I have been looking at §28 of 38 Basic Joseki, which deals with a double keima kakari against a stone on hoshi. I realise that this book is old and restricts itself somewhat, but I am taking it as a way of getting a grounding in joseki.

Most of the variants treated start with tsuke-nobi to pressure a squeezed stone:

$$ ---------------------
$$ - . . . . . . . . . .
$$ - . . . . . . . . . .
$$ - . . a x . W . . . .
$$ - . . b X . . . . . X
$$ - . . d w e . . . . .
$$ - . . O 1 3 z . . . .
$$ - . . v 2 c y . . . .
$$ - . . . . . . . . . .

Other variants are given, with reasons for choosing them, but I cannot formulate a strategic reason for Black to choose tsuke-nobi because White can choose so many different responses to it with different results. Of course Black could play it from habit or just hoping for a certain response, but that seems less satisfactory.

The variants considered

  • Tsuke-nobi: 1-3 as shown.
    • W4@a
      • B5@x — White: side + corner; Black: large top, can swallow marked stone, sente.
      • B5@d — White: corner + small top, sente; Black: side high and strong.
    • W4@b — White: side, option on corner; Black: big top, sente.
    • W4@c, B5@y
      • W6@zFight White: top weak, side medium, sente; Black: corner + cutting stones.
      • W6@aTricky White: side weak, small corner, sente; Black: strong but little territory.
    • W4@d — White: weakish top + side, sente; Black: decent corner, connected to centre.
  • Tsuke-osae: 1-2 as shown, 3@dPeaceful White: see variants; Black: decent corner.
    • W probably connects at v — White: moderately strong on both edges, option to reduce corner; Black: connected to centre, sente.
    • W tries to enclose B (W4@3) — White: weak wall, sente; Black: shut in.
  • Kosumi to 5-5: B1@e, W2@a, B3@xSimple White: decent side + corner; Black: strong top, sente.
  • Kosumi to 3-3: If B has stones on both edges B1@a, W2@e, Black cuts — White: two weak groups, sente; Black: alive in corner.

My question

Examining these variants, it is clear that Black can choose tsuke-osae, 5-5 or (given pincers against both kakaris) 3-3 to achieve the results in the last three options above. But what strategic purpose is served by tsuke-nobi, regardless of where White plays 4 (a/b/c/d)?

What I hope for from the answer

  • It is along the lines of If you want do do X, play A, but if you want to do Y, play tsuke-nobi, i.e. states the strategic goal served, maybe even gives an algorithm for choice. Obviously that will leave me to make some judgement(s), but at least I will know which.
  • It assumes that I restrict myself to the above variants, and presumes my opponent does too, unless that assumption entails a severe loss for me. This because I am trying to impose some sort of limit on how much I learn for an initial grounding in this (and other) joseki. Comments on other variants may also be interesting, but are of subordinate importance to me.
  • 1
    I'm a go noob, but one aspect of this may be that tsuke-nobi works as a kind of a leaning attack. First lean on the one white kakari stone on the left, making the hoshi stone a bit stronger while doing that. Then put heavier pressure on the other kakari stone in the top. Nov 10 at 14:01

2 Answers 2


Black at the 3-3 point is a special strategy that requires supporting stones in the area.

In the fuseki, without supporting stones, black's objective is to separate white, with 3 main options:

  • tsuke nobi against the pincer stone (your diagram)
  • tsuke nobi against the pincered stone
  • 5-5 point

38 Basic Joseki is an older book that favors the first option. In more modern play you'll see the two other options as well. The tsuke nobi keeps the white stones separated, while the 5-5 point will put more emphasis on cental influence.

This page shows a very recent variation from AlphaGo.

  • Thanks, that helps as a start. I have not actually accepted it so far, because what I am actually hoping for, as I have edited my questions to clarify, is which question(s) to ask myself in order to make a choice between the given variants, to which I am provisionally happy to restrict myself while getting a grounding, unless that entails a severe disadvantage.
    – PJTraill
    May 5, 2018 at 13:21
  • Honestly I think you're expecting too much from a forum. Beyond what I said - if you want to split: tsuke nobi, if you want central influence: 5-5 - you would have to look at the rest of the board to understand what black is attacking or defending.
    – Christophe
    May 6, 2018 at 1:45

Actually, corner fighting (so-called Joseki), is still evolving. With AlphaGo and AlphaZero, and the latest other AIs such as FineArt, there's no final conclusion yet.

So, to your question “what strategic purpose is served by tsuke-nobi, regardless of where White plays 4 (a/b/c/d)?” I would argue that, however White continues (a/b/c/d), Black can achieve an acceptable result, so, that's good enough to play tsuke-nobi.

On the strategy? I would rather not worry about it – so long as the journey is entertaining, one doesn't care much where is the destiny.

Or, actually, I even dare say that “strategy” is a word humans use to simplify options – a pincer is to attack, 5-5 is for more air, 3-3 is to attach both side – AIs such as AlphaZero have taught humans that there are many unknown variations and really strong players (best human ones and AIs) can be quite flexible and resourceful so the "destination" is quite unknown.

On 5-5 and 3-3, just my 2 cents, they sound a little bit slow. In the post-AlphaGo era, the best players tend to play a game in which stones conflict with each other, forcing their opponent to respond. In other words, if the pressure is not serious enough, opponent might jolly well tenuki.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .