Go Commentary: Gu Li vs Lee Sedol – Jubango – Game 8

This is the eighth and final game of the MLily Gu vs Lee Jubango.


Gu Li 9 dan (left, at board) and Lee Sedol 9 dan review the final game of their 10 game match.

This game was played on September 28, 2014, in Chongqing, China – which is Gu Li’s hometown.

The score going into this game was 5-2 in Lee Sedol’s favor which meant that this game was a kadoban for Gu and might be the last of the match.

In game 7, Gu had a nice opening, and Gu managed the game very smoothly until the middle game. However, Gu played a careless move, and Lee reversed the game with a sudden attack.

After Gu lost game 7, the atmosphere of the match changed, and a lot of Chinese journalists and Go fans lost interest in the Jubango. That’s because even if Gu won all three of the remaining games, the best result he could achieve was a tie.

We’re writing a book about this match

This commentary, and others, will form the basis for our Go book about Lee Sedol and Gu Li’s jubango. The actual book will contain a more extensive commentary of this game, but you can regard what you see below as a draft (learn more).

Please help us to make our first Go book as good as possible. There are several ways you can help us to improve the commentary below:

  1. Ask questions about the game – if anything is unclear, please let us know so we can explain it better!
  2. Point out any mistakes, even minor typos – our first draft is below. Because this is going to be a book, even small mistakes need to be fixed.
  3. Tell your friends and ask them to help too.

The rules of the game

The time limit for these games is 3 hours and 55 minutes, with 1 minute x 5 times byo-yomi. It’s traditional to subtract 5 minutes from the 4 hour total, because of the 5 x 1 minute periods.

There’s no lunch break scheduled for these games, but food is provided and the players are free to get up and eat whenever they want, throughout the game.

Anyway, let’s have a look at game 8 of the MLily Gu vs Lee Jubango.

Commented game record

Gu Li vs Lee Sedol – Game 8


Download SGF File (Go Game Record)


Related Articles

About Younggil An

Younggil is an 8 dan professional Go player with the Korean Baduk Association. He qualified as a professional in 1997 and won an award for winning 18 consecutive professional matches the following year. After completing compulsory military service, Younggil left Korea in 2008, to teach and promote the game Go overseas. Younggil now lives in Sydney, Australia, and is one of the founders of Go Game Guru. On Friday evenings, Younggil is usually at the Sydney Go Club, where he gives weekly lessons and plays simultaneous games.

You can follow Go Game Guru on Facebook, Twitter, Google+ and Youtube.


  1. Hi Younggil. Is R5 a possible move for 25?

    • Younggil An says:

      Hi Wasuji!

      Yes, it is, but then White will play R6, B R4, W Q7, and Black’s right side stones will get weaker.

  2. Jolyon Kay says:

    I am a relative beginner, and would appreciate fuller comments on why moves are good.

    • Flagellator1974 says:

      “I am a relative beginner, and would appreciate fuller comments (in that book) on why moves are good.”
      I agree.

    • It is certanly not possible to explain all the basic stuff in every game review, it’ll take a whole book. Also the topic – a pro game – is quite advanced itself

      • Younggil An says:

        Yes, right. It’s not possible to explain everything, but we tried to explain as much as possible. 🙂

  3. Lucretiu Calota says:

    Wonderful comments, they make me appreciate the game even more a second time. Their power of reading ahead is amazing, it inspires me to exercise a little more myself. Thank you.

    • Younggil An says:

      I also felt their power and profound reading while I was studying this game carefully. 🙂

  4. ar 2 move 36 “and it will be of greatly assist White in settling his group”.

  5. var 2 move 36 hat was… 🙂

    “Move 78 is worth more than 20 points” – I believe you, but I had to think about it. I hadn’t really thought to quantify this move before your comment, which was good. I think explaining this would be a help to anyone up to (western grade) low dan amateur level, if it is not too off-subject. Just a thought.

    • Younggil An says:

      Yes, I’ll explain more about the move 78 at the book. It’s not too off-subject, and it’ll be helpful to know how to count this sort of big move.

  6. Detailed variations to explain move 84 are really great. As was the comment on the size of move 72 – easily underestimated (I had thought it was about 18).

  7. Move 248 “However, Lee still kept his cool and didn’t in panic, even though the game took a wrong turn.”

    “didn’t in panic” typo also “the game took a wrong turn” is a bit awkward I think.

  8. Last comment reverance spelling (reverence).

  9. Thank you very much for the commentary. It’s as enlightening as always !

    Also congratulations for the improvements to the viewer. If I remember correctly all the most urgent feats have been added. Nevertheless I still have concerns with the full screen mode. I’m using a 5″ 16:9 screen (Nexus 5) and the text box is really small, less than a fifth of the screen height and it’s really difficult to read An Younggil’s comments.

    • Younggil An says:

      Thanks Tim for your feedback, and I’m sorry to hear that the text box is too small for you. It’s bigger than before on my screen.

  10. Great commentary, it made me really appreciate the game and the motives behind many moves. Modern professional games seem to me often just not as beautiful, more messy and fighting than shape oriented than older games, so such commentary helps me to enjoy them as well. Thanks!

    • Younggil An says:

      You’re right Martin. The modern games are more difficult to understand with the intense fighting and trades. You might feel that those games are more exciting and spectacular to watch and follow. 🙂

  11. Thanks for this and all your commentaries, looking forward to the book!

  12. Thomas Woods says:

    I love playing over the games you comment on, you do an excellent job! I cannot wait for the book to come out. Please put me down for a copy.
    Thank you


    • Younggil An says:

      Thanks for your cheering comments Uberdude and Thomas. We’ll try to do our best to make a nice one. 🙂

  13. In the second variation for move 222, after M9 K9, can white play K8? Would this lead to a similar situation as the next variation, with miai at G10 and M11?

    Thanks. Looking forward to the book.

    • Younggil An says:

      That’s a good idea Uisek.

      If White does so, Black will at N10, W J14, B H9, W H13, B O5, W O7, B P7, W N6, B M7, W N7, B G8 and it’ll be one step late ko for Black. It’s a long sequence, and it’s also possible for both, but Lee might have thought the actual game was safer than this variation.

  14. Donald Moorer says:

    The comment on move 8, as it stands, seems to me valid, but black 7 (f17) was never attacked. Was it too slow, was it properly defended by black 9 (k17) (invitation of invasion of white h17) or where there different tactical/strategical goals to be met?

    Are there alternatives for move 78? If 78 wouldn’t be played at j3, do you think it can be marked as losing move?

    (Is it possible to order the comments in scales of ‘education’-levels? (x to 5 kyu, 4kyu to 1kyu, 1dan to 4dan, 5dan to pro-level)

    • David Ormerod says:

      Thanks for your feedback Donald,

      Regarding move 8, we’re going to talk about that more in the book. Briefly though, Black’s response at K17 does serve to defend and Black would have been more likely to play elsewhere if White’s response had been low.

      At 78, White C11 is another eye popping move, but then Black would be able to play at H3, which was kind of miai.

      The value of the two moves can be judged not just in terms of territory, but in terms of the relative strength of groups. White’s bottom left group isn’t that strong yet with the weakness around B2, whereas White’s top left group is completely settled (with practically no weaknesses). That makes 78 in the game more valuable to White.

      Black’s connection at C11 (79) was also very big, but Lee’s next moves in the game showed us that it wasn’t as big as it looked, because he was able to strong arm Black anyway (up to 88). That sequence was really interesting and instructive.

      Whether the difference between the two moves is big enough for it to be the ‘losing move’ is something Younggil will have to answer. My judgment isn’t that good 🙂 (and it also depends on whether White loses the game after playing that way).

      In terms of ordering the comments and labeling them in terms of level, that’s possible in theory. However, the labels would be highly subjective and I’m not sure how useful they’d be as a result. It’s very hard for me to feel the difference between 5k, 4k and 1k, and it’s even harder for Younggil. We’d probably need a small committee of Go players at different levels to help in figuring that out and even then I imagine it would still be highly subjective. How do you think you would use those sorts of labels to help you improve at Go? Do you have something in mind?

  15. Robert Gilman says:

    I’ve come to prefer the format of Ander’s Kierulf’s “Smart Go Books” which provides live diagrams in which you can try your own variations. For myself, I’d love to see you work with him to make your book available in this format. If that doesn’t work out, my next preference would be to have the final commentary available as .sgf files. For me this is a much more convenient way to study the game and variations than setting it up on a regular board.

    Very glad to see that you have taken on this project.

    • David Ormerod says:

      Thanks Robert. Anders talked to us quite early on about making a SmartGo Books version of this book and we agreed. There’s some possibility that it will be ready before the printed book, but I can’t promise anything right now, because we have to discuss it more amongst ouselves (and with Anders).

  16. Flagellator1974 says:

    Move 248: “The game had already become complicated again, and it was because of White’s overplay at A (White 214).”
    That “A” doesn’t marked at H8.
    And thanks for the great commentary!

  17. For w 20, I was surprised by the kick. I was expecting just a jump to f14. For my level it would be helpful to understand better the considerations behind the w 18 attack at c10. But maybe that’s too straightforward for a book about such high level games…

    • David Ormerod says:

      Good questions Vlad 🙂

      The kick at 20 is mostly to make it harder for Black to slide to B14 and make eyespace.

      The pincer at C10 is also an extension from White’s lower left corner enclosure. White wants to make his moyo in the bottom left as wide as possible to put more pressure on Black to invade and fight. Since Black already has one weak group on the left, enticing him to invade is a good strategy. There are other reasons, such as whether it’s easy for Black to counter-attack and I’ll talk to Younggil about putting some variations in the book for you.

      As a rule of thumb, you can extend up to five spaces (like White does here) while still making it hard for Black to invade. That’s because if Black invades around C8 later, White’s lower left star point will already be ‘pincering’ it (from a certain perspective).

      If the game was slightly different and Black had played K17 at K16 earlier, White might think about exchanging G15 for H16 before kicking. That would make it harder for Black to almost connect/enclose with F15 later. Here’s a game where that happened, though the opening is different: https://gogameguru.com/go-commentary-dang-yifei-baek-hongseok-4th-bc-card-cup-final-game-4/

  18. Thanks David. I’ve got this sense that the kick is usually bad and it’s kind of made me reluctant to almost ever consider it; so it’s helpful to get a bit more of an idea of when it’s the right timing for it…

  19. Hi,
    I am not sure if I’m right, but I read that if white plays 202 at q1 then black would play s3.
    It’s pretty complicated but am I right?
    Thank you for this awesome commentary!

    • Younggil An says:

      Yes, you’re right.

      If White plays at Q1 for 202, Black will hane at S3 and White would be in trouble. It’s because S5 and Q2 are miai for Black next.

  20. Maybe I’m not counting this right, but to me move 131 looks like 11 points reverse sente. If White plays here, he will get c18, c19, d19, e18 & captures 1 stone. If Black plays here, he will get f19, g19, g18, g17, h19 & h18.

    • Younggil An says:

      Thanks for your opinion, but it’s hard to follow the sequence you showed.

      Black 131 makes five points difference for both Black and White, so it would be 10 points worth, but it can be changed depends on the surroundings.

      • Oh sorry, that wasn’t supposed to be a sequence of moves. I was just listing the points on the board that go to either side if one or the other plays first. In terms of move sequences it could look like this:

        If Black plays first, the sequence will be like in the actual game. If White plays first, he will get f18 in Sente and Black defends like you showed in the alternative to move 131. Later, White will get g19 in Sente and Black falls back on the second line to avoid Ko. Eventually, Black gets h19 in Sente (White connects at f19). Maybe that is not the correct sequence. But if the sequence is correct, there are 11 points at stake – 4 on White’s side, 6 on Black’s side plus 1 prisoner.

        • Younggil An says:

          Thanks for your explanation.

          Actually, it’s not that important if it’s 10 or 11. I just wanted to show that this sort of endgame is bigger than it looks, because many kyu players don’t know how big the endgame moves are.

          If you learn how to count the value of this sort of endgame moves and try to count by yourself, I’ll be happy enough. 🙂

  21. Actually, the exact count is 10 points: With black H14 on the board (the alternate move), if white captures here, then later after after white H19 black J19 white F19 black can play away at G14 to protect the cut + G15, so black loses nothing as he gains more from G14. Therefore we can assume black will probably get H19 in sente, so it is 10 points…

  22. I haven’t commented here before, but I wanted to say that I was reading your commentaries of the Jubango as they came out, and they were really enjoyable. I like watching pro games, but I’m only 1D, so for some moves I can tell that they are deep and complicated, but I can’t understand them. It was very nice to read your explanations of these moves, and I’m looking forward to the book!

    • Younggil An says:

      Thanks Peter for your kind comment. I’m very happy to hear that the commentaries were enjoyable. 🙂

  23. I thought the when white lived easily in the upper right corner, while black hadn’t captured the white stones, the game turned bad. What about the crude R16 around that time? It is sente against the corner, and takes away white’s method of linking up starting with R16 by white? Could this have helped?

    • Younggil An says:

      That’s a good idea to play at R16 first before White connected underneath. However, if Black plays there, White wouldn’t answer and play elsewhere. Black will atari at T16 next, White will respond at T17 to make a ko. If Black loses the ko, White can still save his side group, so it’s not a good deal for Black.

      That’s why Gu didn’t play there earlier, and Lee eventually came back to connect his side stones.

  24. Anonymous says:

    How do you see the different variations?

    • Younggil An says:

      You can click the red numbers.

      1 is played in the actual game, but 2 or 3 leads the following variations.