Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / can rybka solve it?
1 2 Previous Next  
Parent - By peterhiarcs (**) Date 2009-09-24 11:01 Edited 2009-09-24 11:07
Thanx, I have to look closer, but it's a good trial, except the typo at 11.. Kd6, you propbly mean d7. Here's, what I tried for myself, coming onwards in 2 of 3 cases but then again blind end:

1. Nd4 Qg7+ 2. Kh3 Qxh6 3. Nf4 Kc8 4. Nde6 Kd7 5. h5 c5 6. Kg4 Kc6 7. Kf3 b5 8.
Ke4 bxa4 9. c4 a6 10. Kd3 Kb6 11. Kc2 Ka5
12. Kc3 Kb6 13. b3 axb3 (13... Ka5 $2
14. bxa4 Kb6 (14... Kxa4 15. Nxc5+ Ka5 16. Nce6) 15. a5+ Kxa5 16. Kb3 Kb6 17.
Ka4 a5 18. Ka3 Kc6 19. Kb3 Kb6 20. Ka4 Ka6 21. Nxc5+) (13... a3 $2 14. Kc2 Ka5
15. Kb1 Kb6 16. Ka2 a5 17. Kxa3 Ka7 18. Kb2 Kb6 19. Kc3 Kc6 20. Kd3 Kb6 21. Ke3
Kc6 22. Ke4 Kb6 23. Kd5) 14. Kxb3 Ka7 15. Kc3 (15. Ka4 Ka8 16. Ka5) 15... Kb8
16. Kd3 Ka7 17. Ke4 Kb6 18. Kd3 (18. Kd5 $2 a5)
Parent - By peterhiarcs (**) Date 2009-09-24 11:35
Thanks a lot again, MoKy!
You were a great help, your variant seems alright to me. Here's one little subvariant of mine, just to show, I've got it.
:)

1. Nd4 Qg7+ 2. Kh3 Qxh6 3. Nf4 Kc8 4. Nde6 Kd7 5. h5 c5 6. Kg4 Kc6 7. b3 ...

7... Kd7 8.Kf3 Kc6 9. Ke4 a6 10. Ke3 Kb7 11. Kf3 Kc6 12. Ke4 Kb7 13. Kd5 1-0
Parent - - By MoKy (**) Date 2009-09-23 11:38
Solution

1. Nd4 Qg7+ 2. Kh3 Qxh6 3. Nf4 Kd7 (3... Kc8 4. Nde6 Kb7) 4. Nde6 c6 (4... c5
5. Kg4 Kc6 6. b3 $1 (6. b4 $2 cxb4 7. cxb4 a5 $1 8. bxa5 b5 $1 $11) (6. h5 $2
b5 7. a5 c4 8. Kf5 Kb7 9. Nc5+ Kc6 10. Nce6 $11) 6... b5 7. a5 $1 b4 8. c4 Kb7
9. Nxc5+ Kc7 10. Nce6+ Kb7 11. h5 Ka6 12. c5 Kxa5 13. c6 Kb6 14. c7 Kb7 15. Kf5
a5 16. Ke4 $18) (4... Kc6 5. b4 Kb7 6. Kg4 c5 (6... Kc8 7. h5 Kb8 8. Kf3 Kc8 9.
Ke4 Kb8 10. Kd5 Kb7 11. c4 a6 12. a5 bxa5 13. bxa5 Ka7 14. Kc6 Kb8 15. Kd7 Kb7
16. c5 $18) 7. b5 a6 8. h5 axb5 9. axb5 Kc8 10. Kf5 Kd7 11. Ke4 Ke7 12. Kd5 Kd7
13. c4 $18) (4... a5 5. Kg4 Kc6 6. h5 b5 7. b3 Kb6 8. Kf5 c5 9. Ke4 c4 10. bxc4
bxa4 11. Kd3 Kc6 12. Kc2 Kd7 13. Kb2 (13. c5 Kc6 14. c4 $18) 13... Qxe6 14.
Nxe6 Kxe6 15. Ka3 Kxe5 16. Kxa4 Kd6 17. Kb5 a4 18. Kxa4 Kc5 19. Kb3 h6 20. Kc2
Kxc4 21. Kd2 Kc5 22. Kd3 Kd5 23. c4+ Kc5 24. Kc3 Kc6 25. Kd4 Kd6 26. Ke4 Kc5
27. Kf5 Kxc4 28. Kg6 Kd5 29. Kxh6 Ke6 30. Kg6 $18) 5. Kg4 Ke7 6. h5 Ke8 7. Kf3
Kd7 8. Ke4 Ke7 9. b4 Kd7 10. Kd4 (10. c4 $2 Kc8 $1 (10... Ke7 $2 11. Kd4 (11.
b5 $2 c5 $2 12. Kd5 Kd7 13. a5 $18) 11... Ke8 12. c5 $1 Kd7 13. Kc4 Kc8 14. b5
Kb7 15. Kb4 cxb5 16. Kxb5 $18) 11. Kd4 (11. a5 Kd7 12. axb6 axb6 13. Kd4 Ke7
14. b5 cxb5 15. cxb5 Ke8 $1 $11) 11... Kb7 $1 12. c5 a5 $1 $11) 10... Kc8 11.
Kc4 $1 (11. c4 $2 Kb7 $1 12. b5 (12. Kd3 Kc8 13. Ke4 Kb8 14. Kd4 Kb7 $1 15. c5
a5 $1 $11) 12... cxb5 13. cxb5 Kc8 14. Kc4 Kd7 15. Kd5 Kc8 $1 16. Kc6 Kb8 17.
Kd7 Kb7 $11) 11... Kb8 (11... a6 12. Kd4 Kb8 13. c4 Kb7 14. b5 $1 cxb5 15. cxb5
axb5 16. axb5 Kc8 17. Kc4 $18) (11... Kb7 12. b5 a5 13. bxc6+ Kxc6 14. Kd4 b5
15. axb5+ Kxb5 16. c4+ Kb4 17. c5 a4 18. Kd3 Kb3 (18... a3 19. Kc2 $18) 19. c6
a3 20. c7 a2 21. c8=Q a1=Q 22. Qb8+ Ka3 23. Qa7+ Kb2 24. Qb6+ Ka3 (24... Kc1
25. Qc5+ Kb2 26. Qb4+ Kc1 27. Qd2+ Kb1 28. Qc2#) 25. Qa5+ Kb2 26. Qb4+ Ka2 27.
Kc2 $18) 12. b5 Kb7 13. Kb4 cxb5 (13... c5+ 14. Kc4 Kc8 15. Kd5 Kd7 16. c4 (16.
a5 bxa5 17. Kxc5 Kc8 18. Kc4 Kb7 19. Kb3 Kb6 20. Ka4 a6 21. bxa6 Kxa6 22. c4
Kb6 23. c5+ Ka6 24. c6 Kb6 25. c7 Kb7 26. Kxa5 $18) 16... Kc8 17. Kc6 Kb8 18.
a5 $1 $18) 14. Kxb5 a6+ (14... Kc8 15. Kc6 Kb8 16. c4 Kc8 17. c5 bxc5 18. Kxc5
Kb7 19. Kb5 Kb8 20. Kc6 Kc8 21. a5 Kb8 22. a6 Kc8 23. Kd6 Kb8 24. Kd7 $18) 15.
Kb4 Kc6 16. c4 Kb7 (16... a5+ 17. Kc3 Kd7 (17... Kb7 18. c5 bxc5 19. Nxc5+ Kc6
20. Nce6 $18) 18. c5 bxc5 19. Nxc5+ Kc6 20. Nce6 $18) 17. c5 Kc6 18. cxb6 Kxb6
19. a5+ Kc6 20. Kc4 $18 1-0
Parent - - By Mark (****) Date 2009-09-25 19:48 Edited 2009-09-25 20:34
I still can't find the win for white.  What do you play if black moves 15. ... Kc8 instead of Kc6 in your primary line?

8/1k5p/pp2N2q/4P2P/PK3N2/2P5/8/8 b - -


edit:  IDeA now has +5.50 for Kb8

edit2:  back to the drawing board.  IA has +3.2 after black moves Kc8
Parent - - By MoKy (**) Date 2009-09-25 23:00 Edited 2009-09-25 23:10
Rybka is here rather useless. From the 5th move I mostly analysed alone. The idea is to create black weaknesses vulnerable by undermining, zugzwang theme and approach to critical squares that eventually leads to a passed pawn.
Parent - - By Mark (****) Date 2009-09-25 23:04
I'm trying to use IDeA to analyze and keep track of all the variations.  I've found you have to be pretty interactive, though...
Parent - - By MoKy (**) Date 2009-09-26 00:16
Sadly, now I have found a draw for black.
Parent - - By Dadi Jonsson (Silver) Date 2009-09-26 00:25
I'm not surprised. This is a very complicated endgame. I haven't continued my analysis since (the new) IDeA found the queen trap that got this topic going, but it did evaluate the position as a draw at that point. I'll let it run overnight and see what happens, but this is a very complicated endgame so you need a lot of analysis before you can draw any conclusions. Your analysis was the deepest so far.

Show us the draw you found.
Parent - By Mark (****) Date 2009-09-26 00:53
How do you like the new IDeA?  Is it much different?
Parent - By Roland Rösler (****) Date 2009-09-26 04:48

>This is a very complicated endgame.


It isn´t. If Black waits, it´s over! Black has to push and change pawns as fast as possible. If White King arrives at e4 (e3) and there are three white pawns on the queenside and c-pawn is stiil on c3, it´s over.
Parent - - By peterhiarcs (**) Date 2009-09-24 19:12
Dear Jonas!
Thank you very much again for bringing this marvellous study along but may I ask you once again, now that several people me included have spent some time with it, where you found it?
I think it's great and we should care not to damage the author's effort as for publicizing it according to his own choice.
Parent - - By Fulcrum2000 (****) Date 2009-09-24 19:56
I think I found the source of it:
http://www.schachmatt.de/69-schachraetsel/4240-schachraetsel-gesucht-wo-selbst-schachprogramme-versagen-3.html#post47091

The guy claims this is his first endgame study. WoW, not a bad start I would say :-)
Parent - By peterhiarcs (**) Date 2009-09-24 20:57
:)
I'd also feel encouraged.
Parent - - By Jonas (****) Date 2009-09-25 05:39
yes, its from there... my account there is thinkoutsidethebox...
the problem is i still dont know the author and his account just says konrade and i tried to contact him but he doesnt respond.
so i have no clue who he is and he also didnt tell the solution..
Parent - By peterhiarcs (**) Date 2009-09-25 07:03
Thanks a lot again, thinkoutsidethebox!
:)
Parent - - By MoKy (**) Date 2009-09-26 00:46 Edited 2009-09-26 00:49
[Event "Jonas Study"]
[Result "1/2-1/2"]
[FEN "3k4/p1pq3p/1p5R/4P3/P6P/2PN4/1PN3K1/8 w - - 0 1"]
[PlyCount "26"]

1. Nd4 Qg7+ 2. Kh3 Qxh6 3. Nf4 Kc8 4. Nde6 c5 5. b3 (5. h5 c4 6. Kg4 Kb7
7. Kf3 Ka6 8. Ke4 Ka5 9. Kd4 Kxa4 10. Kxc4 a5) (5. b4 cxb4 6. cxb4 a5 7. bxa5
Kb7 8. h5 Ka6) 5... Kb7 6. h5 Ka6 7. b4 (7. Kg4 Ka5 8. Kf3 c4) 7... cxb4 8.
cxb4 Kb7 9. Kg4 a5 10. b5 Kc8 11. Kf3 Kd7 12. Ke4 Ke7 13. Kd5 Kd7 1/2-1/2

A little repair b2->b3
3k4/p1pq3p/1p5R/4P3/P6P/1PPN4/2N3K1/8 w - - 0 1

Now it should be OK
Parent - By Roland Rösler (****) Date 2009-09-26 04:14
Very good! :-)

>A little repair b2->b3
>Now it should be OK


I don´t think so! :-(
It isn´t only one tempo! When Black plays active how you have shown (Kc8! and c5!), it´s a draw too! Whites king is too far away.

It´s a pity! :-( Nice study, but incorrect!

Nevertheless this study is very good to show the ignorance of the engines. Show me the engine, which plays with Black forward (Kc8, c5) and hold the draw and I will buy it. Rybka (like always) does nothing and look around. But here this behaviour is suicide!
Parent - - By peterhiarcs (**) Date 2009-09-26 13:18
Dear MoKy!
What do you think about a revival by changing the starting position of the study just as for the a-Pawn putting it on a5 instead of a7:

[Event "?"]
[Site "?"]
[Date "2009.09.26"]
[Round "?"]
[White "Peter"]
[Black "MoKy"]
[Result "1-0"]
[Annotator "Martan,Peter"]
[SetUp "1"]
[FEN "3k4/2pq3p/1p5R/p3P3/P6P/2PN4/1PN3K1/8 w - - 0 1"]
[PlyCount "27"]

{1536MB, b.ctg, PCWINXP} 1. Nd4 Qg7+ 2. Kh3 Qxh6 3. Nf4 c6 (3... c5 4. Nde6+
Kc8 5. c4 Kd7 6. Kg3 Ke8 7. h5 Kd7 8. Kf3 Kc6 9. Ke4 Kb7 10. Kd5) (3... Kc8 4.
Nde6 Kb8 (4... c6 5. h5 b5 6. axb5 cxb5 7. Kg4 a4 8. Kf3 Kb7 9. Ke4 Ka7 10. Kd5
Kb6 11. Kd4 Kc6 12. c4 bxc4 (12... b4 13. c5 Kb5 14. Kd3 a3 (14... Qxf4 15.
Nxf4 Kxc5 16. e6) (14... b3 15. Kc3) 15. bxa3 bxa3 16. Kc2 a2 17. Kb2) 13. Kxc4
) (4... Kb7 5. h5 b5 6. b3 bxa4 (6... Kb6 7. Kg3 b4 (7... c5 8. c4) 8. cxb4
axb4 9. Kf3 c5 10. Ke4 c4 11. bxc4 b3 12. Kd3 Ka5 13. Kc3 Kxa4 14. Kb2 Kb4 15.
c5) 7. bxa4 Kc6 8. Kg3 Kb6 9. Kf3 Kc6 10. Ke4 Kb6 11. Kd5 Kb7 12. Kc5 Kb8 13.
Kb5 Kb7 14. Kxa5) (4... b5 5. axb5 a4 (5... Kb7 6. c4 Kb6 7. b3 Kb7 8. h5 Kb8
9. c5 Kc8 10. Kg3 Kb7 11. Kf2 Kc8 12. Ke3 Kb8 13. Kd4 Kc8 14. c6 Kb8 15. Kc5
Kc8 16. b6 cxb6+ 17. Kxb6) 6. c4 Kb7 7. c5 c6 (7... Kb8 8. h5) 8. b6 Ka6 9. h5
Kb5 (9... Kb7 10. Kg4 Ka8 11. Kf5 Kb7 12. Ke4 Kb8 13. Kd3 Kb7 14. Kc3 Kc8 15.
Kb4) 10. b7) 5. Kg3 Kb7 (5... c6 6. h5 Kb7 7. Kf3 Ka6 8. Ke4 Kb7 9. Kd4 Ka7 10.
c4 Ka6 11. Ke4 Ka7 12. c5 bxc5 13. Kd3 Kb7 14. Kc4 Kb6 15. Kb3 Ka7 16. Kc3 Kb6
(16... Ka6 17. Nxc5+) 17. Kc4 Ka7 18. Kxc5 Kb7 19. b3 Kc8 20. Kxc6) 6. h5 Kc8 (
6... Kc6 7. c4 Kb7 8. Kf3 Kc8 9. Ke4 Kb8 10. c5 Ka7 11. cxb6+ cxb6 12. Kd5) 7.
Kf3 Kd7 8. Ke4 Kc6 (8... Ke7 9. Kd5) 9. Kd4 Kd7 (9... b5 10. axb5+ Kb6 (10...
Kxb5 11. Nxc7+ Ka4 12. Nfd5 Qxh5 13. e6)) 10. Kd5 Kc8 11. Kc6 Kb8 12. c4 Kc8
13. c5 bxc5 14. Kxc5 Kd7 15. Kb5 Ke7 16. Kxa5 Kd7 17. Kb5 Ke7 18. Kc6) 4. Nde6+
Kd7 5. h5 Kc8 (5... b5 6. Kg3 bxa4 7. Kf3 Kc8 8. Ke4 Kb7 9. Kd4 Kb6 10. c4 a3
11. bxa3 a4 12. Kc3 Ka5 13. Kd2 Kb6 14. Kc2 Ka5 15. Kc3 Ka6 16. Kb4 Kb6 17.
Kxa4 Ka6 18. Kb4 Kb6 19. Kc3 Ka6 (19... c5 20. Kd3 Ka5 21. Ke4 Ka4 (21... Kb6
22. Kd5) 22. Nxc5+ Kxa3 23. Nce6 Kb4 24. Kd5) 20. Kd4 Kb6 21. a4 c5+ 22. Kd5) (
5... c5 6. c4 Kc6 7. Kg3 Kd7 8. Kf3 Kc6 9. Ke4 Kd7 10. Kd5) 6. Kg2 Kb7 7. Kf3
Kb8 8. Ke4 Ka7 9. c4 Ka8 10. Kd4 Kb7 11. c5 b5 12. axb5 cxb5 13. Kd5 Kc8 14. c6
1-0
Parent - - By Jonas (****) Date 2009-09-27 15:21
i would say, if you change the starting position, try to make it easier. for example by removing a black pawn. it would be interesting if any engine would find it.
Parent - By peterhiarcs (**) Date 2009-09-27 18:50
That's not the point, I didn't want to make it easier, I just wanted to make it a sure win for White and that's what I did, I guess, or does anybody find a way for Black to hold Remis now?
Parent - By peterhiarcs (**) Date 2009-09-28 09:12
But just for fun and to please you and the engines, taking away Black's a-pawn makes it solveable for most engines I tried, easily.
:)
R3 does takes some more time to find the solutionfor than for example Deep Fritz 10 or Stockfish, which both find it almost at once.
Quad 4x2,5GHz, 1024 Mb hash, complete 5MOB tbs, some 6 ones.
3k4/2pq3p/1p5R/4P3/P6P/1PPN4/2N3K1/8 w - - 0 1


Analysis by Rybka 3 32-bit4:

1.Kf3 Dxd3+ 2.Se3 Dxc3 3.Txh7 Dxe5 4.h5
  =/+  (-0.65)   Tiefe: 6   00:00:00  9kN
...
1.Sd4 Dg7+ 2.Kh3[] Dxh6 3.Sf4[] c5 4.Sde6+[] Kd7 5.Kg4 Kc6 6.Kg3 Kb7 7.Kg4 Ka6 8.Sc7+ Ka5 9.Scd5 b5 10.axb5 Dg7+ 11.Kf5 Kxb5 12.e6 Kc6 13.h5
  =/+  (-0.69)   Tiefe: 20   00:02:08  18879kN, tb=23
1.Sd4 Dg7+ 2.Kh3[] Dxh6 3.Sf4[] Ke8 4.Sde6 Kd7 5.b4 Kc8 6.Kg4 Kb8 7.Kh3
  =/+  (-0.69)   Tiefe: 21   00:02:39  23151kN, tb=27
1.Sd4 Dg7+ 2.Kh3[] Dxh6 3.Sf4[] Ke8 4.Sde6 Kd7 5.b4 Kc8 6.Kg4 Kb8 7.h5 Ka7 8.Kf5 c6 9.Ke4 Kb7 10.Sd8+ Kb8 11.Sde6
  =/+  (-0.56)   Tiefe: 22   00:04:04  34383kN, tb=53

And regard, R3 doesn't find the whole truth of the winning PV for a rather long time (I stopped at 30 minutes), neither do other engines I tried.
Positive score for White Fritz gets within minutes, the others take more or less time for this and of course you can take back the winning score from the end of PV given to the engine.

1. Nd4 Qg7+ (1... Kc8) 2. Kh3 Qxh6 (2... Kc8) 3. Nf4 Kd7 4. Nde6 c5 (4... Kc8
5. h5 Kb7 6. Kg3 Kc6 7. Kf3 Kb7 8. Ke4 c6 (8... Kc8 9. Kd5 Kb7 10. b4 Kb8 11.
Kc6) 9. b4 Ka6 10. c4 Ka7 11. b5) 5. Kg3 Kc8 6. h5 Kd7 7. Kf3 Kc8 8. Ke4 Kb7 9.
Kd5 1-0

Here the winning variants, going them back from the end, engines like Fritz stop taking the victim of the Rook at the start.
By the way to take both a- Pawns off the board at the start, the black and the white one, it's Remis again cause the white King again isn't fast enoug to support it's pawns
- - By Kornrade (*) Date 2009-09-28 19:33
Hello everybody,

First of all, thanks for the nice words, and enthusiastic analysis :)
As already mentioned, this is indeed my first endgame study (before this I only composed 3 published puzzles, toying around with queens and knights Matplus [http://www.matplus.net/pub/start.php] MP 33-34/09 , puzzles 1229, 1230 and 1233). This study has been published in Matplus MP 35/09 puzzle 1354, but in the discussions with editor Iuri Akobia, with whom I had a short mail exchange, we did not go very deep. Thus, the only somewhat thorough analysis (until you guys), has been my own, using Fritz 11 as a board to move pieces on.
Until today, I did not know that this puzzles had its own thread in this forum :), and you can imagine that there is a lot of information here that I need to read carefully before being able to answer.
In my home analysis, all lines led to win for white, through zugzwang. But I don't know yet if I have omitted something or not.

And one more thing, I really enjoy that engines are struggling with the puzzle...

All the best,
Kornrade

[Btw, the real name can be found in Matplus, Kornrade is only its translation in German]
Parent - - By Jonas (****) Date 2009-09-30 07:21
well, we are all very impressed by this position.. one of the best i´ve ever analized i can say for myself
Parent - - By peterhiarcs (**) Date 2009-09-30 08:23
So say I!
:)
Therefore I'm going to bring it to Glarean- journal
http://glareanverlag.wordpress.com/
next week with a little story about it's journey through the fora (or can one say forums in English too?)
Thanks for your help, Jonas (or should I say thinkoutsidethebox?)
:)
Parent - - By Mark (****) Date 2009-09-30 10:56
So what is the final conclusion on this, that the original position is a draw, and the position without black's a-pawn (or with a pawn on a5) is a win for white?
Parent - - By peterhiarcs (**) Date 2009-09-30 12:06
Correct.
Parent - - By Roland Rösler (****) Date 2009-09-30 23:35 Edited 2009-09-30 23:37
Black pawn on a6 is a win too and much smarter than pawn on a5! See the link from Kornrade.
Parent - - By peterhiarcs (**) Date 2009-10-01 05:09
So it is, Roland. Just to keep the secret before publicizing new version, I didn't mention Kornrades (Mihai's, as I my say now, that his name is publicized by himself by the link above) fine idea till now. As you saw, the two other versions with and without black a- Pawn are surprisingly more complicated the one (a5) and much easier the other (missing black a- Pawn). But because it was you to bring up the capturing of the Queen here, what do you think about the study itself now, isn't it great? (And am I right, Pa5 is White winning still?)
Parent - By Roland Rösler (****) Date 2009-10-01 14:54

>And am I right, Pa5 is White winning still?


Pawn on a5 is still winning, but the study with pawn on a6 is smarter because of an aesthetical point of view! It looks better! :-)
Parent - - By George Tsavdaris (****) Date 2009-10-01 12:53 Edited 2009-10-01 12:55
So Pawn on a7 is a draw and Pawn on a6 is a win?
Why? :-P

[EDIT]Anyone remember the game of Kramnik in an endgame(i think 6 piece tablebase) where a Rook move like Re1 or something was draw while Re3 or something was a win(for no logical apparent reason), that Kramnik didn't find? Was it Kramnik-Topalov or something?
Parent - By peterhiarcs (**) Date 2009-10-01 13:58 Edited 2009-10-01 14:32
I guess the difference is just as simple as that: if you remember "MoKy's Variant", holding Remis as for the original Pa7:

1. Nd4 Qg7+ 2. Kh3 Qxh6 3. Nf4 Kc8 4. Nde6 c5 5. h5 c4 6. Kg4 Kb7 7. Kf3 Ka6 8.
Ke4 Ka5 9. Kd4 Kxa4 10. Kxc4 a5

That's possible, because Black has the unique way between his own Pawns over a6 to reach White's Pa4 before the white King can reach it and he can stop White from coming up with his b-Pawn with 5... c4. That's the only way for Remis as far as I saw till now.
This doesn't work with a black Pawn on a6 or on a5, the black King cannot pass them and White can attack this black a- Pawn always before the black King is as far as in MoKy's Variant or blockade it till the white King is near enough to help.
Parent - By Roland Rösler (****) Date 2009-10-01 14:32

>So Pawn on a7 is a draw and Pawn on a6 is a win?
>Why?


Look to the drawing line from Moky! Black needs the square a6 for his King to break through!
- By Kornrade (*) Date 2009-09-28 22:21 Edited 2009-09-28 22:40
Yes MoKy, you are right... in my analysis, I lost one tempo with black (bKd8->d7->c8), or otherwise was not active enough with black, and this was sufficient to win for white.

I think that moving the black a-Pawn to a5 would kill a whole tree of variations (including the draw), but I hope that the study would remain as enjoyable as before.

Thank you all for helping with the analysis :)
- - By buffos (Silver) Date 2009-10-01 08:34
Can white win this?

8/7p/kpp1N2q/p3P2P/P2K1N2/2P5/1P6/8 b - -
Parent - - By peterhiarcs (**) Date 2009-10-01 09:42
How do you come do this position?
Parent - - By buffos (Silver) Date 2009-10-01 09:49
Well its not hard. You can reach this position from the initial position (or the modified one).

Just move the pieces.. first c6.. bring the king a5...

Copying from the link above

1. Sd4!! Dg7+ 2. Kh3 Dxh6 3. Sf4 Kc8 4. Sde6 Kb7 5. h5 a5 6.Kg4 c6 (and not Kc6 as in the web page) and from there on.. you can get to the above position.
Parent - - By peterhiarcs (**) Date 2009-10-01 10:19 Edited 2009-10-01 10:21
I see, it's easier for me having the variant too, so I can look at my own ones already saved.
:)
Look here, that's your position, isn't it?

8/7p/kpp1N2q/p3P2P/P2K1N2/2P5/1P6/8 w - - 0 10


And here's again how we came to it

1. Nd4 Qg7+ 2. Kh3 Qxh6 3. Nf4 Kc8 4. Nde6 Kb7 5. h5 a5 6. Kg4 c6 7. Kf3 Kc8 8.
Ke3 Kb7 9. Kd4 Ka6

And here ist the winning continuation:
10. c4 Kb7 11. c5 Ka6 12. cxb6 Kxb6 13. Kc4 Ka7 (13... c5
14. Kd5) 14. Kc5 Kb7 15. b3 1-0

By the way, if you don't believe me, here's what Deep Fritz 10 says in this position, beeing shown the variant's end, even it takes some time for him to take the score back up.

Analysis by Deep Fritz 10:
...
10.c4 Kb7 11.c5 b5 12.axb5 cxb5 13.Kd5 b4 14.c6+ Kb8 15.Kd6 Kc8 16.Kc5 Kb8 17.Kb5
  +-  (7.20)   Tiefe: 25/58   00:06:08  2521mN, tb=13650
Parent - - By buffos (Silver) Date 2009-10-01 10:36 Edited 2009-10-01 10:42
You did not see that it was black to move in that position.

The key is that when white plays Kd4 black should answer Kb7 and then answer c4 with Ka6

So in your variation

1. Nd4 Qg7+ 2. Kh3 Qxh6 3. Nf4 Kc8 4. Nde6 Kb7 5. h5 a5 6. Kg4 c6 7. Kf3 Kc8 8.
Ke3 Kb8 9. Kd4 Kb7

The result is you have to use b3 once to play c4 without the K at a6 but then you dont have another tempo after you take on b6.
Parent - - By peterhiarcs (**) Date 2009-10-01 11:19 Edited 2009-10-01 11:31
What you need, is a simple endgame- technique, called triangle maneuver:

1. Nd4 Qg7+ 2. Kh3 Qxh6 3. Nf4 Kc8 4. Nde6 Kb7 5. h5 a5 6. Kg4 c6 7. Kf3 Kc8 8.
Ke3 Kb8 9. Kd4 Kb7
10. c4 Ka6 11. Ke4 Ka7 (11... Kb7 12. c5 b5 13. axb5 cxb5
14. Kd5 Kc8 15. Kc6) 12. c5 bxc5 (12... b5 13. axb5 cxb5 14. Kd5 Kb7 15. c6+)
13. Kd3 Kb8 (13... Ka6 14. Nxc5+ Kb6 15. Nce6) 14. Kc4 1-0

Again a Fritz- output after you 9... Kb7, again shown the ends of the variants above. Try around a little with him, it's faster in such positions to make him understand than her (Rybka :)) cause of less nullmove.

8/1k5p/1pp1N2q/p3P2P/P1PK1N2/8/1P6/8 b - - 0 1


Analysis by Deep Fritz 10:

...
10...Ka6 11.Ke4 Kb7 12.c5 bxc5 13.Kd3 Ka7 14.Kc4 Kb6 15.Kc3 Kb7 16.Sxc5+ Kc8 17.Sce6
  +-  (4.37)   Tiefe: 22/22   00:00:00  14kN
Parent - - By buffos (Silver) Date 2009-10-01 11:34

> What you need, is a simply endgame- technique, called triangle maneuvre:


I know about that.... :) but i did not see 11. Ke4 which is very good and after 12 c5 the "best" study like continuation is  12...Ka6 13.cb6 Kb7 14. Kd3 Ka6 15. Kc4 and now Kb6 is forced.
Parent - By peterhiarcs (**) Date 2009-10-01 12:03

>> What you need, is a simply endgame- technique, called triangle maneuvre:
> I know about that.... <IMG class="sic sic_sml_pos" alt=:-) src="/mwf/epx.png">


Sorry for the schoolmaster :)
You've got it.
- By Kornrade (*) Date 2009-10-09 15:36
The story of this puzzle can be read here . It is not meant to be a replacement of the Glarean article, but more a completing essay, giving more details of (among others) this thread.
Up Topic Rybka Support & Discussion / Rybka Discussion / can rybka solve it?
1 2 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill