Hi everybody,

I'm currently reading a book about Yose ,

where that "double sente" situation is exposed :

Click Here To Show Diagram Code

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

The sente for black is this one :

Click Here To Show Diagram Code

[go]$$B
$$ | . . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . . X . . . . O , . . . .
$$ | . . . . . O . . . . O . . .
$$ | . . . X . 1 . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ +----------------------------[/go]

I'm ok with that, but the sequence for white is this one :

Click Here To Show Diagram Code

[go]$$W
$$ | . . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . . X . . . . O , . . . .
$$ | . . 2 a . O . . . . O . . .
$$ | . . . X 1 , . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ +----------------------------[/go]

And here I don't understand why black didn't play at "a", is there anything white can do if black reply at "a" ?

Click Here To Show Diagram Code

[go]$$W
$$ | . . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . 6 X . . . . O , . . . .
$$ | . . 5 2 . O . . . . O . . .
$$ | . . 3 X 1 , . . . . . . . .
$$ | . . . 4 . . . . . . . . . .
$$ +----------------------------[/go]

maybe something like that but then, I cannot find a sequence for white to live in the corner ?

So I don't understand why is not the "best" reply here...

PS : There are a few free yose problems given by the authors of the book here :
https://docs.google.com/file/d/0B3rabuV ... JxME0/edit (but it is in french...)

_________________
Converting the book Shape UP! by Charles Matthews/Seong-June Kim
to the gobook format. last updated april 2015 - Index of shapes, p.211 / 216

Hi oca,

One thing to consider: study the end-game sequences if W gets the first line hane D1 ( in your last diagram ).
What's the difference in end-game between the tiger's mouth and the solid connect ?

Hi EdLee,

Thanks for the hint, that's to keep sente is it ?

Click Here To Show Diagram Code

[go]$$W black ends in sente
$$ | . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . .
$$ | . . X . . . . . . . . . .
$$ | . . . . . O . . . . . . .
$$ | . . . X . . . . O , . . .
$$ | . . 2 . . O . . . . O . .
$$ | . . . X 1 . . . . . . . .
$$ | . . 4 3 5 . . . . . . . .
$$ +--------------------------[/go]

Click Here To Show Diagram Code

[go]$$W black ends in gote
$$ | . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . .
$$ | . . X . . . . . . . . . .
$$ | . . . . . O . . . . . . .
$$ | . . . X . . . . O , . . .
$$ | . . . 2 . O . . . . O . .
$$ | . . 6 X 1 . . . . . . . .
$$ | . . 4 3 5 . . . . . . . .
$$ +--------------------------[/go]

_________________
Converting the book Shape UP! by Charles Matthews/Seong-June Kim
to the gobook format. last updated april 2015 - Index of shapes, p.211 / 216

Hi oca, yes, very good. Sente v. gote.

There are probably other reasons, to do with the shape and corner aji. Maybe others can help you out.

The position can, but need not, be a double sente. It depends on whether the black group is weak or strong on the left side and whether the white group is weak or strong on the lower side and in the center. Depending on global context, it could even be a gote. IOW, the book author seems to have imagined a strong white group on the left side and a strong black group in the center and on the far right side.

Double sente depends upon the rest of the board. Typically each player's threat must be greater than anything else on the board. One flaw with older yose books is that they present certain positions as inherently double sente. It has been known for well over 40 years that there is something wrong with the idea of unconditional double sente. Still, writers often follow tradition and classify some positions as double sente. One who does not is O Meien, 9 dan, who does not even mention double sente in his recent yose book.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

I think that is fair comment, but I would also say that I have always assumed that when doing yose problems that they are considered in isolation and so the remainder of the board do not affect the outcome. With this understanding double sente does make sense.

_________________
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).

Quite the opposite. In isolation the only double sente is seki. Otherwise the value of a double sente is undefined, because of division by zero.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

If the follow up move is large enough then the value is undefined, precisely because of the divide by zero.

_________________
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).

If the only consideration is preventing the first line hane, why not here?

Click Here To Show Diagram Code

[go]$$W
$$ | . . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . . X . . . . O , . . . .
$$ | . . . . . O . . . . O . . .
$$ | . . 2 X 1 , . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ +----------------------------[/go]

or here?

Click Here To Show Diagram Code

[go]$$W
$$ | . . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . . X . . . . O , . . . .
$$ | . . . . . O . . . . O . . .
$$ | . . . X 1 , . . . . . . . .
$$ | . . . 2 . . . . . . . . . .
$$ +----------------------------[/go]

Marcel, DrStraw, It is the responsibility of book authors to present correct examples instead of making dubious assumptions.

Bill, I disagree that the value of double sente would be non-existent. Only a specific kind of value is non-existent: the miai value because the number of local excess moves is zero and, in classic maths, we do not divide by zero. However, a double sente does have a different value: the local swing. At a moment, when it is correct to play a double sente because its threat (a second value!) is so big that the opponent must answer, the locally starting player takes the swing value. If he fails to do so, then the opponent takes it and "gains" it for the according expected score change. Therefore, books should not take pride in hiding double sente.

It's not the only consideration. Making the 4-4 stone stronger is another, preventing this sente sequence for example.

Click Here To Show Diagram Code

[go]$$W
$$ | . . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ | . . X . . . . . . . . . . .
$$ | . . 4 3 . O . . . . . . . .
$$ | . . . X 5 . . . O , . . . .
$$ | . . . 6 . O . . . . O . . .
$$ | . . 2 X 1 , . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ +----------------------------[/go]

It is the responsibility of an author to make himself understood. Excessively verbose explanations are a hindrance to learning. If reasonable assumptions are made in order to simplify the learning process than the overall quality of the book is improved. Fortunately, the majority of authors follow this principle. I can only think of a very few who do not.

_________________
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).

Hi Bill, could you elaborate on this a little bit. Thanks.

Since double sente depends on the (global) positional context, examples ignoring it must be accompanied by an explanation of the made simplifying assumptions. One does not learn what double sente is by saying that it was double sente, period. And it really is not that hard for a book author to present a proper version of the example:

Click Here To Show Diagram Code

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

EDIT: oca,

Click Here To Show Diagram Code

[go]$$W
$$ | . . . . . . . . . . . . . .
$$ | . . O . . . . . . . . . . .
$$ | . . . . . . . X . . X . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . . X . . . . O , . . . .
$$ | . . 2 . . O . . . . O . X .
$$ | . . . X 1 , . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ +----------------------------[/go]

Much later on the left lower side, this will be followed up by one of the following gote sequences:

Click Here To Show Diagram Code

[go]$$B
$$ | . . . . . . . . . . . . . .
$$ | . . O . . . . . . . . . . .
$$ | . . . . . . . X . . X . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . . X . . . . O , . . . .
$$ | . . X . . O . . . . O . X .
$$ | . . . X O , . . . . . . . .
$$ | . . . 3 1 2 . . . . . . . .
$$ +----------------------------[/go]

Click Here To Show Diagram Code

[go]$$W
$$ | . . . . . . . . . . . . . .
$$ | . . O . . . . . . . . . . .
$$ | . . . . . . . X . . X . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . . X . . . . O , . . . .
$$ | . . X . . O . . . . O . X .
$$ | . . . X O , . . . . . . . .
$$ | . . 2 1 3 . . . . . . . . .
$$ +----------------------------[/go]

For both together, one can imagine the simplifying average:

Click Here To Show Diagram Code

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

Now compare this with

Click Here To Show Diagram Code

[go]$$W
$$ | . . . . . . . . . . . . . .
$$ | . . O . . . . . . . . . . .
$$ | . . . . . . . X . . X . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . . X . . . . O , . . . .
$$ | . . . X . O . . . . O . X .
$$ | . . . X 1 , . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ +----------------------------[/go]

after which White has the sente / privilege follow-up

Click Here To Show Diagram Code

[go]$$W 4 might also be at a, depending on global context
$$ | . . . . . . . . . . . . . .
$$ | . . O . . . . . . . . . . .
$$ | . . . . . . . X . . X . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . . X . . . . O , . . . .
$$ | . . b X . O . . . . O . X .
$$ | . 4 a X O , . . . . . . . .
$$ | . . 2 1 3 . . . . . . . . .
$$ +----------------------------[/go]

Result: Black has lost 1 point in comparison to initially responding at b.

This is the pure local value explanation. However, b also makes better eyeshape and leaves behind fewer weaknesses for a white invasion if White should become even stronger around the corner. E.g., if Black can later invade the bottom, the white group there will become stronger and then b is in a better position. IOW, b is also the proper move and the better preparation for a possible black invasion.

Another consideration is this aji in case of Black's tenuki:

Click Here To Show Diagram Code

[go]$$W
$$ | . . . . . . . . . . . . . .
$$ | . . O . . . . . . . . . . .
$$ | . . . . . . . X . . X . . .
$$ | . . X . . . . . . . . . . .
$$ | . . . . . O . . . . . . . .
$$ | . . . X 3 . . . O , . . . .
$$ | . . 2 1 . O . . . . O . X .
$$ | . . a X O , . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ +----------------------------[/go]

Usually a is the attacking move, but White can also create this ko.

is not good. White can even play this if there's enough ko threats so black is 1 1/3 points worse off that the above diagram.

Click Here To Show Diagram Code

[go]$$W
$$ | . . . . . . . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ | . . X . . . . . . . . . . .
$$ | . . 4 3 . O . . . . . . . .
$$ | . . . X 7 . . . O , . . . .
$$ | . . 6 5 . O . . . . O . . .
$$ | . . 2 X 1 , . . . . . . . .
$$ | . . . . . . . . . . . . . .
$$ +----------------------------[/go]

Sure. In The Endgame (1976) Ogawa and Davies discuss the divide by zero problem, which suggests a miai value of infinity for a double sente. Kano, in his Yose Dictionary (in Japanese, 1974) calls the following position a double sente, but he knows that there is something wrong with that. In a somewhat confusing passage he says that White has a greater necessity to answer than Black. In fact, it is a 7 point sente for Black.

Click Here To Show Diagram Code

[go]$$B "Double" sente
$$ --------------------
$$ | . . . . . . . O X .
$$ | . . . X O . . O X .
$$ | X X . X O . O O X .
$$ | . X . X X O . O X ,
$$ | X X X X O O O O X .
$$ | O O O O X X X X X .
$$ | . . . O O O . . . .
$$ | . . . . . . . . . .[/go]

See http://senseis.xmp.net/?DoubleSenteIsRelative .

It is true that I am inferring that the authors were aware of the problems before writing the books. But also one of the classical double sente, the double kosumi on the second line that either player can make and threaten to jump or slide into the opponent's territory, often goes unanswered in pro games. If you give that fact any thought you have to conclude that the play is not sente, much less double sente. Certainly the pros who did not answer thought that it was not double sente.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Bill

If I may be a terrier chasing the quarry first flushed out by Ed...

1. You say the problems with double sente has been known for at least 40 years. I infer from your quotation of Kano's 1974 book that this is the source for "40 years". You add that his text is confusing.

1a. Two points to that. First, I didn't find his text confusing. He is not teaching anything in that section. He is just defining terms, and he gives an example of a "pure" double sente before the one you cite. The one you cite he presents to make a clear point: "However, although this is likewise called double sente you need to recognise that here a difference may arise in the right [to play it] depending on the position." He then goes on to explain that if Black has a higher priority move elsewhere, he may take that instead of fully answering in the cited position in gote. Assuming that's the portion you found confusing (and there's little else in the source), I'm puzzled because it seems straightforward to me - it's double sente but in practice White is more likely to get it. Also, it may be worth adding that he does not use the phrase unconditional double sente, though I accept that some such sort of qualification can be inferred.

1b. However, I may have missed something, and in support of that, the 1985 edition of Kano's dictionary reworks this section entirely, which perhaps indicates that he saw a problem with his first effort. But the problem may not have been in the position cited. Rather it may be because he appears to have taken a completely different approach to boundary play counting in general, and to my inexpert eye he has moved a long way towards O Meien, even using some of the same "rights" vocabulary. In the 1985 book Kano also goes well beyond his first effort in explaining why we need to be cautious about regarding the components of a double sente position as sente, and in support of that he cites a completely different kind of position where life & death are not involved, but aji is, and he adds a section on ajikeshi. However, he also clearly explains that this is really a middle game issue and not a boundary play issue and adds that "In a sense it can be called a pre-boundary play problem, but if we do classify it as a boundary play, it is one to which we will give the highest order of priority." And, also worth adding, he tends to be scrupulous in saying "double sente boundary play" and not just "double sente." Again, I see no confusion in this. So my question boils down to this: why should I care about things like dividing by zero, why should I believe there is any practical value in accepting there may (or may not) be some pedantic flaw in the concept of "double sente", especially when I do not add "unconditional" to the term and I do recognise the difference between middle game and endgame?

2. As I just mentioned, Kano appears to have moved towards the O Meien position even before O did. But O took it further and presents a completely different system - Absolute Counting. I'm in no position to judge on the relative merits of any system, but I think I can judge when one system is starkly different from another. On that basis, I fail to see how lack of mention of double sente in O's system is relevant. He doesn't see the need to use it, so he doesn't seem to regard it as relevant either. He doesn't even bother dismissing it. It's like a paper spoon. He just doesn't need it in his system, so why mention it. He's not saying it's not relevant to other systems of boundary play counting. Instead, he is saying his whole system is better than other systems, but he doesn't appear to have convinced the whole go world yet. Is that a fair summary?

What is this theory of O Meien? Is there anything about it in English? Maybe that Zone Press Park book?

To answer your second question first, here is the first of some threads on O Meien's approach by John Fairbairn. viewtopic.php?t=3008

O Meien's theory is basic combinatorial game theory (CGT). This is no accident. Go was an inspiration for CGT, and go players had come up with a number of the insights of CGT long before CGT was developed. (See http://senseis.xmp.net/?CombinatorialGameTheory ).

O's absolute counting (TM) is equivalent to miai value which is equivalent to CGT temperature (not to be confused with the English go term, temperature, which is related but different). In addition to sente and gote, CGT deals with ambiguous positions and plays, which I did not discover, but named. See http://senseis.xmp.net/?AmbiguousPosition . (CGT does not use the terms, sente and gote, but it is easy to define them in CGT terms. )

O's theory does not include the advances in ko evaluation made by Berlekamp in the 1980s and 90s and by me in the late 1990s. But why should it? Berlekamp published his methods in 1994 for mathematicians, in English. I published my approach in 1998, also for mathematicians and in English. A review by me of ko evaluation written for the 4th International Conference on Baduk was published in 2006, in both English and Korean. I would be quite happy if it were translated into Japanese and Chinese.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.