Endgmame question

[lichigo]

Hello,

When I studied in Korea I solved an endgame book and at that time my teacher told me this move in A is less than 4 but more than 3. I am reading robert jasiek's books and when I use this methode I found 1⅔ pt for this move in A. Am I right?

[Harleqin]

I think you cut off too much on the right to surely say.

Note that this would be consistent with your korean teachers using »deiri« counting (swing) but Robert using »miai« counting (difference from average expectation).

[kvasir]

Why was it again that a seemingly handful of Western amateurs promote the miai method while the great majority of Go players everywhere use the delta method?

[RobertJasiek]

Why was it again that a seemingly handful of Western amateurs promote the miai method while the great majority of Go players everywhere use the delta method?

There are both Western and Eastern, amateur and professional players that use traditional ("delta") and / or modern ("miai") endgame theory.

Both methods have their uses but traditional endgame theory is almost only useful for some of the most basic decisions while modern endgame theory can be used for all decisions.

If easily available teaching (verbal, books, videos) and talk is a correct indication, traditional endgame theory is still more popular. Neglecting almost all endgame theory beyond some of the most basic decisions is also still more popular.

People understanding more of endgame theory than only some of the most basic decisions promote modern endgame theory to enable all decisions and avoid almost all mistakes. - Players neglecting endgame theory beyond some of the most basic decisions prefer traditional endgame theory because they are too lazy to improve knowledge of endgame theory significantly beyond the means they already know. People only teaching traditional endgame theory tend to make frequent mistakes in evaluation, characterisation of types, move order and duration of correct local play.

[John Fairbairn]

Why was it again that a seemingly handful of Western amateurs promote the miai method while the great majority of Go players everywhere use the delta method?

This presumably a rhetorical question, but I'd very much like to know the answer myself. O Meien seems to kick against deiri and claims pros in China (by which he may mean Taiwan) all do likewise. But he shies away from calling his own (Chinese) method miai counting. He says 'absolute counting'. Is that really different?

If anyone is prepared to give an answer, I'd prefer it in English and not numberese.

Oh, while we are (supposedly) on the topic of the endgame. I came across a piece of text in a book about Hikaru no Go a couple of days ago which I will use as part of my long-running campaign to get people to say boundary play instead of endgame. It said:

陣地の境界線をはっきりさせるために石を置いていく。これをヨセとおうぞ。

DeepL gives this as : "Stones are placed on the ground to demarcate the boundaries of the camp. This is called yose."

I would give it as: "When we put down stones to clarify the boundary lines of territories, we call this yose." (For the ambiguity hunters, we know its stoneS because of the repetition implied in いく.)

But, either way, "endgame" is conspicuous by its absence.

[RobertJasiek]

AFAIK, miai counting, absolute counting and modern endgame theory are all just different names for the same study field.

"Endgame" includes boundary plays during the opening, middle game and endgame. "Boundary play" occurs during the opening, middle game and endgame.

[lichigo]

So is 1 2/3 is correct ?

[John Fairbairn]

AFAIK, miai counting, absolute counting and modern endgame theory are all just different names for the same study field.

Thanks. I have always wanted to assume that, but Bill Spight always gave me the impression he demurred. And as for O Meien's new term, well I suppose he had a book to sell.

"Endgame" includes boundary plays during the opening, middle game and endgame.

Not to me. If I order a steak I don't expect to get a hamburger.

If you want to argue that the endgame (proper) includes more than the yose, i.e. the boundary plays, I'm fine with that, but it's not a new idea. The Japanese atsui covers many of the other aspects, and timing is also an old big topic. Then we can add kimari to the mix, etc etc.

However, the answer I'm looking for is to the question of why deiri has been the favoured medium in Japan when they were well aware of miai counting. The best explanation I've come up with is that they are both accountancy terms but deiri accounting is more usual (easier?) in real life, and the same attitude just carried over into go. But that's just my own speculation.

[RobertJasiek]

Within modern endgame theory, there are different approaches, such as without or with combinatorial game theory, which is like mechanics without or with considering each elementary particle separately:) Bill might have referred to such advanced study. From the practical point of view, the infinitesemal differences can often be ignored.

[kvasir]

This presumably a rhetorical question, but I'd very much like to know the answer myself. O Meien seems to kick against deiri and claims pros in China (by which he may mean Taiwan) all do likewise. But he shies away from calling his own (Chinese) method miai counting. He says 'absolute counting'. Is that really different?

Maybe a rhetorical question in the sense of creating dramatic effect but not in the sense of making a point. I am also interested in the "why" because we seem to have people adopting the method and the notation, especially in this forum and on sensei's.

I think many pros might just count the end positions in their head, count the difference between positions and basically do dead reckoning (I mean not use any recognizable method). They also seem to be able to switch gear easily and apply different methods. As far as I can tell, pros that do count endgame moves (or teach how to do it) use what one might call the "2x miai counting" because the values are doubled (for gote moves) and they certainly take into account how many moves are played, what is sente and so on. It is possible that Chinese pros, who use half counting (or how to call it), may use same values as miai counting -- I don't really know but I assume they have the half count in mind when the game is about finished.

I wasn't going to write so much, I was just trying to start some discussion. There are so many things that could be said about endgame theory really. I am not taking a firm side, but it seems to cause enough confusion to say -1.25 instead of 2.5 gote for white that you rarely see people do the former. At least that is my experience.

[Harleqin]

So is 1⅔ is correct?

I think so (assuming no further relevant stones to the right), but I'm not an expert in such exact values.

However »more than 3 but less than 4« is also correct (using a different, also popular, counting method, which generally gives about double the numeric value).

Sorry for the derailing discussion.

[lichigo]

Thank you very much. Actually is it also interesting^^
I wasn't sure about the value so thank you.

[RobertJasiek]

$$B initial position
$$ | X X X . . . . . .
$$ | . . X . . . . . .
$$ | . X X X X O O O O
$$ | . X O O O . . . .
$$ | . . . . . . . . .
$$ ------------------

Click Here To Show Diagram Code

[go]$$B initial position
$$ | X X X . . . . . .
$$ | . . X . . . . . .
$$ | . X X X X O O O O
$$ | . X O O O . . . .
$$ | . . . . . . . . .
$$ ------------------[/go]

$$B locale
$$ | X X X . . . . . .
$$ | C C X . . . . . .
$$ | C X X X X O O O O
$$ | C X O O O . . . .
$$ | C C C C C C . . .
$$ ------------------

Click Here To Show Diagram Code

[go]$$B locale
$$ | X X X . . . . . .
$$ | C C X . . . . . .
$$ | C X X X X O O O O
$$ | C X O O O . . . .
$$ | C C C C C C . . .
$$ ------------------[/go]

$$B assuming Black 3, count 5
$$ | X X X . . . . . .
$$ | . . X . . . . . .
$$ | . X X X X O O O O
$$ | . X O O O . . . .
$$ | . . 1 3 . 4 . . .
$$ ------------------

Click Here To Show Diagram Code

[go]$$B assuming Black 3, count 5
$$ | X X X . . . . . .
$$ | . . X . . . . . .
$$ | . X X X X O O O O
$$ | . X O O O . . . .
$$ | . . 1 3 . 4 . . .
$$ ------------------[/go]

To save time, I make some assumptions, which should be verified or refuted.

$$W Variation I, assuming gote
$$ | X X X . . . . . .
$$ | . . X . . . . . .
$$ | . X X X X O O O O
$$ | . X O O O . . . .
$$ | . . 1 C C C . . .
$$ ------------------

Click Here To Show Diagram Code

[go]$$W Variation I, assuming gote
$$ | X X X . . . . . .
$$ | . . X . . . . . .
$$ | . X X X X O O O O
$$ | . X O O O . . . .
$$ | . . 1 C C C . . .
$$ ------------------[/go]

Looking up the count of Black's region, it is 3 2/3. White's region has -3 in the locale. The count is 2/3.

If variation I dominates, the initial gote count is (5 + 2/3) / 2 = 2 5/6 and gote move value is (5 - 2/3) / 2 = 2 1/6.

$$W Variation II, assuming Black 2 - White 3
$$ | X X X . . . . . .
$$ | . . X . . . . . .
$$ | . X X X X O O O O
$$ | . X O O O . . . .
$$ | . 1 2 3 C . . . .
$$ ------------------

Click Here To Show Diagram Code

[go]$$W Variation II, assuming Black 2 - White 3
$$ | X X X . . . . . .
$$ | . . X . . . . . .
$$ | . X X X X O O O O
$$ | . X O O O . . . .
$$ | . 1 2 3 C . . . .
$$ ------------------[/go]

White's region has the count -2 in the locale including the 1 black prisoner stone.

The count of Black's region in this position still needs to be calculated taking into account any later ko-atari on the stone 1.

If the intermediate white string were connected, the count of Black's region would be 2 2/3. Assuming the later ko-atari on the stone 1 is worth 1/6, I expect the count of Black's region to be 2 5/6.

Therefore, the count is 5/6.

If so and variation II dominates, the initial gote count is (5 + 5/6) / 2 = 2 11/12 and gote move value is (5 - 5/6) / 2 = 2 1/12.

White chooses the smaller count of variation I, which thus dominates. The move value is 2 1/6.

I have not proofread the calculations so there may still be mistakes.


EDIT: White has 1 point less in variation II. Correcting calculations accordingly.

[lichigo]

Thank you so much. Actually the picture wasn't clear. So I am porting this again but with the right situation.
And your explaination is wonderful. Thabk you so much.

[John Fairbairn]

I need to lie down after reading the above. And it still ends up as gibberish for me. If I do the calculations my own deiri based way, I very quickly establish that the basic gote-gote deiri value is 3, but White has the potential of an extra small gain with a follow-up move after his gote. So the !true" value is, in my terms, a little under 4.

If we convert that to a miai count we get "a little under 2".

So, I then have two questions:

(1) In practical terms what is the difference between 1.67 and "a little under 2". Or, more specifically, What is it about this difference that makes a too-difficult-for-me and error-prone (error even by a yose maven - see above) calculation worth attempting? The miniscule number of games in which it can make a genuine difference is probably more than outweighed by the number of times the calculations are miscalculated anyway.

(2) The REAL question, which everyone (pros and amateurs) seems to avoid answering, is still: why use one technique over another, especially when we keep getting told you can convert from to the other by dividing or multiplying by 2? Even if there is some obscure reason for saying one is superior to the other that can be demonstrated, it's not enough to say vague things like it helps with strategic choices. It may - but just saying just that is just like saying this margarine tastes like butter (and then wondering why most people and top chefs still prefer butter).