This formation caught my interest in a game between Gu Li and Lee Sedol.

Click Here To Show Diagram Code

[go]$$c Uh... You missed the star point.
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . W . . . . . , . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . W , a . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . W . . . . . , . . . . . , X . . |
$$ | . . . . . . . . . . . . . . X . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

I can't hardly seem to find any information about it. All I got from SL was that it allows the nice cap at a, reducing Black's side while leaving both corners open to invasion on 3-3. I'm surprised to see it played against the top player in the world (still?) after Black made this nice formation on the right.

Gu Li has apparently been playing it for years!

The idea seems the same as the regular San-Ren-Sei, except that if Black approaches the corners from the other side, he's caught in a low rather than high pincer. What are the advantages that make up for the possible capping play?

Discuss.

I feel this is similar to asking, what are the advantages to playing the first diagram as opposed to the second:

Click Here To Show Diagram Code

[go]$$c
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . , . . . . . X . . . |
$$ | . . . . . . . . . . . . . B . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

Click Here To Show Diagram Code

[go]$$c
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . , . . . B . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

It's a different scenario, but it's just a trade off between territory and influence, no?

_________________
be immersed

Assuming it does say really say "nice", etc I would say this illustrates the danger of reading SL (i.e. it is anonymous so you don't know what to trust). In pro play the cap is very, very rare, and SL is ambiguous about leaving "both" corners open. Ultimately it really only leaves one open (i.e. White will get sente to take the other), and so on the principle of virtual territory in large formations (where you map out out an area with large segments A and B on the assumption you will get either A or B and don't care which, while the opponent lives small in the other one) Black would be playing into the hands of Whitebif he invades at 3-3 at the wrong time. There is also the question here of how to relate a cap to Black's formation on the right.

What surprises me a litte is that no pro ever answers the cap with the butting move at the star point, yet this move seems to come instinctively to amateurs. It is interesting to ponder why.



The idea supposedly goes back to Hashimoto Utaro around 1940, but it took off a bit around 2008 and has stayed in the repertoire. I suspect it is not as common as it might be simply because Black doesn't play the appropriate way on the right very often.



I didn't understand this sentence (B/W mixed up?), but taking a stab at it, if Black approaches on the wider side, as surely he must at least consider as the first option, White's standard response (ikken tobi on the other side of the corner) makes the low stone on the centre-side look ideally placed.

The 'nice' is paraphrased (and lazy). But there are indeed dangers to reading SL, although my resources for Go knowledge are limited. Maybe one day SL will grow into a gathering of Go knowledge of Wikipedia-esque quality. We would, of course, need to enforce Wikipedia-like standards.



Off the top of my head, I would think that while it's noble for a superhero to try and protect everyone, it's a sad truth of Go that you can't play on two sides of a move at once, so trying to go for both by butting must seem foolish to them. The idea of giving White the choice of direction to develop from must not seem appealing to them either.



It's very cool to see that this formation has roots dating back so far. Can you comment some more about the appropriate way for Black to play on the right? In this game Lee Sedol was Black, yet he still ended up at a large disadvantage early on. Though this seems to have more to do with Gu Li's extraordinary opening play rather than his formation on the left. Still, I'm not strong enough to rule out its influence.



Yes, I did mix up the colors. I'll edit the original post, thank you.

The most common responses to a corner approach in a regular san-ren-sei formation seem to me to be either the one-space low pincer or the one-space jump. In this case the jump seems to work well in concert with the star-point stone as well, though Black still seems more geared towards the middle of the board. Or rather the entire rest of the board.

A low center stone might look to be a more 'balanced' approach but, the corners being intrinsically vulnerable, White can't expect to make a very large territory. I still feel that I am missing a large piece of the puzzle.

This low middle move instead of making a high san-ren-sei was discussed by Alexander Dinerchtein some years ago. Apparently it was being studied by Korean pros. Sorry I don't have any more precise reference. Anyhow, the low move used to be disparaged as going against the logic of go. The two white corner stones are on the star points, implying an influence-oriented game, but the stone below the star point in the middle of the left side is a territorial move. Dinerchtein said that in response to the capping move the Korean pros felt that White should respond with a two-space jump to the left or right, keeping with the territorial character. In any case I think I can see why Gu Li might not have wanted to play san-ren-sei on the middle star point. That would be very much influence and moyo oriented but Black has a strong position with the low third line stone in the lower right which erases some of the influence of the white san-ren-sei, and Black could respond on the upper side, further erasing white's moyo potential. Black would be at least equal in influence and ahead in territory due to the corner enclosure in the lower right. It does seem to make sense for White to play a territorial move.

I can only give a personal view on this, having no source to quote, but as I said in my first post the weak point may be in saying "the corners" are vulnerable, as opposed to "one of the corners". If, for the sake of disquisition, we assume Black invades one corner now or very soon, it is easy to see, as with all invasions, that White as the encloser should end in sente. With that sente he can add a move that almost secures the other corner, and by now he will have a strong wall in the other corner to help defend it remotely, as well as to give good prospects of future gains in the centre. Putting a value on that package is not easy, but it is rather easy to put a value on Black's package. We can count something like 10-plus for the lone corner stone, 10-minus for the corner-invasion group and 16-points as the average territory value of a low shimari enclosure, so at something like 40 points tops Black is still well shy of the standard target of 60+. Which is worrying enough, but I'd suggest that White's package seems to offer hope of at least a good 40 points already, and with centre-facing strength already he has better prospects for developing further.

Accepting all that, Black would not want to invade soon, and so in a sense neither corner is vulnerable yet. Black has to find a way to play so as to negate White's future prospects (including sente) first and only then invade - a tall order. And during that process, White's coat-hanger formation, being united, seems to offer better immediate development prospects to the sides than Black has. And White also has the powerful ikken tobi from his centre side stone if Black ever falls off the pace.

In brief, a theoretical case can easily be made for White's formation. Whether it's a strong case will depend on actual practice.

When you summarize "White can play the capping play W1 to keep Black's framework under control. The two invasions at b are miai, so it is then harder for Black to take territory, while W1 obstructs very large-scale expansion in the centre." to this "All I got from SL was that it allows the nice cap at a, reducing Black's side while leaving both corners open to invasion on 3-3." leading to a discussion how this is misleading, because you will get only one corner... Well, I guess Charles wrote miai for a reason - but of course two open corners usually means just that as well.

In the game records I have the 3-10 is very rarely played as 5th / 6th move but usually after something else was played on the other side. To find a reason for 3-10 instead of starpoint I would look there. Usually it involves some opposing stones on the adjacent sides which reduce the chances for really large scale moyo building anyway, i.e. precisely the situation where you wouldn't want to play a sanrensei anymore and it has been played by players who played sanrensei, such as Takagawa Kaku, Sakata Eio, even Takemiya Masaki.

I did a pattern search on eidolon, this is the only thing that came up for Gu Li

http://eidogo.com/#1ybps2:0,49

Seems like more of a middle game play.

11 matches just on the pattern above 2011-2012 most by Gu Li. Searching just for White's formation on the left with no Black stone inside the F line, I get 401 matches. With it clear from A to M I get 124 matches for White. From a cursory search, I'm still getting used to Kombilo, I can find examples of it being used against the Orthodox in its various forms, once against the low Chinese and often played not as White's third move but after white has split the side on the right.

I'll mess around with Kombilo and I'll see if it can give my any sensible stats.