Life In 19x19
http://www.lifein19x19.com/

Basic endgame problem -- feedback requested
http://www.lifein19x19.com/viewtopic.php?f=15&t=3056		 Page 1 of 4
Author:  Bill Spight [ Sun Jan 30, 2011 11:38 am ]
Post subject:  Basic endgame problem -- feedback requested
Here is a 13x13 problem that will be simple to some people. My intended solver is around 10 kyu. I would be interested in what you think of the problem, and what level you think it is. Many thanks. :)

Click Here To Show Diagram Code
[go]$$W White to play. Japanese rules. 6.5 komi.
$$ ---------------------------
$$ | . . X O . . . O O . . . . |
$$ | X . X O O O O . O . X X . |
$$ | . X O . . . . O O O X O . |
$$ | . X X X X . X . X O . X . |
$$ | O O O O X . X . X X . X . |
$$ | . . O X . X . X O X . X . |
$$ | . . O X . X . X O X X X . |
$$ | . O O . O X . X O . O X . |
$$ | . X O O X X X O O X . X . |
$$ | . O . O O X X . O X X X O |
$$ | . . O . . . X O O O O X X |
$$ | . . O O X X X O . O O X . |
$$ | . . O . O O X O O . O . . |
$$ ---------------------------[/go]


(Japanese rules, 6.5 komi.)

White to play and win. Show best play for both. (Several options are equivalent.)

Enjoy!
Author:  tj86430 [ Sun Jan 30, 2011 12:01 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
It seems fairly easy, but I'm not sure until i see what the correct answer is :cool:

Probably just fine for 10k.
Author:  Bill Spight [ Sun Jan 30, 2011 12:05 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
It's OK to solve it, too. ;)

(But please hide your solution. :))
Author:  freegame [ Sun Jan 30, 2011 12:19 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
I made attempt.
I did not really calculate the moves,
just (quickly) played out the game


* edit there is a white stone missing at K2. giving white 29.5


will take a 10 kyu quite some time to solve correctly I think.
Author:  tj86430 [ Sun Jan 30, 2011 12:32 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
freegame wrote:
I made attempt.
I did not really calculate the moves,
just (quickly) played out the game

So is the problem to find the next move for white or to play out the entire game? If the latter, then I think it's a big effort for 10k, although doable.

(If the former, then at least my move was the same as yours)
Author:  Bill Spight [ Sun Jan 30, 2011 1:01 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
freegame wrote:
I made attempt.
I did not really calculate the moves,
just (quickly) played out the game



will take a 10 kyu quite some time to solve correctly I think.


Typo: You left out the White stone on K-02, which White needs to live. :)

:b2: is inaccurate. (It would be a mistake, but Black cannot win with correct play, anyway. :))
Author:  illluck [ Sun Jan 30, 2011 1:05 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
I think this is too hard for 10k. I cannot solve this with certainty XD
Author:  Bill Spight [ Sun Jan 30, 2011 1:07 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
tj86430 wrote:
freegame wrote:
I made attempt.
I did not really calculate the moves,
just (quickly) played out the game

So is the problem to find the next move for white or to play out the entire game? If the latter, then I think it's a big effort for 10k, although doable.

(If the former, then at least my move was the same as yours)


Yes, the idea is to play out the rest of the game optimally for both sides. I have edited the problem to make that clear. Thanks. :)
Author:  jts [ Sun Jan 30, 2011 1:10 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
In your diagram, freegame, :b2: should definitely be at

K2


...but only if :w1: doesn't get there first.
Author:  freegame [ Sun Jan 30, 2011 1:20 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
jts wrote:
In your diagram, freegame, :b2: should definitely be at

K2


...but only if :w1: doesn't get there first.


I missed a white stone at k2 when coppying the dagra.
thanks for pointing that out. :tmbup: (put a note in my previous post)
Author:  jts [ Sun Jan 30, 2011 1:43 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
More seriously,
D1 is 2 points (+2 for the capture, up to +3 if B gets E1 first, down to +1 if W gets E1 first) and K6 is 3 points. Since W then gets the first 2-pt move he should play any of them except for D1 - when B finally plays D1, W gets the last 2 pt move at E1. Then I assume B plays L10 and everything else as freegame shows. So W wins by 0.5?
Author:  Mnemonic [ Sun Jan 30, 2011 3:03 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
Well, I'm 10k so let’s see. (Warning: lot of useless text to sort my thoughts)
If I understand you are to play the complete endgame as white perfectly. I did a quick count and the position on the board is close to even or even (at most a 2 point difference). I also think there are no sente endgame moves left.

Step one

Click Here To Show Diagram Code
[go]$$W White to play. Japanese rules. 6.5 komi.
$$ ---------------------------
$$ | . . X O . . . O O . C . . |
$$ | X . X O O O O . O . X X . |
$$ | . X O M . . . O O O X O . |
$$ | C X X X X B X C X O B X . |
$$ | O O O O X . X . X X . X . |
$$ | . . O X . X . X O X . X . |
$$ | . . O X S X . X O X X X . |
$$ | . O O . O X . X O S O X . |
$$ | . X O O X X X O O X . X . |
$$ | . O . O O X X . O X X X O |
$$ | . . O . . . X O O O O X X |
$$ | . . O O X X X O . O O X . |
$$ | . . O W O O X O O . O M . |
$$ ---------------------------[/go]

:ec: are 1 point gote moves. :ex: are 2 point gote moves. The :es: moves are worth 3 and 4 points respectively. Notice that I have added stones to the board to make it easier :D For my board I think the correct sequence is:

Click Here To Show Diagram Code
[go]$$W White to play. Japanese rules. 6.5 komi.
$$ ---------------------------
$$ | . . X O . . . O O a 9 0 . |
$$ | X . X O O O O . O . X X . |
$$ | . X O 5 . . . O O O X O . |
$$ | 7 X X X X B X 8 X O B X . |
$$ | O O O O X . X . X X . X . |
$$ | . . O X 2 X . X O X . X . |
$$ | . . O X 1 X . X O X X X . |
$$ | . O O 3 O X . X O 4 O X . |
$$ | . X O O X X X O O X . X . |
$$ | . O . O O X X . O X X X O |
$$ | . . O . . . X O O O O X X |
$$ | . . O O X X X O . O O X . |
$$ | . . O W O O X O O . O 6 . |
$$ ---------------------------[/go]

My thoughts: 1 is the biggest move since a black move at 3 is worth 4 points. Next 2 is sente so black will play it immediately. (The next largest play is 3 points, answering the move is not losing 4). Then 4 is the next largest play with 3 points. Then both players get a 2 point move. White gets sente so gets 2 of the remaining 1 point moves (he has to connect at a)

Total time spent reading: about 5 min.
Estimated chance that I screwed up: 5%
Conclusion: solving this board is probably 10k (maybe even 15k).

Step Two

Click Here To Show Diagram Code
[go]$$W Weird points
$$ ---------------------------
$$ | . . X O . . . O O . . . . |
$$ | X . X O O O O . O . X X . |
$$ | . X O . . C . O O O X O . |
$$ | . X X X X . X . X O C X . |
$$ | O O O O X . X . X X . X . |
$$ | . . O X . X . X O X . X . |[/go]


The :ec: points are weird (don't know what to call them). If black plays one he gets 2 points in gote, but if white plays them he can only reduce blacks potential by one point in gote :-? Now, if white starts the 2 point period by playing them and black blocks, he gets 1 point in gote for each. They then continue like in my fist diagram where white gets one more point. If he instead ignores whites threats and plays the other circled point, he gets 2 2-point endgame moves and white gets one. Additionally there are then 4 1-point moves and black gets the same amount as white.
First Variation: B: 5 | W: 4
Second Variation: B: 6 | W: 4
So what if white starts with one of the normal 2-point plays? Black has a choice either to play the other 2-point move or one of the :ec: moves. If he plays the normal one, White will definitely play one of the :ec: points. Black ignores (since it seems better) and plays the second :ec: point. We have the same result as the second sequence. If black plays both :ec: he ends up in a worse position, because the number of 1-point moves is odd again and white becomes an additional point.

Click Here To Show Diagram Code
[go]$$W White to play. Japanese rules. 6.5 komi.
$$ ---------------------------
$$ | . . X O . . . O O . b . . |
$$ | X . X O O O O . O . X X . |
$$ | . X O 5 . 7 . O O O X O . |
$$ | 0 X X X X a X 9 X O 8 X . |
$$ | O O O O X . X . X X . X . |
$$ | . . O X 2 X . X O X . X . |
$$ | . . O X 1 X . X O X X X . |
$$ | . O O 3 O X . X O 4 O X . |
$$ | . X O O X X X O O X . X . |
$$ | . O . O O X X . O X X X O |
$$ | . . O . . . X O O O O X X |
$$ | . . O O X X X O . O O X . |
$$ | . . O W O O X O O . O 6 . |
$$ ---------------------------[/go]


Total additional time spent reading: 10 min.
Estimated chance that I screwed up: 5% (I'm still pretty certain)
Conclusion: I would still say that it can be solved by a 15k because it is not complicated, just tiresome. Is it practical in an actual game? HELL NO. I've played tournament games in less time than to find the last 10 moves of this game.


If I have some more time I'll look at the position on the lower side. :D Hope I could help/make you laugh
Author:  daal [ Sun Jan 30, 2011 3:45 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
Bill Spight wrote:
My intended solver is around 10 kyu.


I'm really hoping this is a joke. :-?

I am pushing 5k and am 100% certain that I cannot solve this correctly (without first reading your book). :)
Author:  Loons [ Sun Jan 30, 2011 4:15 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
Hey Mnemonic:
Could you post your answer more unambiguously ? Like, a final two diagrams (1-10 and 11-15) or an SGF ? I have a little bit of trouble following your thoughts. You could just arbitrarily choose between equivalent value moves.

Also, it might be worth noting that the way I, and I imagine freegame and jts all solved this wasn't an extensive calculation; starting with white, we looked for their best play, then did the same for black, then white again etc etc and so came to our solutions that way (maybe with some notes added in posterity).



My solution and slight variations, for reference:
[sgf-full](;GM[1]FF[4]CA[UTF-8]AP[CGoban:3]ST[2]
RU[Japanese]SZ[13]KM[0.00]
GN[White to play. Japanese rules. 6.5 komi.]DT[2011-01-30]PC[http://lifein19x19.com/]AW[da][ha][ia][db][eb][fb][gb][ib][cc][hc][ic][jc][lc][jd][ae][be][ce][de][cf][if][cg][ig][bh][ch][eh][ih][kh][ci][di][hi][ii][bj][dj][ej][ij][mj][ck][hk][ik][jk][kk][cl][dl][hl][jl][kl][cm][em][fm][hm][im][km]AB[ca][ab][cb][kb][lb][bc][kc][bd][cd][dd][ed][gd][id][ld][ee][ge][ie][je][le][df][ff][hf][jf][lf][dg][fg][hg][jg][kg][lg][fh][hh][lh][bi][ei][fi][gi][ji][li][fj][gj][jj][kj][lj][gk][lk][mk][el][fl][gl][ll][gm]C[Diagram from http://lifein19x19.com/

White to play. Japanese rules. 6.5 komi.]PL[W]
;W[eg]
(;B[jh]C[ 2nd pass, slight move differences. 30.5 30 ('cause of E1 area)]
;W[ek]
;B[dc]
;W[lm]
;B[dm]
;W[em]
;B[kd]
;W[fc]
;B[fd]
;W[ka]
;B[la]
;W[ja]
;B[ad]
;W[hd])
(;B[ef]C[ first pass, this move is not necessary. 29.5 29]
;W[dh]
;B[jh]
;W[ek]
;B[dm]
;W[dc]
(;B[em]
;W[lm]
;B[kd]
;W[fc]
;B[hd]
;W[ka]
;B[la]
;W[ja]
;B[ad]
;W[fd])
(;B[lm]C[30.5-30 again this way]
;W[em]
;B[kd]
;W[fc]
;B[ad]
;W[fd]
;B[hd]
;W[ka]
;B[la]
;W[ja])))[/sgf-full]
Edit: tried posting my sgf properly as an attachment and it bugged somehow, so copypasting the text...



Edit:
For Bill:
A collection of such problems segueing into more and more complex endgame exchanges would be a fun/good exercise book, especially for people like me who tend to lose dozens of points in late yose in actual games, I think. I do suspect that a lot of the people saying "too hard/cannot solve with any confidence" think that the problem is to know the score at a glance from the first diagram, which indeed would involve a painful amount of meaningless reading to reach black's score.
Author:  mw42 [ Sun Jan 30, 2011 4:20 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
I got a half-point win for white. I commented on the value of each of the moves. I thought it was pretty easy, but maybe I erred.
Author:  Phelan [ Sun Jan 30, 2011 4:28 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
I thought it was too hard, and from the answers I saw, I even missed some critical points. I need to work on my endgame. :oops:
Author:  mw42 [ Sun Jan 30, 2011 4:33 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
Phelan wrote:
I thought it was too hard, and from the answers I saw, I even missed some critical points. I need to work on my endgame. :oops:


Have a look at mine, I wrote down the point values. There's nothing "tricky" that I saw so it's pretty straight-forward "play-the-biggest-move" yose endgame (but there was something tricky that I didn't see :mrgreen: ).
Author:  Bill Spight [ Sun Jan 30, 2011 4:53 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
mw42 wrote:
I got a half-point win for white. I commented on the value of each of the moves. I thought it was pretty easy, but maybe I erred.



:b12: is a mistake.
Author:  mw42 [ Sun Jan 30, 2011 5:07 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
Bill Spight wrote:

:b12: is a mistake.


Thanks for having a look at my solution, but I must admit your comment baffles me. :-)
Author:  Loons [ Sun Jan 30, 2011 5:19 pm ]
Post subject:  Re: Basic endgame problem -- feedback requested
I'm a little curious; ye fellow sdks who are saying "seems very hard" - could you go into more detail ? Which step(s) are ambiguous, for example. I know I find the e-1 affair a little niggle-y to consciously think about (a two point gote that makes a new two point gote if black plays it? Is that how you describe it?), but not really write-off material.

Bill; I added a small note for you in my previous post. Also, how hard do you find this sort of diagram/problem to create ?

a hint for mw42
mw; saying twelve is incorrect is true, but might be misleading.
Page 1 of 4 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/