Neck and neck: Gu Li vs Lee Sedol jubango – Game 4

Gu Li Lee Sedol MLily Gu Lee Jubango Game 4 t picture

Gu Li 9 dan (left) and Lee Sedol 9 dan at game 4 of their jubango.

Gu Li 9p won game 4 of the MLily Gu vs Lee Jubango, drawing even with Lee Sedol 9p at two all for the match so far.

Game 4 was held on Jeungdo (Jeung Island) in Shinan County, near Lee Sedol’s hometown, and was the first and only game scheduled in Korea.

Before the game, the players paid their respects to the hundreds who died when a ferry tragically capsized on April 16.

The ferry incident occurred near the venue for this match.

Gu Li extends his streak to 4 in a row

Go fans who like to follow the Lee Sedol – Gu Li rivalry will already know that this is Gu’s fourth consecutive victory against Lee in the last two months.

Gu Li Lee Sedol MLily Gu Lee Jubango Game 4 550x351 picture

Do you see the light?” Gu Li (left) seems to have all the answers lately, chalking up four consecutive wins against Lee Sedol.

Lee will be wanting to dismiss these consecutive losses from his mind and focus all his energy on winning the next game, to stop Gu from making off like a runaway train.

However, anyone who’s played Go competitively will know that this is easier said than done.

Lee Sedol MLily Gu Lee Jubango Game 4 picture

After losing his early advantage, Lee Sedol will be keen to put an end to Gu’s winning streak in the next game.

Game 5 will be held in Yunnan Province, China, and is shaping up to be a crucial turning point in the match.

An Younggil’s preliminary comments

Go Game Guru’s An Younggil 8p is in Korea at the moment, and reviewed the game live with other pros. Here are their preliminary comments, which Younggil kindly sent through for GGG readers:

(Note: you can download the game record or scroll down to view the game online and follow along with these comments)

Gu Li started the game with the Micro Chinese Opening.

Lee Sedol Gu Li MLily Gu Lee Jubango Game 4 1 550x367 picture

Gu Li (right) plays his first move against Lee Sedol.

White’s approach from the side, at 8, has been getting more popular recently.

White 10 was unusual. Approaching at 14 is more common in this position, but Lee chose this move instead.

It looks like both players had researched this pattern before the game. The result up to 27 is playable for both.

Black 37 and 39 were good tesuji, and the game was still even up to Black 49.

Black 57 and 59 formed a nice combination, but White 58 and 60 were good responses.

White 66 was very sharp, and the result up to here was favorable for White.

White 72 and 74 were practical moves, and Lee might have thought that the game was good for White at this point.

White 76 was somewhat questionable, and the game became more or less even again here. White should have played at 87 instead.

White 82 was also a mistake. White should still have played at 87. Lee must felt that the game was still alright for him, but Black 83 was sharp, and Black 85 and 87 were very big. Those moves helped Black’s top group and also aimed to cut at M17.

White 88 doesn’t look good either. Black was happy to reinforce the corner territory with Black 89.

Black 113 was a brilliant move, and the game became slightly favorable for Black. The game was still very close, but Black was slightly better after Black 121.

Black 167 was the finishing blow, and the game was practically over here. If white connects at 173, Black will cut at 172, and he’ll be able to gain some more endgame points by squeezing.

Up to Black 179, Black was winning the game by around 2.5 points, so Lee Sedol resigned.

Lee Sedol Gu Li MLily Gu Lee Jubango Game 4 550x368 picture

Gu Li (right) defeated Lee Sedol by resignation in 179 moves.

Nearly 90% chance of 9 games or more

Given that each player has already won two games, and that they need to win six games to win the match, we can do some crude calculations about how many games we can expect to see from here.

We have to assume that, in each game between these two, each player has a 50% chance of winning. In other words, that each game is like a coin toss.

Based on these assumptions, there’s roughly a 12.5% chance that one player will win the next four games and that the match will end after just eight games. This means there’s something like an 87.5% chance that we’ll see nine games or more.

The probability that the match will end at nine games is about 50% and the chance of ten games at this point is about 37.5%.

Of course, this is a gross oversimplification which doesn’t take player psychology, health and other external factors into consideration. Go games aren’t really like coin tosses (which is what makes this crude), but it does provide some approximate numbers and should be reassuring to those of us hoping for nine or ten games.

The MLily Gu vs Lee Jubango

Two of the world’s top Go players, Lee Sedol and Gu Li, will play a jubango throughout 2014, to decide which of them is the stronger player.

A jubango is a 10 game match between two players. The term originates from the Japanese language and has been imported into English language Go parlance. The first player to win six games wins the match.

The official name for this event is the MLily Gu vs Lee Jubango. MLily is a mattress and bedding company that also sponsors the MLily Cup.

Go Game Guru is writing a book about this match and posting news and commentary about each game as it happens.

Game record

Gu Li vs Lee Sedol – Game 4

Download SGF File (Go Game Record)

David Ormerod, with Younggil An and Jingning Xue


About David Ormerod

David likes teaching, learning, playing and writing about the game Go. He's taught hundreds of people to play Go, including many children at schools in Australia. In 2010 David was the Australian representative at the 31st World Amateur Go Championships. He's a 5 dan amateur Go player and is the editor of Go Game Guru. You can find David on Google+ and follow Go Game Guru on Facebook, Google+ or Twitter.


  1. Marek Jasovsky says:

    thanks for game record and some comments by Younggil!

  2. Connection says:

    Ah how nice to see such a game, I am sure I would think that I messed up at some point when a result like that up to O7 would have happened in some of my games, because black got solid points and even sente to make his bottom group with j4 stronger, but when pros play this way it must be my judging about such situations that is wrong, I hope I can improve it. I also struggle to understand move E6, why not play something like H7? I hope someone can explain me the meaning.

    • David Britt says:

      As a 5k player I would also have thought the sequence up to 27 was much better for B. I don’t know if I would say B got sente (27 seems pretty important). In fact to my mind sente seemed like the only real benefit to W. W has no base and B has solid 4th line stones on either side of the W shape, so it’s not like the influence is that great either. The sequence from 70 to 75 only seems to confirm that. I’d be curious to hear Younggil’s or David’s thoughts on how one would make use of that shape.

      On a side note, I’m glad the series is even again.. let’s hope it stays that way till the end!

      • bobiscool says:

        I’m not a stronger player, but one thing to keep in mind always, is that you can’t just view a result “by itself”, in isolation.

        Remember, black started out with a lot more stones in that area. So if both players are very good, black should come out better in that local area, since white might have gained elsewhere.

    • Younggil An says:

      That’s a good question. The result looks better for Black, but his lower side area is still a bit weak even after Black 27. White can also reduce the lower right corner in the endgame (for example, White 156 was sente in the game). In addition, Black spent more moves in this area, so he got a better result. However, White got sente, and played somewhere else, so the result was still even.
      For the questions of the E6, if White pushes at H7, Black would peep at F6 in sente, and D5 is a good follow up. Black can invade the corner afterwards.

      • Connection says:

        Thank you for the explanation Younggil An.

      • David Britt says:

        Thanks for your reply Younggil. I’ve spent a little more time trying to understand this position, since it still seems strange to me. I noticed that black has to play a gote/reinforcing move twice as a result of the W group, first at move 27 and then again at 75. Perhaps giving up sente that second time demonstrates the weakness of the B group at bottom. That allows W to take a huge corner and 5th line territory on the left side. (Or, as you suggest in your comments, would have allowed W to play at 87.)

        This brings up another question.. what about B playing move 27 at J3? That would make the group more solid and would give more ability to play in the bottom left corner, either with the slide or the invasion. The sequence of B C8, W D2 would also be more playable if the left side became interesting.

        • Younggil An says:

          That’s a good idea David to play at J3. The shape is flat, and it looks a bit low, so Gu didn’t play I think. Actually, playing on the 3rd line in this case is conceivable, and it’s a possible idea. However, White can still exploit Black from outside with attaching at J4 etc. Black can’t easily hane, because the position is weak, so it can help White to develop the center.

  3. Zeno van Ditzhuijzen says:

    Connection, I was wondering the same thing (E6 at H7). My best guess is some combination in which black gets to make life/steal points from the corner. I think; h7, f6, f7, d5. Now white has to choose to protect the corner aji or the fatal push inside his teritory. Dillema time.

  4. It’s too bad that I’m too blind to see it myself, but why can’t w play f15 for e15 and then h18 instead of d15?
    If I were b, that would kill me :-).

    • Younggil An says:

      If White tries to capture the top group with H18, Black can cut at M17, and it’s hard for White to deal with. I’ll show you the details in the commentary soon. Thanks. :)

  5. Did black really have to take gote at move 79 to make the E2-F2-D2 exchange? The cut at L6 doesn’t seem that dangerous, is there a placement inside black’s shape I’m not seeing?

  6. David Ongaro says:

    From an endgame standpoint alone S79 (e2) is big. It’s at least 14 points in gote (just count d2, c2, c1, b1, e1, b2 as blacks privilege and analogous for white if he gets to play f2 first). But there are middle even middle game tactics left with the tsuke at c2, so its probably even bigger than 14 points.

    • Younggil An says:

      David Ongaro’s explain is correct. If Black doesn’t play there, White would just hane at F2, B G2, E2, and the corner territory would become large.

  7. David Ongaro says:

    (Sorry the above sequence should read d1, c2, c1, b1, e1, b2. Note that black has a tactic to live in the corner with Ko should white decide to block d1 at c1.)

    @Stefan: It’s not possible to kill black as long as black can m17 in sente – especially if you take your own liberty with the f15, e15 exchange. Remember that there is a cut at k12 so white should be careful not to die himself.

    • Thx, David. Now I see it. without the lib at f15 w cant answer m17 at m19.
      Those pros really know what they’re doing :-)

      “MLily is a mattress and bedding company” – not so hard to guess when you look at the paravent. It’s a little hard on the players. That lady sleeping happily is almost as contagious as somebody yawning.

  8. E2 is indeed bigger than just the (large) endgame value. Black can definitely do something if white plays elsewhere because of the aji of, for example, C3-D3; D5-C5; C4 (mentioned earlier) – black just needs another stone supporting D5 to make this work, and, for example, an attachment at E9 seems therefore to be very difficult to answer. E9-D9; D10 and white will find it hard to defend all round, or get adequate compensation in the centre from an exchange. Hence white responded locally at F8. E2 looks like a very nice move to me. Especially as black went on to win! :-)

  9. Incidentally, Younggil, is E2 a really really obvious move to a pro, or do you have to think about it a bit?

    • Given the endgame follow-ups it’s a move of 15 points with tally 2. See for the corner aji.

    • In such a relatively peaceful position pros are extremely aware of the score. So if E2 is the move to keep territorial balance, it will very probably get played. Otoh, if E2 were not quite sufficient, b would rather start a precarious fight somewhere. Better to be the underdog in a fight, than to trail in a counting game. And if the opponent “blinks” b might get a small profit elsewhere first and then get E2. As amateurs, we have the liberty to say “gee, f18 is big too, and it’s good for my group”. Or “let’s see what a crosscut on the left side gets me”. But pros are restricted by the requirement of perfection. :-)

    • Younggil An says:

      That’s not really obvious Hippo. Actually, Black 79 and 81 weren’t expected. It’s because Black seemed to be a bit behind. However, Gu thought the game was playable, and he chose that easy way.

  10. At L19 there is also quite some amazement with the idea that the result up to 27 is even. Up to amateur 5d people are convinced the three stones at the bottom negate White’s influence and later on the thick group makes small territory in front of it.

    It seems there is a substantial gap in amateur positional judgment to cover here.

    • As a wall w seems to get nothing. Better to think about it as a reducing group. B is pretty low on the right, w captured P7, and this was bs sphere to begin with. If w could also extend from the wall or invade the bottom, it would be a horrible result for b. As it stands, it evens out. It’s true, that latter w got a very confined position here. but b had to invest pretty heavily for that. W could tenuki against O9, and again after N3. I agree with you, that pros are mysteriously strong. But I think most strong amateurs would assess the potion after 28 (C15) as fairly equal.

    • David Ormerod says:

      It’s taking me time to wrap my head around that too. Initially I thought the result seemed better for Black.

      During the game, the pros on Baduk TV also said that the position was even, so I was trying to understand that.

      I told myself that White could still invade around K3, that Black had played more stones in this area, that they’d traded one stone each, that the wall made it hard for Black to invade at 3-3 in the lower left (without being attacked) and that White had sente (though the endgame potential Younggil mentioned above didn’t occur to me at the time). White also has one almost absolute ko threat in the lower right. However, I still felt the position was a bit difficult for White.

      In fact, I expected White to play 16 at M4 or N4 to try to diffuse Black’s power in this part of the board and avoid the direct cut at 19 completely, for the time being.

      Anyway, after thinking about the position some more, I started to see that Black’s territory in the lower right wasn’t that big – in fact maybe it’s even a little over-concentrated. And while Black’s happy to have time to play J4 (and White’s influence is reduced), White’s influence still turns the whole bottom area into a kind of no man’s land, devoid of any significant potential for either player. Black can’t really attack White’s wall either, only harass it a little later on.

      So, maybe the best way to view this result is in terms of potential? White has more potential in the open parts of the board and White’s lower left corner is bigger than it looks, because of White’s lower right influence. What do you think when viewing the game from that perspective Dieter? What do other people think?

      It would be helpful to talk about this more to find an explanation that will make some sense to people who are reading the book.

      • I thought the result up to 27 was a total failure for white when I first saw it. But after reading the comments here I think I am starting to see that it is not bad for white. Black’s terrority is really only on the third line, and Q10 looks inefficient.

        Another way I tried to convince myself was by pretending that 7 was played at 9 and the result followed from black playing a two-space pincer on a normal knight approach by white to the corner (P3 M3 P6 R7), followed by some exchanges that solidify both sides, which seem to favor white since they helped a weak white group get much stronger while black did not really get a whole lot in return. Looking at it from this perspective, I actually feel this is pretty good for white. But I’m not sure if the knight approach would’ve been a good choice for white in this hypothetical opening, so this reasoning is perhaps not quite valid.

      • Connection says:

        All these explanations make very much sense, it is just hard as first, especially as kyu to see it in this light, because just theoretical if such a position would happen in a game and you would try to get help on KGS or somewhere else from a stronger player, I would bet my money on it that you would hear sentences like “How much points did you get?”, “Your influence is nullified by this black group, *insert generic proverb about extending from walls here*”, “Yes, its strong but how will you use it after black J4? Nothing to attack(!) and you can’t build a framework with it either!” At least I had such experiences and of course its nothing bad in itself, because disagreements can often help to shed new light on things, but if you’re weaker and you get teached these things as a new player, that are seen wrong by a majority of much stronger players it will not help your development but rather hinder it at some point and that of other players if you will teach them the same stuff because it makes sense to you, because you have never met anyone stronger that could explain to you that this way of thinking is actually not good because it leads you to misjudge situations like those that happened in this game.

        I actually remembered a time where I was teached something new, that didn’t make that much sense to me as 9k, but I tried anyway to play after it, and early in the games I couldn’t really see how it helped me, but later I could actually see how it helped me to win/play better/make it harder for my opponent, I think now I got rid of a bad habit that day and it was an amazing feeling.

        I also found the topic in “Life in 19×19″ about Game 4, and someone mentions that “(…) By 75 the overconcentration of white stones at the bottom right is painful.” And I agree, the group really looks painful, and in the pre-comments it is mentioned that move w72 and w74 were practical and Lee probably thought the game was going good for white, but I have a hard time to see other ways for white to react and in the topic there are no variations on it, or if there aren’t really any other moves for his group, isn’t it true that Lee messed up and should have reacted to b63 already?

        • Connection says:

          “I actually remembered a time where I was teached something new, that didn’t make that much sense to me as 9k, but I tried anyway to play after it, and early in the games I couldn’t really see how it helped me, but later I could actually see how it helped me to win/play better/make it harder for my opponent, I think now I got rid of a bad habit that day and it was an amazing feeling.”

          With that I wanted to say, that I will hope now not to misjudge similiar situations in my game or get demotivated, and resign to early to desperate measures, or even resign the game, and maybe see how such strong groups will actually aid me in my game or I will just learn how they can if I just stay calm in such games from now on. Maybe it will be again like getting rid of a bad habit

          • Younggil An says:

            That’s nice experience Connection! Thanks for sharing your own thought and experience here. :)

      • Yes, David, I indeed try to look at it from the perspective of potential. I’m also more aware of thickness having multiple purposes, not only “to attack”, but to nullify.

        But still … It’s easy to come up with justifications when a pro has evaluated it as even. It’s another thing to apply such kind of positional judgment in another game. It’s also difficult to interpret “pros like territory” if you think about this as even. It seems they like territory but they like strong groups too.

        • Stefan says:

          Pros are sometimes a little cavalier in their explanations. They know we amateurs can never really grasp the position. So if they have to chose between a correct explanation, and a superficially consistent one, they kinda like to go with consistency. ;-)

        • Younggil An says:

          Yes, “pros like territory” is not easy to interpret, Dieter. Some pros like territory, and don’t care much about his weak groups like Cho Chikun 9p. However, in some point, when there’re two options, pros rather choose territorial one if both are possible, because it’s easier to manage the game afterwards. Anyway, it’s still hard to explain well.

          For Stefan’s comment, it’s yes and no. I don’t try to explain in details every time, but if my student is higher dan level, I try to explain correctly. If the listener is weaker, I try to explain on his level, so that he can understand better. :)

  11. Anonymous says:

    I kind of always trust the pros, when they give an evaluation. Even if i really cannot understand. I take it as something that point to something that i still do not grasp, and that i was not even aware of not grasping. So it is a very good opportunity to progress. I also believe that in some ways, the evaluation (at least in term of winning chances) is not the same according to the level, e.g thickness can be difficult to use, so the game might be even for pros, but, let say, for strong sdk, the one with the territory will have better chances to win, because well, territory is territory^^.
    I know for myself, i would have been desperate with w position after j4. So i’m very happy to have a good opportunity to learn, and i really looking forward for the commentary from An Younggil, which is always very good and enlightening.

  12. Anonymous says:

    As a beginner what would happen if B25 plays at W26 (O7)? It seems that O7 is a ponnuki and thus very good for white?

    • Younggil An says:

      That’s a good question. If Black extends at O7, White will cut at R6, B R5, W R9 to capture the two stones on the right side. That’s why Black reinforced at 25.

  13. Stefan says:

    B O9 at move 63 was a point of beauty. But w R15 was a great tactical reply. Could b have played R16 instead of O9, or am I too much of a tight wad?

    • Younggil An says:

      Black 63 was a good place. White 64 and 66 were a nice combination, but Black took the sente. Playing at R16 instead would be a bit slow.

  14. After black j4, white’s central group is well placed and strong enough to enable white to destroy all black’s futur attempts to make points in the center. This group has a strong influence even if j4 erases white’s potential on the side. So, in the game, white tried to prove that this influence is a good compensation for the points that black gets in the corner and on the bottom side. That’s why Gu Li played j17, to create a potential for black, and that’s why white answered o17 (only move to destroy this potential).

    I prefer black because I think the good moves are easier to find. Lee’s strategy in this game was great :)

  15. Stefan says:

    It is said that Lee Sedol and Gu Li are good friends. And they certainly respect each other greatly. Gu Li, especially, seems to save his best play for when he’s playing Lee Sedol. But now the series has become really tense. I think for them to save their friendship, the will have to suspend it until the end of the match. And probably only a 5-5 result can leave it fully intact.

  16. Anonymous says:

    Hi An Youngil
    I have a question to ask you.
    In ebook Commented games of Lee Seedol Volume 3,game 2(17 LG cup) was commented in this book?
    Thank you very much

    • Younggil An says:

      I don’t think so. Lee Sedol was not on the final of the 17th LG Cup. You can see the final, game 2 of 13th LG Cup instead.

  17. Gogameguru willn’t post commented junbango game 4???

Speak your mind