Using Joseki to respond to non-joseki

For lessons, as well as threads about specific moves, and anything else worth studying.
User avatar
Loons
Gosei
Posts: 1378
Joined: Tue Apr 20, 2010 4:17 am
GD Posts: 0
Location: wHam!lton, Aotearoa
Has thanked: 253 times
Been thanked: 105 times

Using Joseki to respond to non-joseki

Post by Loons »

I am thinking about this with respect to teaching my bot how to play, but it struck me that gainful joseki study is something us non-computers have put a lot of thought into for ourselves.

Obviously these reasonable sequences have something to do with eachother:
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 4 . .
$$ | . . 1 , . . . . . ,
$$ | . . . 3 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 5 , . . . . . ,[/go]
Click Here To Show Diagram Code
[go]$$Bc
$$ --------------------+
$$ . . . . . . . . . . |
$$ . . . . . . . . . . |
$$ . . 8 . . 7 . . . . |
$$ , . . . . . , 6 . . |
$$ . . . . . . 9 . . . |
$$ . . . . . . . . . . |
$$ . . . . . . . . . . |
$$ . . . . . . . . . . |
$$ . . . . . . . . . . |
$$ , . . . . . , . . . |[/go]
Click Here To Show Diagram Code
[go]$$Wcm10
$$ , . . . . . , . . . |
$$ . . . . . . . . . . |
$$ . . . . . . . 3 . . |
$$ . . . . . . . . . . |
$$ . . . . . . . . . . |
$$ . . . . . . . 2 . . |
$$ , . . . . 5 , . . . |
$$ 4 . . . . . 1 . . . |
$$ . . . . . . . . . . |
$$ . . . . . . . . . . |
$$ --------------------+[/go]
Or similarly
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . 4 . . . . . .
$$ | . . 5 . . 2 . . 6 .
$$ | . . . 1 . . . . . ,
$$ | . . . . . . . . . .
$$ | . . 3 . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
Click Here To Show Diagram Code
[go]$$Bcm7
$$ | . . . , . . . . . ,
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 3 . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . 1 . . . . . ,
$$ | . . . . . 2 . . 4 .
$$ | . . . 5 . . . . . .
$$ | . . . . . . . . . .
$$ +---------------------[/go]
Click Here To Show Diagram Code
[go]$$Wcm12
$$ --------------------+
$$ . . . . . . . . . . |
$$ . . . . . . . . . . |
$$ . . . . 3 . . 6 . . |
$$ , . . . . . 1 . 4 . |
$$ . . . . . . . . . . |
$$ . . . . . . . 2 . . |
$$ . . . . . . . . . . |
$$ . . . . . . . 5 . . |
$$ . . . . . . . . . . |
$$ , . . . . . , . . . |[/go]
Now these were all at least kind of joseki, though one could imagine only knowing one sequence and kind of inferring the others.
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.
Kirby
Honinbo
Posts: 9553
Joined: Wed Feb 24, 2010 6:04 pm
GD Posts: 0
KGS: Kirby
Tygem: 커비라고해
Has thanked: 1583 times
Been thanked: 1707 times

Re: Using Joseki to respond to non-joseki

Post by Kirby »

Can you explain more clearly what you mean?

I don't follow. Are you saying that, if you know one joseki, you can infer others?
be immersed
User avatar
Loons
Gosei
Posts: 1378
Joined: Tue Apr 20, 2010 4:17 am
GD Posts: 0
Location: wHam!lton, Aotearoa
Has thanked: 253 times
Been thanked: 105 times

Re: Using Joseki to respond to non-joseki

Post by Loons »

If I only know this joseki:
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 4 . .
$$ | . . 1 , . . . . . ,
$$ | . . . 3 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 5 , . . . . . ,[/go]
And my opponent plays this:
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 3 . .
$$ | . . 1 , . . . . . ,
$$ | . . . a . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . b , . . . . . ,[/go]
I expect 'a' or 'b' to be reasonable moves.

Is this a way you use joseki in games?
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.
User avatar
Loons
Gosei
Posts: 1378
Joined: Tue Apr 20, 2010 4:17 am
GD Posts: 0
Location: wHam!lton, Aotearoa
Has thanked: 253 times
Been thanked: 105 times

Re: Using Joseki to respond to non-joseki

Post by Loons »

And by extension can I expect 'a' and 'b' to be reasonable?
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . . . .
$$ | . . 1 , . . . 3 . ,
$$ | . . . a . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . b , . . . . . ,[/go]
And given all that, can I infer:
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 1 , 2 . . 3 . ,
$$ | . . a . . . . . . .
$$ | . . . b . . . . . .
$$ | . . . c . . . . . .
$$ | . . . d . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
Though now the next moves look quite different.

Edit: So rather than saying I want to infer non-joseki, I want to infer new joseki from the little joseki I know.
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.
DrStraw
Oza
Posts: 2180
Joined: Tue Apr 27, 2010 4:09 am
Rank: AGA 5d
GD Posts: 4312
Online playing schedule: Every tenth February 29th from 20:00-20:01 (if time permits)
Location: ʍoquıɐɹ ǝɥʇ ɹǝʌo 'ǝɹǝɥʍǝɯos
Has thanked: 237 times
Been thanked: 662 times
Contact:

Re: Using Joseki to respond to non-joseki

Post by DrStraw »

I don't follow at all. There may be some commonality between some of the sequences you show but I don't see how knowledge of one of them can imply knowledge of the others. I think that different basic principles are at play.

Having said that, I do think that knowledge of some joseki can help with understanding of other positions which show similarities but which are not intrinsically the same pattern.
Still officially AGA 5d but I play so irregularly these days that I am probably only 3d or 4d over the board (but hopefully still 5d in terms of knowledge, theory and the ability to contribute).
User avatar
Loons
Gosei
Posts: 1378
Joined: Tue Apr 20, 2010 4:17 am
GD Posts: 0
Location: wHam!lton, Aotearoa
Has thanked: 253 times
Been thanked: 105 times

Re: Using Joseki to respond to non-joseki

Post by Loons »

I am very interested in how other people find this obscure.

I would say the 'only joseki I know' has genuinely encoded in it the value of the knight's move approach, an enclosure, extensions, pincers and presses.

When my opponent deviates from joseki I know, typically the first place I consider is where I believed my opponent was meant to play next, and if not that stone, perhaps the usual move after that.

Edit: I know lots and lots of joseki, and there are arbitrarily many cases - especially in contact fights - where a given tewari would be crazy.

I guess what I may be confused about is that I think we might recognise the value of a move because of its place in joseki we know, versus recognising the value of a joseki because of the value of moves it contains.

So I am cherry-picking when I say that these sequences seem - for all their difference - tremendously similar
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 4 . .
$$ | . . 1 , . . . . . ,
$$ | . . . 3 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 5 , . . . . . ,[/go]
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 3 . .
$$ | . . 1 , . . . . . ,
$$ | . . . 4 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 1 , 2 . . 3 . ,
$$ | . . . . . . . . . .
$$ | . . . 4 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
Last edited by Loons on Thu Jul 02, 2015 4:44 pm, edited 1 time in total.
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.
Kirby
Honinbo
Posts: 9553
Joined: Wed Feb 24, 2010 6:04 pm
GD Posts: 0
KGS: Kirby
Tygem: 커비라고해
Has thanked: 1583 times
Been thanked: 1707 times

Re: Using Joseki to respond to non-joseki

Post by Kirby »

Loons wrote:If I only know this joseki:
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 4 . .
$$ | . . 1 , . . . . . ,
$$ | . . . 3 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 5 , . . . . . ,[/go]
And my opponent plays this:
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 3 . .
$$ | . . 1 , . . . . . ,
$$ | . . . a . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . b , . . . . . ,[/go]
I expect 'a' or 'b' to be reasonable moves.

Is this a way you use joseki in games?
Ah, it's an interesting idea. I never thought about it that way before.
I guess the idea is based on the proverb that "the opponent's good move is a good move for me".

I don't know if it always works, but it might be good for identifying candidate moves.

For example, if I only know this:
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . 3 . . . .
$$ | . 4 . 1 . . . . . ,
$$ | . . . . . . . . . .
$$ | . . 2 . . . . . . .
$$ | . . 5 . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
and my opponent plays this:
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 5 . . 3 . . . .
$$ | . 4 . 1 . . . . . ,
$$ | . . . . . . . . . .
$$ | . . 2 . . . . . . .
$$ | . . a . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
then 'a' seems questionable:
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . 5 . . 3 . . . .
$$ | . 4 . 1 . . . . . ,
$$ | . . . . . . . . . .
$$ | . . 2 . . . . . . .
$$ | . . 6 . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
be immersed
User avatar
Loons
Gosei
Posts: 1378
Joined: Tue Apr 20, 2010 4:17 am
GD Posts: 0
Location: wHam!lton, Aotearoa
Has thanked: 253 times
Been thanked: 105 times

Re: Using Joseki to respond to non-joseki

Post by Loons »

I think of contact moves as a special case, expanding contact to mean 'eight-connected'.

Edit: Also, your '5' is not a tenuki-point in joseki.

AND '6' is bad suji but in the right direction.
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.
User avatar
ez4u
Oza
Posts: 2414
Joined: Wed Feb 23, 2011 10:15 pm
Rank: Jp 6 dan
GD Posts: 0
KGS: ez4u
Location: Tokyo, Japan
Has thanked: 2351 times
Been thanked: 1332 times

Tip: Styles can be applied quickly to selected text.

Post by ez4u »

The most common reply to all 3-4 based pincers is a jump straight up: 2-space jump against 2-space or 3-space pincers, 1-space jump from a 1-space low pincer. The one exception is the 1-space high pincer where the keima shoulder hit against the pincer stone has been played slightly more often than the 1-space jump. I do not think that your method will yield any of these replies. This seems a major weakness.

Also where do you start? By far the most common reply to the diagonal response to the original approach is to tenuki. Your 2-space extension ranks fifth. So which 'reasonable' sequences will you seed your program with?

Going back to your original question, is this how I think about joseki? No.
Last edited by ez4u on Thu Jul 02, 2015 5:03 pm, edited 1 time in total.
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21
Kirby
Honinbo
Posts: 9553
Joined: Wed Feb 24, 2010 6:04 pm
GD Posts: 0
KGS: Kirby
Tygem: 커비라고해
Has thanked: 1583 times
Been thanked: 1707 times

Re: Using Joseki to respond to non-joseki

Post by Kirby »

Loons wrote:I think of contact moves as a special case, expanding contact to mean 'eight-connected'.

Edit: Also, your '5' is not a tenuki-point in joseki.

AND '6' is bad suji but in the right direction.
Like I said, it's an interesting idea, and can give you candidate moves - just like the proverb can ("the opponent's good move is a good move for me").

But I'd consider it a heuristic more than a definite rule.
be immersed
User avatar
ez4u
Oza
Posts: 2414
Joined: Wed Feb 23, 2011 10:15 pm
Rank: Jp 6 dan
GD Posts: 0
KGS: ez4u
Location: Tokyo, Japan
Has thanked: 2351 times
Been thanked: 1332 times

Re: Using Joseki to respond to non-joseki

Post by ez4u »

Loons wrote:I think of contact moves as a special case, expanding contact to mean 'eight-connected'.

...
Careful! Haven't you just declared 4 below an 8-connected special case?
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 3 . .
$$ | . . 1 , . . . . . ,
$$ | . . . 4 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21
User avatar
Loons
Gosei
Posts: 1378
Joined: Tue Apr 20, 2010 4:17 am
GD Posts: 0
Location: wHam!lton, Aotearoa
Has thanked: 253 times
Been thanked: 105 times

Re: Using Joseki to respond to non-joseki

Post by Loons »

ez4u wrote:
Loons wrote:I think of contact moves as a special case, expanding contact to mean 'eight-connected'.

...
Careful! Haven't you just declared 4 below an 8-connected special case?
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 3 . .
$$ | . . 1 , . . . . . ,
$$ | . . . 4 . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
My joseki choice was biased as "variation of the first joseki in Takao's" with a 4-4 variation tacked on. That is indeed a special case and a reasonable move. My definition of reasonable is something like "not unambiguously a mistake in a joseki dictionary". This is the first joseki given after "two space jump and counter pincer : see the high pincer".
Click Here To Show Diagram Code
[go]$$Wcm5
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . X . . O . 4
$$ | . . O , . . . . . ,
$$ | . . 1 X . . . 5 . .
$$ | . . . 2 . . . . . .
$$ | . . 3 6 . . . . . .
$$ | . . 7 8 . . . . . .
$$ | . . . 9 0 . . . . .
$$ | . . . , . . . . . ,[/go]
Click Here To Show Diagram Code
[go]$$Wcm15
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . X . . O . X
$$ | . . O , . . . . . ,
$$ | . 5 O X . . . O . 6
$$ | . 3 2 X . . . . . .
$$ | . 4 O X . . . 7 . .
$$ | . . O X . 8 . . . .
$$ | . . . O X . . . . .
$$ | . . . 1 . . . . . ,[/go]
NB. I am exploring this idea. I would guess the special case about contact moves would be something like, if what would be a later move in the contact fight was played before the earlier move, that invalidates the other move(s) as candidates.

If both players "only know" this joseki but black accidentally played this 3, I think the following would be implied:
Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 4 . 5
$$ | . . 1 , . . . . . ,
$$ | . . . a . . . 6 . .
$$ | . . . . . . . . . .
$$ | . . 3 . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
Because while 'a' is an earlier move in the joseki than 5 or 6, it is contact-linked to 3. The next available moves not contact linked to 3 are 4 and 5 (original 3 and 8).

Now, this is not the best sequence in the world but it's also not that unreasonable. The heuristic that gave it is very minimal. I am wondering if the heuristic can be generalised in an egalitarian way to cover even more cases relatively reasonably.
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.
User avatar
Loons
Gosei
Posts: 1378
Joined: Tue Apr 20, 2010 4:17 am
GD Posts: 0
Location: wHam!lton, Aotearoa
Has thanked: 253 times
Been thanked: 105 times

Re: Using Joseki to respond to non-joseki

Post by Loons »

Click Here To Show Diagram Code
[go]$$Bc
$$ +---------------------
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . . 2 . . 3 . 6
$$ | . . 1 , . . . . . ,
$$ | . . . x . . . 7 . 8
$$ | . . . . 4 . . . . .
$$ | . . 5 . . . . 9 . .
$$ | . . . . . . . . . .
$$ | . . . . . . . . . .
$$ | . . . , . . . . . ,[/go]
Perhaps more appropriate. Discounting 'x' as touching a stone, and excluding stones 'x' would have touched, we get something like a counter pincer joseki.
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.
User avatar
Loons
Gosei
Posts: 1378
Joined: Tue Apr 20, 2010 4:17 am
GD Posts: 0
Location: wHam!lton, Aotearoa
Has thanked: 253 times
Been thanked: 105 times

Re: Using Joseki to respond to non-joseki

Post by Loons »

Sorry for harping and harping on about it, but I would like to clarify:

Joseki are in no sense a mystery. They are in some sense forcible sequences that each side can feel satisfied with unless they made a mistake in inciting it.

Satisfaction comes in the form of the most points, thickness, sente, aji, relationship to other corners, fighting shape ... And the struggle to get these implies a lot of tesuji knowledge, counting and fighting skill.

It is in no sense surprising that one joseki almost explicitly says what to do in another joseki by move priority and position in many cases, because high priorities are high priorities are high priorities.
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.
Kirby
Honinbo
Posts: 9553
Joined: Wed Feb 24, 2010 6:04 pm
GD Posts: 0
KGS: Kirby
Tygem: 커비라고해
Has thanked: 1583 times
Been thanked: 1707 times

Re: Using Joseki to respond to non-joseki

Post by Kirby »

Ok, dude.
be immersed
Post Reply