On "Practical endgame counting: what's happening..."

[RobertJasiek]

Citation reference: viewtopic.php?p=252610#p252610

The book "the Endgame" [...]
this text is the underestimated shortcut to amateur dan-level play
Link: https://twitter.com/zbaduk/status/1002439000160907264

Dan level play does not only consist of endgame. With respect to endgame skill, the book The Endgame has raised my endgame skill from 8k to 5k level of endgame skill. Beyond that, its limitations blocked my further improvement of endgame skill, when I was 5k to 4d (overall strength). In particular, it kept my reverse sente weak. Therefore, I do not at all share your experience "shortcut to amateur dan-level play".

In your link, you wrote in June 2018 that The Endgame would have been the only English book explaining endgame values. This was wrong at that time because Mathematical Go Endgames appeared in 1994 (but, for direct application of theory, is rather useless during games because calculating infinitesemals is too slow, even if one understands them). Endgame 2 Values appeared only one month later in July 2018:)

Since you still refer to the link, the list of English books explaining endgame values should be updated, sorted in increasing order by how much they explain about the theory of endgame values:

- The Endgame
- Mathematical Go Endgames
- Rational Endgame
- Endgame Problems 1
- Endgame 3 Accurate Local Evaluation
- Endgame 2 Values

[xela]

For completeness, here are some more English books that mention endgame values (although in some cases, "explain" may not be the right word).

- The Theory and Practice of GO has a chapter on "End Plays". Interestingly, he uses the word "Upperhand" (always capitalised) instead of "sente". I don't recommend reading this book as a way to improve your go skill; it's just of historical interest.
- Basic Techniques of Go has a chapter on the endgame. There's a brief explanation of counting.
- 200 Endgame Problems is mostly about endgame tesuji; I don't understand why the author decided to include a short section on counting!
- Lessons in the Fundamentals of Go has a chapter on "endgame pointers" which includes some hints on how to count.
- Get Strong at the Endgame treats how to count as assumed knowledge. In principle you should be able to deduce the method from the problems and solutions, although the number of mistakes would make this challenging.
- Lee Chang-Ho's Endgame Techniques: I haven't read this, but I believe volume 1 has a chapter on counting.

[Bill Spight]

Sadly, as I discovered many years ago, Japanese endgame texts had calculation errors on almost every page. English texts had miscaculations, as well. Often this was the result of miscategorization, guessing wrongly whether a play was sente or gote, or calculating double sente. (If you realize that the count comes first, it is obvious that double sente is incoherent, since a gainful double sente does not have a count. Davies, I believe, pointed out that double sente implies division by 0.) Among recent texts, in Japanese O Meien gets the calculations right, and in English Jasiek does. Tormanen probably does, as well, but I have not read his book.

[RobertJasiek]

Japanese endgame texts had calculation errors on almost every page. [...] Often this was the result of miscategorization, guessing wrongly whether a play was sente or gote, or calculating double sente.

Previously, you mentioned that Japanese endgame texts also made mistakes as to how long sequences should be played successively. Was that an occasional or frequent mistake in them?

[dfan]

Lee Chang-Ho's Endgame Techniques: I haven't read this, but I believe volume 1 has a chapter on counting.

Yes, Volume 1 is mostly about counting. It's quite good, with many examples explained clearly and patiently. It uses deiri counting, for those who care.

[Knotwilg]

The title of this thread makes it a duplicate of the thread it refers to.

You may want this thread to be about "which books and publications are there on endgame theory"

[Bill Spight]

Japanese endgame texts had calculation errors on almost every page. [...] Often this was the result of miscategorization, guessing wrongly whether a play was sente or gote, or calculating double sente.

Previously, you mentioned that Japanese endgame texts also made mistakes as to how long sequences should be played successively. Was that an occasional or frequent mistake in them?

That's a general error, based, I believe, on the fact that humans are best a depth first search. It is not confined to endgame texts.

In the context of practical counting during play with limited time, however, despite the analytical errors caused by guessing whether a play or sequence of play is sente or gote and by reading deeply, I recommend both precisely because they play to human strengths.

OC, I also recommend having a broad scope, for instance, with my saying, Tenuki is always an option, and my heuristic of comparing two plays by considering them to be miai (roughly), asking what happens if your opponent replies to one play by playing at or near the other point. OC, neither of these pertains to local counting.

[RobertJasiek]

The title of this thread makes it a duplicate of the thread it refers to.

Discussion of books only in the books forums is a consequence of the forum rules. For details, ask via email or rec.games.go using your real name.

[RobertJasiek]

Citation reference: viewtopic.php?p=252625#p252625

As a minimum, I expect a theory book on the endgame to explain the difference between the following two concepts:

A theory book, or a series / collection of theory books! If some concept is explained in one volume (or another book discussing a topic more closely related to a particular concept), it need not always be explained in all volumes.

When a local sente is answered by another local sente elsewhere (and subsequently both answer the local sente, or continue following up), we can speak of the process of mutual damage.

E.g., mutual reduction is also a topic of middle game strategy or positional judgement. For the late endgame, there are alternative concepts.

For books about mutual reduction, see Positional Judgement 1 Territory, Fighting Fundamentals, Positional Judgement / High Speed (Cho), Endgame 1 Fundamentals and others.

[...]
- a move which, when unanswered, leaves a follow-up move with a higher value than the original move
- a move which requires an immediate answer (i.e. playing elsewhere, even another local sente, does not offset the damage done by the follow-up) [...]
There may be multiple local sente but at most one global sente.

Others have already pointed out problems with your attempted definitions and the possibility of several equally good best moves.

Let's call the first "local sente" and the second "global sente".
[...]
If it is correct, then let me know which books have this notion.

Endgame 2 Values explains both concepts. Endgame 3 Accurate Local Evaluation, Endgame Problems 1 and planned later books of these series consistently and frequently speak of "local gote / local sente / ambiguous" or more specific types (such as "Black's long sente" or "double long gote") throughout the books, explain and define the terms and characterise them by up to five alternative value conditions.

I would benefit more from clear concepts than from correct calculations. The latter follow from the former anyhow and are merely a matter of precision than clarity.

Right, for learning the concepts. However, acquiring the skill of fast calculation also needs much practice of performing calculations.


EDIT: add books for mutual reduction.