Go problems don't bring any result?

Cassandra · Post by **Cassandra** » Sat Dec 17, 2016 6:08 am

RobertJasiek wrote:Cassandra, identifying obvious failures does not necessarily involve shape analysis.

Dear Robert, "Shape Analysis" serves identifying obvious CANDIDATE moves.

It is not necessary at all to waste your time with moves that will fail for sure.

Cassandra · Post by **Cassandra** » Sat Dec 17, 2016 6:26 am

RobertJasiek wrote:The stronger I have become, the less useful shapes have become.

In his afterword to Igo Hatsuyōron, Inoue Dōsetsu Inseki said that once you practise "combinations" (i.e. "move sequences") the study of the "formations" (i.e. "shape") becomes even more important.

So, what do you mean with "less useful" ? E.g. massively decreased value-to-time-ratio when trying to look any deeper ?

John Fairbairn · Post by **John Fairbairn** » Sat Dec 17, 2016 7:48 am

Robert

Ok, a lot clearer now, thanks, but I still feel (a) regular is a very peculiar word to use and (b) you are repeating Aesop's fable of a mountains straining to produce a mouse.

If I've understood it now, though, you are saying your version of standard reading can be enhanced (i.e. pulled back from brute force) by using e.g. visual inspection of nakade shapes. If that's the case, all the more reason to learn the necessary shapes and priyomes. No?

Cassandra · Post by **Cassandra** » Sat Dec 17, 2016 9:41 am

"Shape Analysis" does include "reading", doesn't it ?

Cassandra wrote:
$$B Looking forward
$$ -------------------------------
$$ . . . . O . . . . M M . . . . .
$$ . . . X O O . O O O . . Y . . .
$$ . . X . X O O X O X . X . . . .
$$ . . . . X X X X X X . . X . . .
$$ . . . . . . . . . . . . . . . .
$$ . . . . . . . . . . . . . . . .
Click Here To Show Diagram Code
[go]$$B Looking forward $$ ------------------------------- $$ . . . . O . . . . M M . . . . . $$ . . . X O O . O O O . . Y . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]
Black's marked stone at the right is looking forward to supporting any friendly stone at the crossed intersections.

Shape Analysis (1).

Click Here To Show Diagram Code: [go]$$B Looking forward (1) $$ ------------------------------- $$ . . . . O . . ? . M . . . . . . $$ . . . X O O . O O O . . Y . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

Reading (1).

Click Here To Show Diagram Code: [go]$$B Looking forward (1) $$ ------------------------------- $$ . . . . O . . ? 6 1 3 4 . . . . $$ . . . X O O . O O O 2 5 X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

Click Here To Show Diagram Code: [go]$$B Looking forward (1) -- False eye at the right -- Black to move $$ ------------------------------- $$ . . . . O . . ? O . 7 O . . . . $$ . . . X O O . O O O O X X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

Shape Analysis (2).

Click Here To Show Diagram Code: [go]$$B Looking forward (2) $$ ------------------------------- $$ . . . . O . . . . . M . . . . . $$ . . . X O O . O O O ? . Y . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

Reading (2).

Click Here To Show Diagram Code: [go]$$B Looking forward (2) -- False eye at the right -- Black to move (two times) $$ ------------------------------- $$ . . . . O . . . . 4 1 2 . . . . $$ . . . X O O . O O O ? 3 X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

There is only one first move to consider.

Cassandra wrote:
$$B Dead eye shape (1)
$$ -------------------------------
$$ . . . . O T S T . . . . . . . .
$$ . . . X O O T O O O . . X . . .
$$ . . X . X O O X O X . X . . . .
$$ . . . . X X X X X X . . X . . .
$$ . . . . . . . . . . . . . . . .
$$ . . . . . . . . . . . . . . . .
Click Here To Show Diagram Code
[go]$$B Dead eye shape (1) $$ ------------------------------- $$ . . . . O T S T . . . . . . . . $$ . . . X O O T O O O . . X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]
At the left, there is a dead triangle four-point eye shape visible which central point is still unoccupied.

Click Here To Show Diagram Code: [go]$$B Dead eye shape (1) -- Occupying the vital point $$ ------------------------------- $$ . . . . O . 1 . . . . . . . . . $$ . . . X O O . O O O . . X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

The rest of the sequence is straightforward.
There is no need to read (again) what you already know.

RobertJasiek · Post by **RobertJasiek** » Sat Dec 17, 2016 11:44 am

If "shapes" are meant to be "shapes + reading", we get something more powerful. Only shapes in tactical problems have the disadvantage that very similar shapes can have opposite results, different kinds of first moves or move orders etc. I have not done study for "shapes + reading" as to whether such is avoided.

Is it known what exactly Dosetsu meant when speaking of "shapes"? Static shapes, dynamic shapes a la John, "shapes + reading" a la Cassandra or something else? Hence can we be sure what he meant when saying something amounting to "if practising /move sequences/, the study of /shape/ becomes even more important"?

Shapes having become less useful as I became stronger: as a kyu and partly as a low dan, still I was rather much of a believer of Japanese literature / verbal ideas, among which shapes took a prominent role and, due to weak translations, mostly in the static sense. I would use shapes for every purpose and consider good shape itself to be something worth striving for. The stronger I became the more exceptions popped up; eventually I found myself in a situation in which abandoning shape adherence became the better option in most cases. This was even before I heard of modern Korean fighting style.

A small number of kinds of shape applications have survived, esp. terminal shapes for abbreviating reading (but I had to learn to be careful not to overlook the environment and side conditions), thick shapes as good candidate moves to always be at least considered, the simplest model shapes as samples for functional classes such as throw-in, shapes of rules arcana.

The most noteworthy omission is non-terminal shapes in tactical problems; I have basically lost faith in them; instead, I consider move candidates due to functions / purposes / aims and rather expect the unexpected shapes to also be valid candidates if only there is some function (or at least two threats). Also techniques can be worth considering and would be much more powerful than shapes.

Before recognising the greater importance of regular reading, I had the plan to study problem collections for their move functions and techniques in order to find a higher level structure of combined techniques. John and Thomas (Cassandra) have tried, AFAICS, to some, IMO, still modest extent. I think that the effort must be 10 times greater to possibly find a more powerful structure, something Dosetsu might have seen. Even if such amounted to a collection of some thousands of combinations of shapes, techniques, special kinds of sequences, decision-makings of reading, we would reveal a sort of language with a greater chance of competing with regular reading more often and, so the hope, provide faster problem solving. I do not expect a quick discovery; 10 times more effort is no over-statement, and it could be more.

Regular reading is already rather efficient due to its built-in simplifications. Furthermore, regular reading is enhanced by ca. a dozen of additional, sometimes applicable simplifications. Quite likely, more study of the method and its application can provide more additions. The cumbersome enhancement is also using the various techniques; to make this efficient, conditions for the applicability and combined use must be found and described.

Who said that learning nakade shapes could be forgone?:) Basics must not be omitted. Same for standard shapes, such as problems after josekis.

Bill Spight · Post by **Bill Spight** » Sat Dec 17, 2016 11:45 am

First problem I have composed after midnight.

OK, here is my solution mainline diagram.

Click Here To Show Diagram Code: [go]$$B Black to play and kill $$ ------------------------------- $$ . . . . O . 1 5 . 4 3 7 8 . . . $$ . . . X O O . O O O 2 6 X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

@ 7

More later about chunking.

I suppose that Regular Reading ™ would also start with

, because it is obvious. And indeed it is for experienced players, even those who do not do tsumego. But what if it were not obvious? It would show up with shape analysis that indicates two potential eyes adjacent to it, which it prevents. All you have to do is see them.

RobertJasiek · Post by **RobertJasiek** » Sat Dec 17, 2016 11:56 am

In your reading, move 4 is your first chosen interesting next move candidate at that branch. This makes it part of your regular reading. When you will have read the subtree (incl. 6@7 and other necessary variations), you come back to the branch at move 4 and choose your next interesting move candidate, such as 4@5, 4@6 or 4@7, find that all move 4s fail, conclude that move 3 succeeds, study alternatives for move 2 - et viola.

Bill Spight · Post by **Bill Spight** » Sat Dec 17, 2016 12:09 pm

RobertJasiek wrote:In your reading, move 4 is your first chosen interesting next move candidate at that branch. This makes it part of your regular reading.

Thanks, Robert.

For everyone else, I remarked that I liked

, but was not sure that it was part of Regular Reading. (I edited that out because I thought that it did not add much to the discussion.) It is not exactly clear what moves are part of Regular Reading or not. Regular Reading obviously prunes the game tree, except for small problems.

Why do I like

? Consider this alternate line of play.

Click Here To Show Diagram Code: [go]$$B Black to play and kill $$ ------------------------------- $$ . . . . O . 1 7 . 6 3 5 8 . . . $$ . . . X O O . O O O 2 4 X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

@ 5

A transposition, but I think that in this line

is obvious, but in my mainline,

Click Here To Show Diagram Code: [go]$$B Black to play and kill $$ ------------------------------- $$ . . . . O . 1 5 . 4 3 . . . . . $$ . . . X O O . O O O 2 6 X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

is not so obvious.

Knotwilg · Post by **Knotwilg** » Sat Dec 17, 2016 12:28 pm

I recommend the Hitachi go problem series, elementary first.

http://www.hitachi.co.jp/Sp/tsumego/index-e.html

1. there are 1077 problems and counting, so there's plenty to do before you do the second iteration and "remembering" will converge with "knowing"
2. you can look at the solutions; as said, you won't remember any of them the second time around; rather you will know the solution by solving many similar problems
3. they are within reach of single digit kyu (actually most players IMO)

I took the latest problem as an example

Click Here To Show Diagram Code: [go]$$B Black to live in the corner $$--------------- $$| . . . X O . . $$| . . O X O . . $$| X X . X O . . $$| O O X X O . . $$| . O O O . . . $$| . . . . . . . $$| . . . . . . .[/go]

Please try before reading on.

1) A newcomer may try brute force here. They may try all of the points except 1-1 and 2-1, and have to calculate about 5! = 120 moves.

2) A more experienced player knows that connecting is a dead end and the solution will be either 2-2 or 1-2.
We start with 2-2. It looks very promising, but there's a catch. If White plays 1-2 next, there will be a ko. This may be something a relatively experienced player still misses.

3) So, we play 1-2. It doesn't take much to understand that this is the answer. I'm sure that most, if not all, players of my level will solve this problem.

I would rate it as a 9k-6k problem, which some players in this category may think to have solved too soon, playing 2-2.
In retrospect, which of my 3 approaches resembles yours most?

dfan · Post by **dfan** » Sat Dec 17, 2016 12:42 pm

Black has to partition his eyespace in two, and the two eyes will clearly be horizontally rather than vertically related. His biggest other issue is the 3-3 cutting point and lack of liberties on the right.

I don't even consider connecting at the 3-3 point.

My thinking is influenced by the fact that I know that this is a tsumego problem, not an actual game position. 2-2 looks like it is clearly a "try" (chess problem jargon for a move you are supposed to consider first but which has a refutation). White's descent giving atari is the easiest response to look at, but doesn't work. I have learned to look at the hane as well in these sort of situations (filling opponent's eyespace while threatening low-liberty strings) and indeed it causes a ko.

The only other clearly productive move is the 1-2. At this point I brute-force read White's three interesting tries, none of which work.

In a game I would follow the same sort of reasoning but would probably be less immediately suspicious of the 2-2 point.

Bill Spight · Post by **Bill Spight** » Sat Dec 17, 2016 2:09 pm

Thanks, Dieter.

Click Here To Show Diagram Code: [go]$$Bc Black to live in the corner $$--------------- $$| C 1 . B O . . $$| . . O B O . . $$| X X . B O . . $$| O O B B O . . $$| . O O O . . . $$| . . . . . . . $$| . . . . . . .[/go]

My first thought was

, but I am a bit pressed to say why. (Actually, I saw a White stone there. Go figure.

) Because of the proverb? I don't think so. Familiarity? I don't think so. This problem is unlike any I have ever seen, to the best of my memory. I think it is because

guarantees the eye on the A-19. It also saves the

stones from atari, but so does B-18.

Click Here To Show Diagram Code: [go]$$Bc Black to live in the corner $$--------------- $$| . 2 . X O . . $$| C 1 O X O . . $$| X X . X O . . $$| O O X X O . . $$| . O O O . . . $$| . . . . . . . $$| . . . . . . .[/go]

OTOH,

on B-18 does not guarantee the eye on A-18.

Jhyn · Post by **Jhyn** » Sat Dec 17, 2016 2:09 pm

This is a side point so I hide it.

Knotwilg · Post by **Knotwilg** » Sat Dec 17, 2016 5:07 pm

Jhyn wrote:This is a side point so I hide it.
Knotwilg wrote:I recommend the Hitachi go problem series, elementary first.

http://www.hitachi.co.jp/Sp/tsumego/index-e.html
Hi Knotwilg,

Do you know if there is a way to get/buy the problems in batch so as to be able to solve them without an Internet connection? I could download automatically every page but that looks rude, to say the least. I couldn't find any solution in the English version and wondered if you knew of one, maybe in the Japanese version. I can pay a membership fee if needed.

Sorry I don't. I'm connected often enough to work them online.

Cassandra · Post by **Cassandra** » Sun Dec 18, 2016 1:27 am

Just a tiny complement to Bill's thoughts ...

Click Here To Show Diagram Code: [go]$$B Dead eye shape (1) $$--------------- $$| T T . X O . . $$| T S Q X O . . $$| X X . X O . . $$| O O X X O . . $$| . O O O . . . $$| . . . . . . . $$| . . . . . . .[/go]

Click Here To Show Diagram Code: [go]$$B Dead eye shape (2) $$--------------- $$| T S T X O . . $$| T T O X O . . $$| X X . X O . . $$| O O X X O . . $$| . O O O . . . $$| . . . . . . . $$| . . . . . . .[/go]

Click Here To Show Diagram Code: [go]$$B Snaky eye shape $$--------------- $$| C C C X O . . $$| C ? W X O . . $$| X X C X O . . $$| O O X X O . . $$| . O O O . . . $$| . . . . . . . $$| . . . . . . .[/go]

Think twice before creating a bent shape in the corner.
You might be unable to (easily) access the resulting vital points.

Click Here To Show Diagram Code: [go]$$B Labyrinthine eye shape $$--------------- $$| C ? C X O . . $$| C C W X O . . $$| X X C X O . . $$| O O X X O . . $$| . O O O . . . $$| . . . . . . . $$| . . . . . . .[/go]

Please note the value of the 1-2-points for easily creating an eye in the corner.

Bill Spight · Post by **Bill Spight** » Sun Dec 18, 2016 5:59 pm

A bit on chunking

Chunking has to do with memory, and short term memory in particular. When we are solving go problems we utilize short term memory (working memory), and probably long term memory, as well. Our short term memory is limited; we can hold only a handful of things in short term memory at one time. However, the amount of information in these things is not fixed. By combining things into larger units we can hold and process more information in working memory. These larger units are called chunks, and the process of forming chunks is called chunking.

Let me illustrate chunking with the problem I composed the other night.

Click Here To Show Diagram Code: [go]$$B Black to play and kill, mainline $$ ------------------------------- $$ . . . . O . 1 5 . 4 3 7 8 . . . $$ . . . X O O . O O O 2 6 X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

@ 7

We can make a chunk out of

-

,

. It forms a recognizable pattern of play. Even beginning go players can do so. That is fine as far as remembering this sequence of play is concerned. However, that sequence is meaningful, and the meaning is good for chunking. The sequence achieves a purpose.

Here is the position after

.

Click Here To Show Diagram Code: [go]$$Bm9 A half eye $$ ------------------------------- $$ . . . . O . X X . O C . W . . . $$ . . . X O O . O O O O O X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

The marked point is a half eye. White to play can make an eye there, Black to play can prevent an eye there. The fact that this is a half eye may be something that is already in long term memory, or it may be something that we have discovered while working on the problem. This gives a meaning to

. It takes away the potential eye.

We discovered the meaning of

by considering White's play, even though it is not White's turn. This suggests the well known heuristic of allowing your opponent to make two moves in a row while reading. While doing so may be inefficient in any given case, it may help both in chunking and in coming up with candidate moves.

OC, since we are only talking about one play, we have no chunk yet. But note that

is not necessary to the half eye.

Click Here To Show Diagram Code: [go]$$Bm9 A half eye $$ ------------------------------- $$ . . . . O . X X . O C . . . . . $$ . . . X O O . O O O O O X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

Now let's back up to the position before

.

Click Here To Show Diagram Code: [go]$$Wm8 Two stones in atari $$ ------------------------------- $$ . . . . O . X X . O X B . . . $$ . . . X O O . O O O O O X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

What about this position? How many eyes does White have with the two stones in atari? Zero. Black does not have to do anything. If White takes the two stones, Black can just throw in at

.

Now we have a chunk in the play sequence,

-

. We also have a chunk in the shape of the two stones in atari on the first line and the helper stone on the second line. We also have a chunk combining the shape chunk and the play chunk. This zero eye occurs frequently in life and death problems, and experienced solvers will have these chunks in long term memory. It may not occur so often in real games, because the players do not bother to play the sequence out to that point.

Now let's back up one more play.

Click Here To Show Diagram Code: [go]$$Bm7 A half eye $$ ------------------------------- $$ . . . . O . X X . O B . . . . . $$ . . . X O O . O O O O O X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

Now we have a half eye again. The key point is the same for Black, but White can play the hane.

IMX, a lot of players have a blind spot for under the stones play, and could easily miss it. However, if they have the shape chunk for the zero eye in long term memory, they could see the right play.

We have another chunk with the sequence,

-

,

, and another shape chunk for the half eye.

Now let's look at a different, but related variation.

Click Here To Show Diagram Code: [go]$$Wm4 Variation $$ ------------------------------- $$ . . . . O . X 1 . 2 B 3 . . . . $$ . . . X O O . O O O O 4 X . . . $$ . . X . X O O X O X . X . . . . $$ . . . . X X X X X X . . X . . . $$ . . . . . . . . . . . . . . . . $$ . . . . . . . . . . . . . . . .[/go]

threatens to make an eye with 5, so Black plays there. Now

puts two Black stones in atari, but

makes a zero eye shape. This shape is a variant of the other zero eye shape. If White captures Black has the same throw-in at

. The shapes are not exactly the same, but with a little abstraction, they can be combined into one chunk. The same holds for the sequence chunks. Different plays capture the two stones, but they produce the same half eye.

Now let us suppose that the solver overlooked

in the main line, but reads out this variation. Here Black does not have to make an under the stones play to form the zero eye. The player may then transfer the chunks learned here to the other variation.

Chunking is something that we all do automatically. However, we may enhance the process by asking about the meaning of plays, by forming sub-problems and subgoals. Also by abstraction and labeling. The terms, half eye and zero eye, help us to fix the shapes and plays in memory.

Life In 19x19

Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?

Re: Go problems don't bring any result?