Looking for a good book on the endgame

[singular]

I have 'The Endgame' by Ogawa and Davies but I find it obtuse. My mind doesn't grip what the authors are saying even after repeated reads. I always end up with more questions than answers for whatever reason. Compare this book to Attack and Defense which does a great pedagogical job of leading the reader from one point to the next-- thats the kind of book I'm looking for for the endgame.

Does there exist something like this? A book that introduces endgame methods gradually and provides exercises for them? Anticipates how the reader might get lost and pre-empts that lostness with patient explanation? I would like an Endgame for Babies book if it existed.

I've seen good reviews for 'Get Strong at the Endgame' but I'm concerned it's too advanced for my shonky 9k-ness.

[skydyr]

Well, what are the questions that it brings up? Perhaps we could provide clarity.

[hyperpape]

One aspect of the endgame is understanding how not to follow your opponent around. The best discussion of that that I have seen in Lessons in the Fundamentals of Go, by Kageyama.

As far as how to evaluate endgame sequences, that is generally very precise material, and will be challenging just like The Endgame.

[Bill Spight]

Attack and Defense . . . does a great pedagogical job of leading the reader from one point to the next-- thats the kind of book I'm looking for for the endgame.

Does there exist something like this? A book that introduces endgame methods gradually and provides exercises for them? Anticipates how the reader might get lost and pre-empts that lostness with patient explanation? I would like an Endgame for Babies book if it existed.

What questions do you have?

BTW, are you confident in dealing with damezumari at the end of the game? You might like Dame Filling for Dummies. ( http://www.lifein19x19.com/forum/viewto ... =15&t=9103 )

[bogiesan]

Janice Kim's last book in her series has a similarly obtuse section on the endgame. She even offers some nerdy calculation protocol you can run in your head for every branch of every move's tree to figure out if a half point is good for you or not. Impossibly dense stuff but she presents it with class and humor.

[Boidhre]

I'm enjoying this at the moment, most of the problems are "here is the normal endgame move, the answer is 4 points better" when showing you the correct answer, kind of stuff. Interesting and fun and good for finding blind spots in your reading.

200 Endgame Problems: http://www.slateandshell.com/SSSH001

I also got Lee Chang-ho’s Endgame Techniques on SmartGo but didn't read through it yet. From a casual glance it seemed to be a "from the ground up" approach to the yose.

[RobertJasiek]

The English endgame literature is still not as good as you seek. Better books need to be written. Exactly the kind of book you want does not exist yet, IMO. So what can you do? You need to bite through the existing literature. I suggest you to start with the simple books on the topic.

The IMO most useful start for your level is First Fundamentals, which discusses several topics incl. endgame principles for the opening and middle game, and an endgame chapter. For you, the greatest advantage is: the book does not confront you with values and counts of endgame moves or sequences yet. The disadvantage for you: the book has only a few problems.

http://home.snafu.de/jasiek/First_Fundamentals.html

So maybe you should get a second book with lots of easy endgame problems, but, AFAIK, the (English) endgame problem books rely on values and counts and are not as easy as you wish. So you have little chance but to bite your way through them anyway, if you want to practise lots of problems. (There are Asian books on endgame tesujis, but this is a too specialised topic to start learning the endgame at your level.)

How about a different approach? Postpone endgame values until you are 5 kyu. Before, study only value-less endgame literature.

[emeraldemon]

You might like Dame Filling for Dummies. ( http://www.lifein19x19.com/forum/viewto ... =15&t=9103 )

I would 100% buy this book.

This is a slightly different style, but have you seen Fujisawa's tesuji dictionary about the endgame?

http://senseis.xmp.net/?FujisawaTesujiDictionary

[Javaness2]

Have you lost your mind????
Seriously. Buy this book http://livre.fnac.com/a6910594/Dai-Junf ... -jeu-de-go now and destroy even the smallest idea of purchasing any other book.

[singular]

Thanks everyone for your insights and suggestions. I apologize if my response is a bit long.

Bill-- I am definitely not confident with dame-filling! I didn't even think of that as an endgame issue to apprehend at this time, but I now realize dame-filling has always been one of those things at which I am crap. Going over the problem you posted it is clear to me that I would benefit greatly from some kind of dame-filling literature. Is there any? The usual shortage-of-liberties stuff in problem books is not like the kind of problem you posted, a whole board problem with many groups in close contact and a variety of weaknesses. The problem you posted is directly applicable to real game situations and how to sequence moves. It was a bit hard for me (I got it kinda badly wrong), but I enjoyed it and would like to get to the point where I could take a problem like that in stride. I would be a much, much better Go player. Any plans to write a book?

Boidhre-- I wondered about that book, and now looking at the sample pages I learned something about the endgame already. It looks like I will get it, thanks!

Robert-- Thanks for the recommendation and advice. I am however looking for a way to get into counting and evaluating positions, and I know it won't be easy (as hyperpape pointed out), but there's got to be a better way than in The Endgame (which is inscrutable). I will look at First Fundamentals for the reasons you outlined.

As far as the endgame is concerned I think the main thing I am after with the literature is to have plenty of easy practise to supplement the teaching. Easy in order to bang home the lessons, but also becoming progressively harder and applicable to real-game situations.

This is what I'd love to see, my pie-in-the-sky wishlist:

Practise:

> Practise calculating values, starting from ultra-basic positions. Overlearning trivial counts and moving on from there.

> Practise dame-filling! This is a new thing for me!

> Practise tesuji.

> Ultimately, practise sequencing whole-board endgame.


Lessons that answer various questions:

> How to calculate value? Some areas are more obvious than others . . . . they could be put into categories or something.

> How to weigh value and sente (play a big move or a sente move?)

> Damezumari: how to put the mind in order about dame-filling. It can be confusing for me with so many groups in contact and so many different kinds of weaknesses to cover. Some guidelines or principles would help. I always get this wrong and pay for it.

> Semedori and tiny ko. When are they worth something and what are they worth?

> Tedomari-- when does endgame start? And at the end of endgame is the final move worth anything? Why?

> Tesuji-- how to avoid making shapes vulnerable to tesuji, and how to exploit vulnerable shapes.

> Which moves are endgame moves? (Not always obvious).


I appreciate that I (and others at my level and below) should ignore endgame for now, but I feel drawn to it. Maybe because I play on KGS alot and most of those games tend to be even. Endgame usually decides the winner. I also think many of my other weaknesses could be addressed by getting better at endgame. Also, I'm just curious, and I hope others will benefit from the topic.

[singular]

Javaness-- Thanks for the recommendation. Unfortunately my French is worse than my Go, and I can't see any sample pages

[Bill Spight]

Bill-- I am definitely not confident with dame-filling! I didn't even think of that as an endgame issue to apprehend at this time, but I now realize dame-filling has always been one of those things at which I am crap. Going over the problem you posted it is clear to me that I would benefit greatly from some kind of dame-filling literature. Is there any? The usual shortage-of-liberties stuff in problem books is not like the kind of problem you posted, a whole board problem with many groups in close contact and a variety of weaknesses. The problem you posted is directly applicable to real game situations and how to sequence moves. It was a bit hard for me (I got it kinda badly wrong), but I enjoyed it and would like to get to the point where I could take a problem like that in stride. I would be a much, much better Go player. Any plans to write a book?

I am unaware of any dame filling literature to speak of. But I remember when I was a 5 kyu and first began studying pro games, I was often unable to fill the dame correctly from the final game diagram so that I got the correct score. It was actually pretty good practice to try to figure out where protective plays were necessary and why. Go programmer Erik van der Werf claims that players around 5 kyu get the score right around 98% of the time. That's actually quite good. But, OC, they are unaware of their errors in dame filling. Dan players are near perfect, but there are some very difficult dame filling positions out there.

Some years ago I thought about writing a yose book aimed at 10 kyu players. So I went online and took a look at 10 kyu vs. 10 kyu game records. In most games there were opportunities for significant swings near the end of the game, because of damezumari. A lot of yose problems rely on damezumari, but they are usually A level problems or B level problems, not C level problems (using Segoe's categories). I had thought of teaching the basics of endgame calculations, but what good does it do to be able to tell the difference between a play that gains 1 pt. and one that gains 1/2 pt., when a 5 point or 10 point swing is staring you in the face?

What to do at your level?

First, always play the dame out. How can you improve your dame filling when you don't fill the dame? It does not take long to do, unless somebody makes a mistake. Your opponents often will. So will you. Look for dangers, look for opportunities. Your game will improve.

Second, study the end of high level games that were played out. Go to the dame stage and then see if you can predict the final score correctly. If not, look for the protective play or plays that you missed. Very important: Do not be afraid to try moves out on the board. You should try to develop your reading, but you have already gotten reading practice when you tried to predict the final score, and when you went back and looked again for threats and protective plays. Your aim now is to discover what you are not seeing.

Third, study your own games for mistakes and opportunities at or near the dame stage. You can't be sure of the score, but your game will improve. Imagine entering the endgame 5 points behind, but confident that you will have good chances to win the game.

Fourth, do problems: semeai, yose tesuji, and life and death.

Once you are aware of the dangers and opportunities in the late endgame, you will be at least a few stones stronger, and then you can worry about endgame calculation.

Good luck!

And thanks for your questions. I am very busy this year, but next year I do plan to do a book.

[Boidhre]

I am very busy this year, but next year I do plan to do a book.

[skydyr]

> How to calculate value? Some areas are more obvious than others . . . . they could be put into categories or something.

> How to weigh value and sente (play a big move or a sente move?)

> Damezumari: how to put the mind in order about dame-filling. It can be confusing for me with so many groups in contact and so many different kinds of weaknesses to cover. Some guidelines or principles would help. I always get this wrong and pay for it.

> Semedori and tiny ko. When are they worth something and what are they worth?

> Tedomari-- when does endgame start? And at the end of endgame is the final move worth anything? Why?

> Tesuji-- how to avoid making shapes vulnerable to tesuji, and how to exploit vulnerable shapes.

> Which moves are endgame moves? (Not always obvious).

So to tackle a few of your questions, at least, an endgame move is a move that is only worth points. The contrast is with moves in the middle game, which concern the life and death of groups. Once all the groups' statuses are decided, it's time for the endgame. There are, of course, sometimes very large endgame moves that should be taken before the last middle game moves, but as for classifying a move as one or the other, the question is whether a group is unsettled or of uncertain status and can be attacked, in which case it's up to you to attack or defend as appropriate, or move on to endgame if it is not true.

Regarding Tedomari, as mentioned, the endgame starts when all the groups statuses are decided. This doesn't mean that you don't have forcing moves against those groups, but just that with proper play, they won't die. It is also possible for a game to move into the endgame and then, as a result of a cut or somesuch, go back to the middle game, because there are undecided groups. Tedomari itself is taking (and by extension, arranging to take) the last move of a particular value before all the moves drop significantly in value. This requires counting, or at least a sense of the size of moves, and is probably best explored through problems. One thing that's related though you didn't mention it is miai. If there are two moves left at a certain value, make sure you get one of them! Especially if both are double-sente, as giving your opponent both is a huge loss.

As for Semedori and dame filling, these also need to be read out, but generally, if you have two eyes on one side of a capture, you don't need to capture, and if you have one, you eventually will if your opponent can fill all the other liberties. If this is the case, then you need to count how many liberties you have, and how many the dead stones have, and start taking liberties when you are one away from losing the capturing race. Dame filling is similar, expect that if you have one eye, plus various loose connections and false eyes, you are generally okay, and if you don't, you need to read out how many liberties you have with the last string, and how many moves it will take you to confirm your connection, and then start connecting one liberty before you lose. You generally also want to connect from the living group to the string, rather than the other way around, so that each move your opponent makes threatens to capture less and less stones. Again, problems help here.

Regarding sente, gote, and value:

Traditionally, moves are grouped into 3 or 4 categories: double sente, sente, reverse sente, and gote.

First, a move is sente because the follow up is significantly larger than all the other moves available on the board. Sometimes this means that it threatens to kill a group, sometimes not. Also, once played, the other player may be able to intersperse a sente move with a bigger followup prior to responding to your threat.

Double sente are moves that are sente for both sides. Play these first, as soon as the threat is big enough.

Sente moves are sente for you, but if your opponent plays there first, it's gote for them. Generally these are your privelege to play, and to value them, you count the difference in territory and prisoners from when you play first and when your opponent plays first.

Reverse sente moves are the opposite, sente for your opponent, gote for you. You count them the same as sente moves, but generally, your opponent will take them first, since by taking it you give up sente.

Gote moves are gote for both players. Traditionally, these are counted the same way as sente moves, and then the value is halved for the purpose of deciding when to play it. They are the least important moves to play, since neither player is in a rush to take or prevent them, and they give up sente.

The counting method I am describing here is known as deiri counting, if you want to look it up, and there is a slightly confusing page on it at sensei's library: http://senseis.xmp.net/?DeiriCounting. This may be worth looking at for the first example, at least, to get a better idea of how the count works. You may also get good value out of looking at the endgame section there: http://senseis.xmp.net/?Endgame. There are several exercises with answers and other resources that may help you bridge the gap between where you are and the written literature.

Hopefully this helps a bit, at least.

[skydyr]

Javaness,
Do you know of a source for that book that doesn't include 15+ euro shipping to the US?