Page 1 of 1

Mathematical Go Endgames

Posted: Fri May 24, 2013 6:08 am
by hyperpape
I see that this book now has a new subtitle. Does anyone know for sure if the change is just cosmetic?

Re: Mathematical Go Endgames

Posted: Fri May 24, 2013 6:36 am
by daal
If anyone is interested in this sort of stuff, follow the link and you can read a huge part of the book (only pages 47-145 are not shown). I guess this makes sense. Anyone still reading by page 45 will certainly buy it. :)

Re: Mathematical Go Endgames

Posted: Fri May 24, 2013 6:43 am
by Bill Spight
hyperpape wrote:I see that this book now has a new subtitle. Does anyone know for sure if the change is just cosmetic?
The copyright belongs to A. K. Peters, the original publisher, dated 1994, and the new publisher is Ishi Press in NYC. I am reasonably certain that the content is the same as the original. (Peters retired.) Besides Berlekamp has said nothing to me about a new edition.

Re: Mathematical Go Endgames

Posted: Fri May 24, 2013 8:46 am
by Thunkd
I bought a lot of go books that included this book. After a brief glance through it, it didn't seem like anything that was particularly helpful for me. Was that a too cursory decision? I had planned to try to sell this, but if someone wants to make a case that I've given the book short shrift I could be persuaded to reconsider my decision.

Re: Mathematical Go Endgames

Posted: Fri May 24, 2013 9:54 am
by Bill Spight
Thunkd wrote:I bought a lot of go books that included this book. After a brief glance through it, it didn't seem like anything that was particularly helpful for me. Was that a too cursory decision? I had planned to try to sell this, but if someone wants to make a case that I've given the book short shrift I could be persuaded to reconsider my decision.
It is a good book, but not for everybody. There are concepts that, IMO, every pro should know, but few do. As a practical matter, it will add about 2 pts. to your game, not just from the material, but because it generalizes. That's not much to us, but it's about one half a pro rank. :)

Re: Mathematical Go Endgames

Posted: Fri May 24, 2013 10:38 am
by gowan
Bill Spight wrote:
hyperpape wrote:I see that this book now has a new subtitle. Does anyone know for sure if the change is just cosmetic?
The copyright belongs to A. K. Peters, the original publisher, dated 1994, and the new publisher is Ishi Press in NYC. I am reasonably certain that the content is the same as the original. (Peters retired.) Besides Berlekamp has said nothing to me about a new edition.
Ishi Press still exists? I thought it was defunct after the mess that led to the creation of Kiseido.

Re: Mathematical Go Endgames

Posted: Fri May 24, 2013 11:28 am
by Boidhre
gowan wrote:
Bill Spight wrote:
hyperpape wrote:I see that this book now has a new subtitle. Does anyone know for sure if the change is just cosmetic?
The copyright belongs to A. K. Peters, the original publisher, dated 1994, and the new publisher is Ishi Press in NYC. I am reasonably certain that the content is the same as the original. (Peters retired.) Besides Berlekamp has said nothing to me about a new edition.
Ishi Press still exists? I thought it was defunct after the mess that led to the creation of Kiseido.
It seems to publish Chess stuff now: http://www.newinchess.com/Complete_Ches ... -2996.html

Wikipedia says it was restructured and aimed at the chess market after the mess.

Re: Mathematical Go Endgames

Posted: Fri May 24, 2013 11:57 am
by hyperpape
Bill Spight wrote:
Thunkd wrote:I bought a lot of go books that included this book. After a brief glance through it, it didn't seem like anything that was particularly helpful for me. Was that a too cursory decision? I had planned to try to sell this, but if someone wants to make a case that I've given the book short shrift I could be persuaded to reconsider my decision.
It is a good book, but not for everybody. There are concepts that, IMO, every pro should know, but few do. As a practical matter, it will add about 2 pts. to your game, not just from the material, but because it generalizes. That's not much to us, but it's about one half a pro rank. :)
What do you base the 2 points on? Are pros typically leaving that much on the table in endgame?

Re: Mathematical Go Endgames

Posted: Fri May 24, 2013 2:27 pm
by Bill Spight
hyperpape wrote:
Bill Spight wrote:
Thunkd wrote:I bought a lot of go books that included this book. After a brief glance through it, it didn't seem like anything that was particularly helpful for me. Was that a too cursory decision? I had planned to try to sell this, but if someone wants to make a case that I've given the book short shrift I could be persuaded to reconsider my decision.
It is a good book, but not for everybody. There are concepts that, IMO, every pro should know, but few do. As a practical matter, it will add about 2 pts. to your game, not just from the material, but because it generalizes. That's not much to us, but it's about one half a pro rank. :)
What do you base the 2 points on? Are pros typically leaving that much on the table in endgame?
If you look at pro games in the 19th century, with no time limits, better pros (5 dan and up) hardly ever dropped a point in the endgame. If you look at games in the 20th century, with time limits, but limits that are generous by today's standards, top pros hardly ever dropped a point in the late endgame. However, they frequently made inaccuracies that did not drop a point, but could have. E. g., a move loses on average 1/8 of a point, but 7/8 of the time will lose nothing. Where you start to see errors at that time in top play is when plays gain 3 pts. or more. In these days of quick pro games, I shudder to think. ;)

I am basing my estimate on generalization and the fact that getting the ideas down can save a lot of time. (If not for generalization, the book is worth less than one point. ;))

Re: Mathematical Go Endgames

Posted: Fri May 24, 2013 5:20 pm
by jts
Are you suggesting that the average player loses 2 points per game because he doesn't know any of Berlekamp's techniques or that the theoretical maximum difference between Berlekamp's techniques and standard rules of thumb is 2 points?

Posted: Sat May 25, 2013 3:12 am
by EdLee
gowan wrote:Ishi Press still exists? I thought it was defunct after the mess that led to the creation of Kiseido.
What was the mess ? (before my time; curious.)

Re: Mathematical Go Endgames

Posted: Sat May 25, 2013 4:47 am
by Bill Spight
jts wrote:Are you suggesting that the average player loses 2 points per game because he doesn't know any of Berlekamp's techniques or that the theoretical maximum difference between Berlekamp's techniques and standard rules of thumb is 2 points?
I was thinking, not of the average player, but of the player who might be thinking of studying the book. And I would hardly apply the term theoretical to a guesstimate. ;)

Re: Mathematical Go Endgames

Posted: Sat May 25, 2013 4:58 am
by Bill Spight
gowan wrote:
Bill Spight wrote:
hyperpape wrote:I see that this book now has a new subtitle. Does anyone know for sure if the change is just cosmetic?
The copyright belongs to A. K. Peters, the original publisher, dated 1994, and the new publisher is Ishi Press in NYC. I am reasonably certain that the content is the same as the original. (Peters retired.) Besides Berlekamp has said nothing to me about a new edition.
Ishi Press still exists? I thought it was defunct after the mess that led to the creation of Kiseido.
Well, it is Ishi Press International. I guess there is a distinction between that and Ishi Press. It is certainly not the Ishi Press of old.