In the comment section of this gogameguru post Younggil An mentions that
and later that



On the other hand, I recall reading that Michael Redmond considered the endgames in the classical Castle games to be close to optimal. However I can't find a reference to this judgement (except some L19 threads without links, e.g. here). Here is an interview where Redmond mentions his study of the Castle games, but says nothing about the correctness.

Anyway, I found Younggil An's comments surprising, how about the rest of you? What would be an example of modern endgame technique? (I'm not talking about combinatorial game theory etc here --- because I don't think that is what Younggil An is talking about )

I guess classical go didn't have the pressure of time/ clock. How much time a match lasted in those days? Of course the reading in the middle game or endgame was so great.

_________________
si no puedes triunfar, muere gloriosamente

I have not heard An's opinion before. It may well be that modern pros have a better understanding of the endgame than pros of 50 or 100 years ago. OTOH, because of the length of games in the late 19th century, top players hardly ever lost a game because of endgame errors.

I have noticed that, in the timed games of the 20th century, even 9 dans often made errors when plays gained around 3 points (think of the hane-and-connect sente/reverse sente). And it is also true that, even if modern pros understand the endgame better than any group of pros in history, because of the short time limits they may not produce better endgames.

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

Visualize whirled peas.

Everything with love. Stay safe.

My interpretation of this is that there has been a translation mistake for yose. The references to the 'endgame' should be references to 'boundary plays', which is a very different thing.

Boundary plays start in the middle game, and even in the opening, and it was in the middle game that Yi Ch'ang-ho was famed for his so-called endgame ability.

Eventually someone will believe me when I keep pointing out that yose != endgame.

Edo players did make mistakes in yose, but they hardly ever made any in the endgame (typically the last 100 moves or so).

Also in those comments, An said this:

"In general, endgame is strongly related to reading, middle game and Life and Death, so if you study those parts, your endgame will get stronger too."

Amateurs often think of the endgame as being about making the biggest play. But as Takagawa said, The endgame is fighting strength.

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

Visualize whirled peas.

Everything with love. Stay safe.

Well, John, since Japanese amateurs also confuse yose with shuban, I think you have an uphill battle.

Merry Christmas!

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

Visualize whirled peas.

Everything with love. Stay safe.

Even allowing for various meanings of the term "endgame", I still can't make sense of An's comments: what is this part of the game that professionals of old didn't pay attention to compared to these days, and what would be a concrete example of skills and techniques in that phase of the game that has been developed in modern times? (Conversely, in the case of fuseki / joseki it is easy to give lots of examples.)

Translation was correct. Before changho lee, the strength of KOREAN PROFESSIONAL players were below current level. Many pros made mistakes at the end but was not noticed as much until lee. Korea didn't have any two day matches so many mistakes were common. Japanese top players were given sufficient time so no mistakes.

_________________
"The more we think we know about
The greater the unknown"
Words by neil peart, music by geddy lee and alex lifeson

Bill: You keep saying that and I keep letting it go, but it's wrong. All (almost) endgame plays are yose, but not all yose moves occur in the endgame.

Thus in the endgame (final moves) phase it is perfectly legit for them to talk about yose, which, don't forget, has a clear and relevant meaning (pulling together) for them in their ordinary language - it's not an out-and-out technical term for them as it is for us.

And when they see a reference to yose in the opening or middle game, you do NOT hear them cry, "Hey, I'm confused!"

To get back on my hobby horse, too many westerners think about yose in terms of numbers and counting, which is almost always possible only towards the end of the game. Japanese think of yose in terms of sealing off boundaries, which is possible in any phase of the game It is FAR more common in Japanese texts to talk about yose moves with terms such as 'thick' and ajikeshi than it is to say how big they are. And if they really do want to stress it is the endgame alone they are talking about, they have a very unconfusing term: shuuban.

Sennahoj: The modern feel for the game has benefited from a greater incidence of plays on the fourth line or higher than in Edo times, and they have learnt to seize territory-making opportunities earlier than in Edo times. Part of the technique of doing this is also learning to make thick plays in the centre earlier than their Edo counterparts. Knowing there is something to look for gives them a big edge. A Treasure Hunt with no clues is a lot harder than one with clues, and it's even harder if you are not even aware there is even a hunt going on.

It is clear that Edo games had the advantage of time and modern games have the advantage of greater knowledge. However, in both cases, research should be done to assess which endgame is better.

Non-numerical and numerical perception of boundary play are both important. If non-numerical perception cannot reveal an answer, numerical calculation is required to get the answer. Both also rely on reading.

Both Asian and Western endgame knowledge have improved. For most of us, it is hard to know in detail how Asian knowledge has been improved because too little is translated. The more theory is involved the more it seems that Western endgame understanding is running ahead.

Well, John, there are a couple of things in the background to clear up. First, there are prescriptive linguistics and descriptive linguistics. AFAICT, as far as go terms are concerned, we both fall in between pure prescriptive linguistics and pure descriptive linguistics, but my impression is that I am closer to descriptive linguistics than you are. Second, words, even technical terms, have more than one meaning, and speakers switch between them, usually effortlessly.

One example in go is ji in Japanese, territory in English. In each language the term has a specific technical meaning in the rules, which applies after play is over. But there is also a related, non-technical, fuzzy meaning which applies while the game is still going on. Thus, players can refer to (effective) plays inside the opponent's territory. In the technical meaning there can be no such effective play; there may be sente which do not alter the score, but that's it. Nobody except a rules lawyer would object to the non-technical meaning and insist that what people are calling territory actually is not.

There are also a number of meanings of sente and gote. There is a technical meaning of sente under which double sente which gain points do not exist. The failure to distinguish between different meanings of sente and gote has caused confusion and misunderstanding.

Now, as an amateur descriptive linguist I observed Japanese amateur go players use the term, yose, to refer to the endgame. They never used the term, shuban, at all. To the best of my recollection some of them and probably all of term used yose to refer to boundary plays before the endgame, as well as boundary plays in the endgame. They switched effortlessly between these two meanings, as human speakers do. I only learned the term, shuban, later on, through reading. It is not just I who have noted the meaning of the endgame of go for yose, Japanese lexicographers have, as well. It appears in Japanese dictionaries. To be sure, these are not lexicons of the technical terms of go, but I believe that they do reflect the non-technical language of Japanese amateur go players.

Now, I support you in spreading the word about the technical meaning of yose as boundary play, but I am too much of a descriptivist to support the idea that foreigners should observe niceties that native speakers do not.

Edit: When I said that Japanese amateurs confused yose with shuban, I did not mean that they were confused, I meant that they failed to make a distinction between the two.

Also, as far as multiple meanings of a word are concerned, this is more likely to occur and less likely to cause a problem when the two meanings apply to different kinds of things. That is the case with yose as a play and yose as a phase of the game. It makes perfect sense to refer to a yose (play) occurring before the yose (phase of the game). Speakers switch between the two meanings with ease. They do not get confused.

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

Visualize whirled peas.

Everything with love. Stay safe.

The word "endgame" is used in four meanings: 1) game phase, 2) all boundary plays, 3) a local endgame being a local part of the board, 4) a sequence of boundary plays (an endgame sequence). They can be can be discriminated as 'endgame phase', 'boundary plays', 'local endgame [position]' and 'endgame sequence', respectively, but usually language is lazy and speaks of 'endgame' in all meanings.

This is a bit funny...

"I don't know what's going on in Asia, but it seems Western endgame understanding is running ahead"

Certainly endgame in the west is better than where it was, but there's no reason to believe it's running ahead.

It's not like there is no East-West communication about the endgame.

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

Visualize whirled peas.

Everything with love. Stay safe.

oren, I can compare what I see in the literature, comments and reports about the hard to find literature. Result: Western endgame knowledge is running ahead of Asian endgame knowledge. (Note: knowledge, not endgame playing strength, because the knowledge is either too specialised / abstract or too little applied by Western players.)

On the topic of shuban versus yose, here's a modern example of the extremely high-level of shuban in today's go. It's game three of the 17th Tianyuan-Ch'eonweon Match.

Could someone explain 3.75 komi to me?

3.75 komi is a Chinese half counting komi equalling 7.5 full counting komi.

http://home.snafu.de/jasiek/int.html#Counting
http://home.snafu.de/jasiek/int.html#SpecialCommentsTR

John and logan, thanks for examples!

How can a Japanese-English translation issue be instrumental to understanding a comment by a Korean pro?
Do you think An Younggil speaks about "endgame" because he has no better word for it and should talk about "boundary plays" instead because he's really not talking about "the late stages of the game"?