Page 1 of 1

Endgame problem question

Posted: Fri Feb 07, 2014 11:48 am
by happysocks

I chose to save the 3 black stones instead of capturing the 6 white stones which is the correct answer but then found myself wondering about how to calculate the point swing. The capture is easy = 12 points, but saving the the 3 black felt bigger because black can then further reduce white's area and this is where things become foggy for me because there seems to be so many different ways for both black and white to play here. Black get the first move in whites area after white saves it's 6 six stones so it seems the area will be reduced to almost nothing but is there a way to figure out the end result precisely?

Re: Endgame problem question

Posted: Fri Feb 07, 2014 12:41 pm
by happysocks
Unless I'm missing some white options it looks like the 20 points that would be whites if black chose not to save the 3 stones will be at best 1 point if black does save them with best reducing play?

Re: Endgame problem question

Posted: Fri Feb 07, 2014 1:06 pm
by Bill Spight
Do you know how to calculate the count (expected net score) of an independent region of the board?

Re: Endgame problem question

Posted: Fri Feb 07, 2014 1:53 pm
by happysocks
Think so but "expected net score" is where things become hazy sometimes (admittedly I have a tendency to see things more complicated then they actually are but that may be somewhat common for beginners?). Anyway post 2 was relating an attempt to do so (I thought).

Re: Endgame problem question

Posted: Fri Feb 07, 2014 2:30 pm
by Bill Spight
happysocks wrote:Think so but "expected net score" is where things become hazy sometimes (admittedly I have a tendency to see things more complicated then they actually are but that may be somewhat common for beginners?). Anyway post 2 was relating an attempt to do so (I thought).
By net score I mean the difference between the score for Black and the score for White. By expected net score I mean that we calculate it by assuming that if the best local play is gote, then the odds are 50:50 whether Black plays first or White plays first, and if the best local play is sente, then the sente is played and answered.

As for your post 2, I think that you have made a reasonable estimate for play in a game. Precise calculation takes time. :)

Re: Endgame problem question

Posted: Fri Feb 07, 2014 5:46 pm
by happysocks
Mmm, got it. Thanks.

"Precise calculation takes time." Am perhaps gaining a little better appreciation of this equation. :mrgreen:

Re: Endgame problem question

Posted: Fri Feb 07, 2014 11:05 pm
by Bill Spight
Well, I think I have best play in the problem position. :)

Here are some variations with my calculated gains.



Does my recommended line agree with the official answer?

Re: Endgame problem question

Posted: Sat Feb 08, 2014 3:18 pm
by happysocks
This is from SmartGo's collection of problems (# 1714 in the 12kyu problems) and all it say's about black 1. Q6 is "solved" (black 1. O8 "Black can do better") but is quite interesting to see the breakdown of variations in the lines you illustrated. Thanks again. :)