back to the main game, it’s not simple how black should fix his shape. The connection of black 1, here, is what most amateur players would come with, but it lacks imagination. Black has to add another move at around black 3; otherwise white would peep at A, forcing black B, and the black group could be under attack in no time. Even with black 3, here, white is easily able to live with something like C, later. The fact that white is able to make the peep of A anytime, forcing a black empty triangle with B, should warn black that something has gone wrong.

Here is how the game continued. Black 21 is an excellent move, forcing white to take an empty triangle instead. Then black 23 continues the idea, sacrificing the black stone at B which wasn’t needed. Black fixes his wall in sente with black 25 and 27, and finally connects with 29, getting an incredibly strong wall. In the game, white went on to create a living group on the lower side with white 30-34, to which I replied with living in the lower left corner with 35.

White 1 here is how the joseki usually concludes. In this game, black would play at around 2, then, starting to create a fine moyo on the lower side, and also killing off the lone white stone. This game might still be even, but I definitely prefer black myself.

In the real game, black 35 was a little bit hasty. We got Kamimura sensei (9 dan professional) to review the game afterwards; his suggestion was that black play as below.

As the white lower-side group is still not alive, a pincer like black 1 here is big. White cannot break free with the usual technique of white 2 and 4; up to black 21 here, white is completely dead.

The next idea for answering black 1 is the bump of white 2. The sequence up to white 10 above is one way the game might turn out. Next, black will probably play at A to force the white two-space extension of B, and then live in the corner with C. If white played C instead of 10, black would play at around 10 himself to create a fine right-side moyo.

The game went on like this. White got to live a bit too easily on the lower side with 38, and though black gets to play the pincer of black 39, the white lower-left group is in no danger, having plenty of opportunities to escape or to form two eyes. Up to 50, the opening is slightly better for white. The game continued peacefully for a while, after which the opponent ended up playing a little bit slowly. Eventually the white stone of 50 got under a serious attack, and black got a decisive advantage from attacking it.

10 thoughts on “Class D, third week; insei-level opening theory”

  1. Thank you for quite an interesting piece of joseki development. In the GoGod database, there’s only one pro game where white pincers at 6. It was Qiu Jun against Zhou Heyang in 2008.

    Cheers, Vesa

  2. Thank you very much for all your interessting blog posts!

    You are a very good writer and every time I read your posts, I want to become stronger – espacially when I see your games here! : D Very creative and efficient play!

    Wish you all the luck and great success with your insei training and please don’t stop writing about it =)

  3. Hi Antti, great blog. When I was in China in the summer of 2009 a professional player told me that the move which is 27 in your game diagram was a little crude (I played there too, as it feels like good timing). He said that one of the reasons for this is that after black plays the move which is 35 in your game black has a back-door connection at the first row, if he has omitted playing 27. He also added that it’s pretty much black’s right to play there and that white gains nothing from pushing there himself. I guess it’s a small detail, but since that is kind of what this post is about I thought you might be interested.

    1. Wow, that’s an interesting detail! Up until yesterday, I had mostly concerned myself with Variation 5, which was how I believed the joseki should be played. Now that we’ve seen that it’s also possible for white to move to live on the lower side instead, it certainly looks like the atari of 27 need not be played immediately. Even without the atari having been played, if white goes to take the corner like in Variation 5, black will get the atari in sente anyway.

      Thanks for pointing this out!

      1. This Joseki is also included in “The 21st Century of Basic Joseki, Volume One”.
        Indeed Black does not play the Atari and White usallly takes the corner whereafter Black plays one point lower than you did.

        Funny though connecting with the empty triangle (your variation 4) is also listed there and it says equal ^^ After White drops down, Black pincers one point under the middle hoshi.

        Then again players like Sakata Eio played “Joseki” which the book calls unfavourable.

  4. While you’ve been winning almost constantly so far, this is the second game you’ve reviewed where you thought your opponent had an advantage coming out of the opening. Do you think the Japanese insei you’re playing are more oriented towards the opening than you are, or Western players in general? Or is that just coincidence.

    1. It might be a coincidence, too, but I guess the insei are relatively stronger in the opening than in other parts of the game. I feel like I’m mostly coming up ahead in middle game fights and yose. In most amateur games, anyway, the game is usually decided in middle game fights.

  5. Always enjoy reading your blog. Thank you!

    Hope you can manage to get to the B class or even higher during your stay in Japan : )

  6. Hi Antti, and thank you for your blog, is awesome :)

    As far as I know after 21B atari W should not connect (very bad shape and the variations you and Charlie have discussed). This happened to me in a tournament and the preocupation because of the underneath connection was like I was lacking a move somewhere, I think this avriation is not possible for W.

    So after B21 W should 24 instead (counteratari), B will extend (not capture), W will connect, B also connects (with atari), W extends in the bottom, B will capture and W will reinforce 14 stone extending it.

    I believe this an accepted variation, W is bigger but B shape has no defects and sente. It would be nice if you tell me your thoughts about this


Leave a Reply

Your email address will not be published. Required fields are marked *

Please confirm you are a human by solving this: