More rumours on endgame books

[RobertJasiek]

Reference:
http://www.lifein19x19.com/forum/viewto ... 84#p128284

I'm personally waiting for Robert Jasiek's endgame book with some good expectations, but never know if he'll write them.

An endgame book is in my pipeline of books I want to write during the following years. Study for three of my earlier books has provided me with enough conceptual background for a thorough endgame book. While you await my endgame book, I await Bill Spight's:) I suppose his will be rather numerical, explaining mathematics of fractions, infinitesemals and temparatures. I intend a presumably more pragmatic approach with an accuracy up to about 1/3 of a point. There is so much to write and learn about the endgame that IMO players must not be overloaded with too complicated mathematics. Bill is the expert for the CGT. I see my aim more in making the practically relevant game tree decisions and mathematics available for everybody's understanding. If I can do that, your expectations would be justified. However, writing about a theory-heavy topic is demanding. Despite my own optimism, I will need months to write such a book, so that it can meet, e.g., your expectations. In the meantime, maybe Bill surprises us with achieving endgame explanation for everybody's easy understanding.

One thing is sure: the current English language endgame literature, although useful, does by far not convey everything needed to become a dan WRT to endgame knowledge. As SL and various news / forum messages indicate, quite some additional knowledge is floating around and awaits a refined, comprehensive explanation. This is one of the strong motivations for me to write an endgame book. However, you are not alone in not knowing when I will actually write it:)

[karaklis]

I intend a presumably more pragmatic approach with an accuracy up to about 1/3 of a point. There is so much to write and learn about the endgame that IMO players must not be overloaded with too complicated mathematics.

These statements obviously contradict. I find that counting with fractions does not help much. At the end only whole points are counted.

[quantumf]

I intend a presumably more pragmatic approach with an accuracy up to about 1/3 of a point. There is so much to write and learn about the endgame that IMO players must not be overloaded with too complicated mathematics.

These statements obviously contradict. I find that counting with fractions does not help much. At the end only whole points are counted.

And yet, they do matter. A move worth 2/3 is generally better than one worth 1/3, and moves worth these values are completely standard and occur in many games. If fractions are distressing, then there are perhaps alternatives (although I'm not aware of any similarly succinct ones) but somehow you have to know that this small move is worth more than that small move. Perhaps Robert will address this numeric discomfort.

[RobertJasiek]

Every endgame book above beginner level should teach the difference between 1 point endgames, 1/2 point endgames and 1/3 point kos. In fact, these are already essential for good positional judgement during opening and middle game. However, endgame books can dispense with distinguishing 7/8mini from 7/8up of a point. (Except that it would still be interesting to read about such from Bill.) During a game, it is not worth spending minutes on finding the bigger of the two; the thinking time is better spent for avoiding all confusions between 1 and 1/2 point moves.

[karaklis]

I remember that a German 5d (forgot his name) had written a go book long time ago that also included a chapter about the endgame. I liked this chapter because the author dismissed the usage of such fractions. When using fractions you assume that e.g. one point can be gained in three moves, and then you call it 1/3 point. But at the end it is one point or none, depending on whether the opponent responds to your move or not. At the end Schrödinger's cat is dead or alive and not in a state between.

[jts]

I remember that a German 5d (forgot his name) had written a go book long time ago that also included a chapter about the endgame. I liked this chapter because the author dismissed the usage of such fractions. When using fractions you assume that e.g. one point can be gained in three moves, and then you call it 1/3 point. But at the end it is one point or none, depending on whether the opponent responds to your move or not. At the end Schrödinger's cat is dead or alive and not in a state between.

That's not really the point of the thing. If you can read out the endgame to the last play, more power to you; but for the rest of us, it's easier to know which plays are biggest in fractional counting and play the biggest point. If there are a few points of difference between best and worst play, we will get those points.

[aokun]

I look forward to more endgame books, Bill's or Robert's. I will buy them. It is the weakest part of the literature in English and it makes a difference to one's strength. Beyond that, it makes a difference to the enjoyment of the game. A proper game of go, unless it ends early, is a protracted struggle and in the struggle comes much of its satisfaction. A lot of us in the West fight vigorously till all the major questions are decided and then, lacking the tools to keep at it, wander vaguely through the process of cleaning up the boundaries with little understanding of the opportunities. Then we count up wondering till the last moment who won.

At least two books are called for, the complete theory based in CGT and the practical playing guide.

But at the end it is one point or none, depending on whether the opponent responds to your move or not. At the end Schrödinger's cat is dead or alive and not in a state between.

And yet QM is correct. Yes, a measurement produces an allowed answer not a state in between, but the states in between are the central feature of QM and its usefulness as a theory, not the fate of the cat, is the main thing. CGT and QM are not the only fields revolutionized by mixed answers. Game theory's central result involves equilibria in mixed strategies, even though "at the end" you only get to pick one strategy. Sometimes the correct choice is to turn your choice over to a deck of cards. I don't know much about fuzzy logic, but gather than decision rules and deductions can be made better by modelling binary things with uncertainty.

I think fractions, correctly calculated, belong in go to the degree it is practical for us to calculate them.

[RobertJasiek]

a German 5d (forgot his name) had written a go book long time ago

Maybe Karl-Friedrich Lenz? He wrote a paper, in which he advocates usage of integers. (Just for the history. I am not convinced because we cannot always read to the end of a game.)

[deja]

Robert, just wanted to add my encouragement/support for you to go forward with your endgame book, and to extend my thanks for what you've already published. I don't have all of your books yet but am planning to complete the set fairly soon.

[karaklis]

Maybe Karl-Friedrich Lenz? He wrote a paper, in which he advocates usage of integers. (Just for the history. I am not convinced because we cannot always read to the end of a game.)

Right, that was the name.

I am beginning to understand. The usage of fractions is actually a shortcut to reading because you don't know whether your opponent responds or not. Knowing the common several hundreds of endgame shapes and their values would probably be very useful for your endgame play. These can be practised just like you can practise L&D or tesuji. Interesting. I'll be looking forward to the coming book then.

[RobertJasiek]

Knowing the common several hundreds of endgame shapes and their values would probably be very useful for your endgame play.

Yes. Ugh. I was planning to include a few dozens of regularly occurring shapes (such as 6 move josekis) with values and already frightened of having to do their reasonably accurate calculations. Several hundreds? Nice dream, but this alone would fill another book. I plan to write about endgame theory (and its application to examples). For a pure shape / value collection, the following is a good start:

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

[John Fairbairn]

Knowing the common several hundreds of endgame shapes and their values would probably be very useful for your endgame play. These can be practised just like you can practise L&D or tesuji. Interesting. I'll be looking forward to the coming book then.

A book on this by a Meijin, Genan Inseki, exists. It's on the GoGoD CD. Rob van Zeijst took the business a step further and memorised reportedly about 1000 positions and their values when he was trying to become a pro. Pros have confirmed to me that memorising lots of positions as per Inseki/Rob is normal, but I don't know how many is normal.

[LokBuddha]

where in GoGod cd is Inseki endgame manual?

[John Fairbairn]

where in GoGod cd is Inseki endgame manual?

It's in the New In Go archive, which of course means it's also still online (item No. 21).

Sorry to be technologically snobbish by offering it in an old-school format there, but New In Go is free and with a ridiculously high content-to-noise ratio (T Mark - does that sound technological enough?).