Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka 2.3.2a Problem
- - By NM (*) Date 2007-06-30 19:20
I was testing 2.3.2a vs 2.3.1. I have the following position.

Black to move
2.3.1 White: Kf2 Qf3 ph5        2.3.2 Black: Kh7 Qf5 pg7 ph6

Rybka 2.3.2 depth 33 -3.67 Qxf3??
Rybka 2.3.1 depth 29 -0.48 Qc5

2.3.2 played Qxf3 and drew shortly. I am curious why 2.3.1 understood this position, while 2.3.2 did not.
Parent - - By noctiferus (***) Date 2007-06-30 19:32
would you please post the FEN?
Parent - - By Uly (Gold) Date 2007-06-30 20:00
8/6pk/7p/5q1P/8/5Q2/5K2/8 b - - 0 1
Parent - - By NM (*) Date 2007-06-30 20:17
Yeah I have no idea how to post fen, but the position above is correct.
Parent - By noctiferus (***) Date 2007-06-30 20:23 Edited 2007-06-30 20:42
Are you using a Chessbase GUI?
Go to the position.
Edit\ copy\copy position,
then paste in the post.  :)
Parent - - By noctiferus (***) Date 2007-06-30 20:20
Thanks, Vytron. The reason of the request is that I'm unable to find this behaviour.
With SPV, at 15 ply, my 232a starts to consider Qc5+ as best.
Parent - - By NM (*) Date 2007-06-30 20:56
Even now if in chessbase if I set up that position, then click add rybka 2.3.2a. I get

depth 15
1. -3.38 Qxf3
2. -3.38 Qf6
3. -3.38 g6

depth 20 
1. -3.38 Qxf3
2. -3.38 Qf6
3. -3.38 g6

depth 25
1. -3.04 Qxf3
2. -3.04 Qf6
3. -3.04 g6

depth 30
1. -3.04 Qxf3
2. -3.04 Qf6
3. -3.04 g6
Parent - - By Felix Kling (Gold) Date 2007-06-30 21:02
This is a problem of the new pawn endings evaluation, you should use at least the 3 and 4 men tablebases, so that it can evaluate the kpk endings right...
Parent - By noctiferus (***) Date 2007-06-30 21:27 Edited 2007-06-30 21:39
perfect. with nalimovs switched off, 231 is ok, while 232a gives wrong eval.

note : Analysis by Rybka 2.3.2 32-bit  20 pl:
1. ³  (-0.55): 1...Qxf3+ 2.Kxf3 g6 3.hxg6+ Kxg6 4.Kg4 h5+ 5.Kg3 Kg5 6.Kh3 h4 7.Kh2 Kh6 8.Kh1
2. ³  (-0.55): 1...g6 2.hxg6+ Kxg6 3.Qxf5+ Kxf5 4.Kg3 Kg5 5.Kh3 h5 6.Kg3 h4+ 7.Kh3 Kh5 8.Kh2
3. ³  (-0.55): 1...Qf6 2.Kg3 Qxf3+ 3.Kxf3 g6 4.hxg6+ Kxg6 5.Kg4 h5+ 6.Kg3 Kg5 7.Kh3 h4 8.Kh2

is less overestimanting
Parent - - By noctiferus (***) Date 2007-06-30 21:05
very strange.

New game - Rybka 2.3.2a 32-bit, 120'/40+60'/20+30'
8/6pk/7p/5q1P/8/5Q2/5K2/8 b - - 0 1


Analysis by Rybka 2.3.2a 32-bit  19ply:

1. ³  (-0.45): 1...Qf6 2.Kg3 Qg5+ 3.Kh3 Kg8 4.Kh2 Qd2+ 5.Kh3 Qd7+ 6.Kg3 Qd6+ 7.Kh3 Qe5 8.Kg4
2. ³  (-0.45): 1...Qc5+ 2.Kg3 Qe5+ 3.Kg4 Qg5+ 4.Kh3 Kg8 5.Kh2 Qd2+ 6.Kh3 Qd7+ 7.Kg3 Qd6+ 8.Kh3
3. ³  (-0.45): 1...Qe5 2.Kg2 Kg8 3.Kh3 Qg5 4.Kh2 Qd2+ 5.Kh3 Qd7+ 6.Kg3 Qd6+ 7.Kh3 Qe5 8.Kg4

(Enrico, Milano 30.06.2007)
Parent - By Felix Kling (Gold) Date 2007-06-30 21:06
I bet you have some tablebases installed :)
Parent - By noctiferus (***) Date 2007-06-30 21:08
F.Kling is right. I switched off nalimovs, and I'm getting your results
Parent - - By Debaser (***) Date 2007-06-30 22:26
Another one, Rybka 2.3.2a likes Rxf5 and missed a win vs Shredder

8/6k1/4K1bp/4Rp2/7P/8/8/8 w - - 0 64
Parent - - By Gaмßito (****) Date 2007-07-01 06:18 Edited 2007-07-01 06:54
Of course, Rybka's line are just wrong.  
 
The program goes very deeply but wrongly, and it does not help in anything here.

8/6k1/4K1bp/4Rp2/7P/8/8/8 w - - 0 1


Analysis by Rybka 2.3.2a 32-bit :

1.Re3 Bh7
  +-  (1.97)   Depth: 5   00:00:00
1.Re3 h5 2.Rc3
  +-  (2.00)   Depth: 6   00:00:00  2kN
1.Re3 h5 2.Rc3 Bh7
  +-  (2.05)   Depth: 7   00:00:00  2kN
1.Re3 h5 2.Rc3 Bh7 3.Ke5
  +-  (2.12)   Depth: 8   00:00:01  3kN
1.Re3 h5 2.Rc3 Bh7 3.Ke5 Bg6
  +-  (2.17)   Depth: 9   00:00:01  4kN
1.Re3 h5 2.Rc3 Bh7 3.Ke5 Bg6 4.Rc7+ Kh6
  +-  (2.17)   Depth: 10   00:00:01  6kN
1.Re3 h5 2.Rc3 Kh7 3.Ke5 Kh6 4.Kf4 Bh7
  +-  (2.21)   Depth: 11   00:00:01  16kN
1.Re1 h5 2.Ra1 Kh7 3.Ra7+ Kh6 4.Ra5 f4 5.Ra3 Be4
  +-  (2.51)   Depth: 11   00:00:01  61kN
1.Re1 h5 2.Ra1 Kh7 3.Ra7+ Kh6 4.Ra5 Kg7 5.Ra7+ Kh6 6.Ra5 Kg7 7.Ra7+ Kh6
  +-  (2.57)   Depth: 12   00:00:01  63kN
1.Re1 f4 2.Rf1 Bd3 3.Rxf4 Kg6 4.Ke5 Kh5 5.Rd4 Be2
  +-  (2.58)   Depth: 13   00:00:01  79kN
1.Re3 f4 2.Rf3 Bf7+ 3.Ke5 Be8 4.Rxf4 Kg6 5.Rb4 Bd7 6.Rd4
  +-  (2.61)   Depth: 13   00:00:02  92kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (2.79)   Depth: 13   00:00:02  133kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (2.79)   Depth: 14   00:00:03  143kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (2.79)   Depth: 15   00:00:04  159kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (2.79)   Depth: 16   00:00:04  186kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (2.91)   Depth: 17   00:00:05  225kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (2.91)   Depth: 18   00:00:06  266kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (2.96)   Depth: 19   00:00:06  341kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.14)   Depth: 20   00:00:07  462kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.14)   Depth: 21   00:00:08  585kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.14)   Depth: 22   00:00:10  746kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.14)   Depth: 23   00:00:13  988kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.14)   Depth: 24   00:00:16  1321kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.14)   Depth: 25   00:00:20  1752kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.14)   Depth: 26   00:00:28  2556kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.15)   Depth: 27   00:00:39  3530kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.15)   Depth: 28   00:00:49  4529kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.15)   Depth: 29   00:01:05  5958kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.15)   Depth: 30   00:01:28  8110kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.15)   Depth: 31   00:02:07  12026kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.15)   Depth: 32   00:02:41  15352kN
1.Rxf5 Be8 2.h5 Bc6 3.Rf7+ Kg8 4.Ke7 Bb7 5.Rf8+ Kg7 6.Rf5 Be4 7.Rf7+ Kg8
  +-  (3.15)   Depth: 33   00:03:24  19537kN

(, AMD 30.06.2007)

Of course, that line with Rxf5 could be very good only if the Bishop goes to e8, like Rybka prefers :-)  
This is a problem related with the lack of knowledge. The program even does not knows that the black King can go to h8, even to lost the pawn in h6, and the position is clearly a draw because white cannot promote its pawn neither to take out the black king of the corner. The program simply does not calculate this line, it does not believe that black will go to take on f5 never, because it is completely sure that, if the opponent capture, white ''will win this easily''.This demonstrates that, pruning so much and goes deeper without enough knowledge, it is not very useful, or it won't help.

Note also that, after a long search, Rybka 2.1 can choose Rc5, Ra5 or Rb5 too, (not Rxf5) because it has the same score and probably it play the move randomly. 2.3.2a also can choose Rc5. It has the same score as Rxf5 after a long search.
 
Vas probably removed some code here, because older versions as 2.2 or 2.1c, can even know that after Rxf5, it is dead draw. 
 
Regards, 
Gambito.



Parent - - By Gaмßito (****) Date 2007-07-01 06:21
Depth 34

Analysis by Rybka 2.3.2a 32-bit :

1. +-  (3.20): 1.Rxf5 Be8 2.h5 Ba4 3.Rf7+ Kg8
2. +-  (3.20): 1.Rc5 f4 2.Rc7+ Kg8 3.Kd5 f3 4.Kd4 Bh5 5.Ke3 Bg4 6.Kf2 Kf8 7.Rh7 h5
3. +-  (3.20): 1.Rb5 f4 2.Rb7+ Kg8 3.Kf6 Bd3 4.Rb8+ Kh7 5.Rb3 Bc2 6.Rf3 Be4 7.Rxf4 Bd3

(, AMD 01.07.2007)

Regards,
Gambito.
Parent - - By Gaмßito (****) Date 2007-07-01 06:36 Edited 2007-07-01 06:53
Depth 32

And 5 moves with the same score. Any movement of those, can be played.

Analysis by Rybka 2.3.2a 32-bit :

1. +-  (3.20): 1.Rxf5 Be8 2.h5 Ba4 3.Rf7+ Kg8
2. +-  (3.20): 1.Rc5 f4 2.Rc7+ Kg8 3.Kf6
3. +-  (3.20): 1.Ra5 f4 2.Ra7+ Kg8 3.Kf6 Bd3 4.Rg7+ Kh8 5.Rd7 Bc2 6.Rd2 Be4 7.Rd4 Bc2
4. +-  (3.20): 1.Re1 f4 2.Rg1 Kh7 3.Kf6 f3 4.Rg3 Bh5 5.Rg7+ Kh8 6.Rg1 Be8 7.Rf1 Bh5
5. +-  (3.20): 1.Rb5 f4 2.Rb7+ Kg8 3.Kf6
6. +-  (2.90): 1.Re3 f4 2.Rf3 Bc2 3.Ke5 Kg6 4.Kxf4 Kh5 5.Re3 Bg6 6.Kg3 Bc2 7.Re5+ Kg6
7. +-  (2.74): 1.Ke7 f4 2.Re2 f3 3.Re3 Bh5 4.Re4 Kg6 5.Rf4 Kg7 6.Ke6 Kg6 7.Kd5 Kg7
8. +-  (2.01): 1.Re2 Bh5 2.Rd2 Bg4 3.Ke5 Kg6 4.Kf4 Bh3 5.Rd1 Bg4 6.Re1 Kf6 7.Kg3 Bh5

(, AMD 01.07.2007)

Regards,
Gambito.
Parent - - By Debaser (***) Date 2007-07-01 10:39
Thanks for your analysis ;)

So I have a question, if several moves have the same score, is it chosen randomly?

Then why appears Rxf5 always the first, and is finally the move done?
Parent - - By Pia (****) Date 2007-07-01 15:37
Because for that move highest score was found first.
Parent - By Debaser (***) Date 2007-07-03 07:45
OK, thanks. I suppose that is better than randomly ;)
Parent - - By Uri Blass (*****) Date 2007-07-03 08:15
No

I think that Vas did not remove code here.

Vas added code.

program with only material evaluation is not going to play Rxf5
Rybka does it because it has knowledge and not because it has not knowledge.

Uri
Parent - - By Gaмßito (****) Date 2007-07-03 08:50 Edited 2007-07-03 09:09
Uri,

I do not understand why you said that 2.3.2 played Rxf5?? based on its knowledge.

I think that some code was remove there, because the older versions can see instantly that after the bad Rxf5, black has an easy draw, while 2.3.2 cannot see it in a long time. Also it is possible that the new added code, is making some conflict there. But, then why the older versions can see it in 0 seconds? What do you believe?

Analysis by Rybka 1.2f 32-bit:

1...Bxf5+ 2.Kxf5 Kf7 3.Ke4 h5 4.Kf3 Kg7 5.Kg2 Kf7 6.Kf3 Kg7 7.Kg2 Kf7 8.Kf3
=  (0.00)    Depth: 59   00:00:32  20641kN
(, AMD 03.07.2007)

Analysis by Rybka 2.1c 32-bit:

1...Bxf5+ 2.Kxf5 Kf7 3.Ke4 Kg6 4.Kf3 Kh7 5.Ke4 Kg6 6.Kf3 Kh7 7.Ke4 Kg6 8.Kf3
=  (0.00)    Depth: 38   00:00:40  22788kN
(, AMD 03.07.2007)

Analysis by Rybka 2.2 32-bit:

1...Bxf5+ 2.Kxf5 h5 3.Kg5 Kf7 4.Kxh5 Kg7 5.Kg5 Kh7 6.h5 Kg7 7.h6+ Kh7 8.Kh5
=  (0.08)    Depth: 37   00:00:33  3837kN
(, AMD 03.07.2007)

Analysis by Rybka 2.3.1 32-bit :

1...Bxf5+ 2.Kxf5 h5 3.Kg5 Kh7 4.Kxh5 Kg7 5.Kg5 Kh7 6.h5 Kg7 7.Kf5 Kh6 8.Kg4
=  (0.00)    Depth: 46   00:00:37  2339kN
(, AMD 03.07.2007)

Analysis by Rybka 2.3.2a 32-bit :

1...Be8 2.Ke7 Ba4 3.Rf4 Bd1 4.Rf1 Be2 5.Rg1+ Kh7 6.Kf6 Bd3 7.Rg7+ Kh8 8.Rc7
+-  (3.23)    Depth: 40   00:30:00  168mN
(, AMD 03.07.2007)

Regards,
Gambito.
Parent - - By Uri Blass (*****) Date 2007-07-03 10:37
You may be right that some code was removed but it seems that some code was also added instead of it(otherwise it could see a score of +1).

2.3.2a can see 2.92 is this winning position immediatly when an older version see only +1 or even less than +1 at small depth.

It seems that rybka have some bonus for having more pawns in pawn endgame and it is already known for a long time.
I already posted an example when rybka cannot see simple draw in the following link
http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?tid=1309

I showed that link in reply in another forum here
http://ancients.correspondencechess.com/index.php?topic=117.0

New game
8/8/8/2k5/8/8/1KP5/8 w - - 0 1


Analysis by Rybka 2.3.2a 32-bit :

1.Kc3 Kd5
  +-  (2.92)   Depth: 5   00:00:00
1.Kc3 Kd5 2.Kd3
  +-  (2.98)   Depth: 6   00:00:00
1.Kc3 Kd5 2.Kd3 Ke6
  +-  (3.06)   Depth: 7   00:00:00
1.Kc3 Kd5 2.Kd3 Ke6 3.Kd4
  +-  (3.06)   Depth: 8   00:00:00
1.Kc3 Kd5 2.Kb4 Ke5 3.c4 Kd6
  +-  (3.11)   Depth: 9   00:00:00  1kN
1.Kc3 Kd5 2.Kb4 Ke5 3.c4 Kd6 4.c5+ Kd7
  +-  (3.51)   Depth: 10   00:00:00  2kN
1.Kc3 Kd5 2.Kb4 Ke5 3.c4 Kd6 4.c5+ Kc6 5.Kc4
  +-  (3.80)   Depth: 11   00:00:00  2kN
1.Kc3 Kd5 2.Kb4 Kd6 3.c4 Kd7 4.c5 Kc8 5.c6
  +-  (3.82)   Depth: 12   00:00:00  4kN
1.Kc3 Kd5 2.Kb4 Kd6 3.Kb5 Kc7 4.Kc5 Kb7 5.Kd6 Kb6 6.c4 Kb7
  +-  (3.98)   Depth: 13   00:00:00  5kN
1.Kc3 Kd5 2.Kb4 Kd6 3.Kb5 Kc7 4.Kc5 Kb7 5.Kd6 Kb6 6.c4 Kb7 7.c5
  +-  (3.99)   Depth: 14   00:00:00  7kN
1.Kc3 Kd5 2.Kb4 Kc6 3.c4 Kb6 4.c5+ Kc6 5.Kc4 Kd7 6.Kd5 Kc7 7.c6 Kd8
  +-  (3.99)   Depth: 15   00:00:00  9kN
1.Kc3 Kd5 2.Kb4 Kc6 3.c4 Kb6 4.c5+ Kc6 5.Kc4 Kd7 6.Kd5 Kc7 7.c6 Kd8
  +-  (3.99)   Depth: 16   00:00:00  10kN

(Uri, MyTown 03.07.2007)
Parent - By Vasik Rajlich (Silver) Date 2007-07-03 18:41
I don't know if it's an important difference, but indeed this is new code.

Vas
Parent - By ernest (****) Date 2007-07-01 20:58

> Another one, Rybka 2.3.2a likes Rxf5 and missed a win vs Shredder


Well, where is the win?
What program finds it?
Parent - - By plicocf (***) Date 2007-06-30 23:20
Before a draw with repetitions moves, this is what Rybka shows,
an higth score and after the zero score.

Paulo Soares
Parent - By Jack-Ireland (**) Date 2007-06-30 23:50
OK, this is surely not normal.
Parent - - By DamirD81 (***) Date 2007-06-30 23:58
You must be doing something wrong, because in this position Rybka played Tc5 on my comp with +-3.02 for white.
Parent - By Pia (****) Date 2007-07-01 04:43
Your comp's with tablebases then. Switch NalimovUsage to Never and you'll see Debaser isn't wrong. 2.3.2a plays Rxf5 +- (3.19)
Parent - By plicocf (***) Date 2007-07-01 16:02
I was looking for the position posted by NM, but i don´t place the position on the board,
I just saw that it´s a draw.
Parent - - By Pia (****) Date 2007-07-01 04:48 Edited 2007-07-01 05:36
Analysis by Rybka 2.3.2a 32-bit :

1...Qf5xf3+ 2.Kf2xf3 g7-g6 3.h5xg6+ Kh7xg6 4.Kf3-g4 h6-h5+ 5.Kg4-f3 Kg6-f5 6.Kf3-g3 Kf5-g5 7.Kg3-h3 h5-h4 8.Kh3-g2 Kg5-g6 9.Kg2-h2
  -+  (-3.86)   Depth: 59   00:07:24  20060kN

(It was a little fake, however. Very little.)

>Before a draw with repetitions moves, this is what Rybka shows, an higth score and after the zero score.


>Paulo Soares


Ah, yes, after 9.5 moves made the score is 0.
Parent - - By BigBen (****) Date 2007-07-01 09:02
Hi,
    I think this is a neutral outlook `problem` .... Just switch to `slightly optimistic` and I think like my machine your comp will play Rc5 ....

I never use the neutral outlook for my games, I cant put my finger on the problem but lots of the problems I see on the forum concerning 2.3.2a vanish just by altering the outlook

Regards
Tony
Parent - By Pia (****) Date 2007-07-01 13:37
Of course. Every found problem is for particular Rybka setting.

In this position 2.3.2a plays Rxf5 in Slightly Optimistic mode too.

8/6k1/4K1b1/4Rp1p/7P/8/8/8 w - - 0 1
Parent - - By plicocf (***) Date 2007-07-01 16:10
Look for this, Pia:
Quad 6600 HT=128 Mb

Analysis by Rybka 2.3.2a 32-bit :
1...Qxf3+ 2.Kxf3 g6 3.Kg4 Kg7 4.hxg6 Kxg6 5.Kg3 h5 6.Kh3 Kg5 7.Kg3 h4+ 8.Kh3
  -+  (-4.11)   Depth: 21   00:00:01  63kN
1...Qxf3+ 2.Kxf3 g6 3.Kg4 Kg7 4.hxg6 Kxg6 5.Kg3 h5 6.Kh3 Kg5 7.Kg3 h4+ 8.Kh3
  -+  (-4.10)   Depth: 22   00:00:01  89kN
________________________________________________________________________
Analysis by Rybka 2.3.2a 32-bit TBS= rarely, convekta tablebases(50 Gb):
1...Qc5+ 2.Kg2 Qg5+ 3.Kf2 Qf6 4.Ke2 Qb2+ 5.Ke3 Qc3+ 6.Kf4 Qf6+ 7.Ke4 Qc6+ 8.Kf4
  =/+  (-0.47)   Depth: 21   00:00:55  7278kN, tb=10550
1...Qc5+ 2.Kg2 Qg5+ 3.Kf2 Qf6 4.Ke3 Qb6+ 5.Kd3 Qb5+ 6.Kd4 Qb4+ 7.Ke5 Qe7+ 8.Kd4
  =/+  (-0.47)   Depth: 22   00:01:05  8686kN, tb=14218
_______________________________________________________________
Analysis by Rybka 2.3.2a mp 4 trheads TBS= rarely, convekta tablebases(50 Gb):
1...Qf6 2.Kg2 Qxf3+ 3.Kxf3 g6 4.hxg6+ Kxg6 5.Ke4 Kg5 6.Kf3 h5 7.Kg3 h4+ 8.Kg2
  -+  (-3.74)   Depth: 21   00:00:01  320kN, tb=50
1...Qf6 2.Kg2 Qxf3+ 3.Kxf3 g6 4.hxg6+ Kxg6 5.Ke4 Kg5 6.Kf3 h5 7.Kg3 h4+ 8.Kg2
  -+  (-3.74)   Depth: 22   00:00:01  332kN, tb=84
_________________________________________________________________________
Analysis by Rybka 2.3.2a mp 4 Threads :
1...Qxf3+ 2.Kxf3 g6 3.hxg6+ Kxg6 4.Kg4 h5+ 5.Kh4 Kh6 6.Kh3 Kg5 7.Kg3 h4+ 8.Kh3
  -+  (-4.10)   Depth: 21   00:00:00  49kN
1...Qxf3+ 2.Kxf3 g6 3.hxg6+ Kxg6 4.Kg4 h5+ 5.Kh4 Kh6 6.Kh3 Kg5 7.Kg3 h4+ 8.Kh3
  -+  (-4.10)   Depth: 22   00:00:01  201kN

I think it´s good for all when shows a test position:
1- The computer used.
2- Hash table.
3- Ponder= on or off. ( If a game)
4- Configuratin of the engine.
5- Any sugestion?

Paulo Soares
Parent - By Pia (****) Date 2007-07-01 20:27 Edited 2007-07-01 20:34
That's weird behavior for mp with TB you have, indeed.
Parent - - By Uri Blass (*****) Date 2007-07-01 11:09
The new rybka is simply weaker in pawn endgames(at least in positions that I compose for her)

Here is another example

Rybka cannot see the draw that is very easy for glaurung

New game
8/k7/Pp1p1p2/1PpPpPp1/2P1P1Pp/3K3P/8/8 b - - 0 1


Analysis by Rybka 2.3.2a 32-bit :

2...Kb8 3.Ke2
  +-  (3.80)   Depth: 5   00:00:00
2...Kb8 3.Ke2 Kc8
  +-  (3.80)   Depth: 6   00:00:00
2...Kb8 3.Ke2 Kc8 4.Kd2
  +-  (3.80)   Depth: 7   00:00:00
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7
  +-  (3.80)   Depth: 8   00:00:00
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 9   00:00:00
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 10   00:00:00
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 11   00:00:00
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 12   00:00:00
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 13   00:00:00
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 19   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 27   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 28   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 29   00:00:00  3kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 30   00:00:00  4kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 47   00:00:00  7kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 47   00:00:00  7kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 47   00:00:00  7kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 47   00:00:00  7kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 47   00:00:00  7kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 47   00:00:00  7kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 47   00:00:00  7kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 47   00:00:00  7kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 47   00:00:00  7kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 51   00:00:00  9kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 54   00:00:00  9kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 54   00:00:00  9kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 54   00:00:00  9kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 54   00:00:00  9kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 54   00:00:00  9kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 56   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN
2...Kb8 3.Ke2 Kc8 4.Kd2 Kc7 5.Ke3 Kb8 6.Kd2 Kc8 7.Ke2 Kc7 8.Ke3 Kb8 9.Kd2
  +-  (3.80)   Depth: 59   00:00:00  10kN

(Uri, MyTown 01.07.2007)

New game
8/k7/Pp1p1p2/1PpPpPp1/2P1P1Pp/3K3P/8/8 b - - 0 1


Analysis by Glaurung 2-epsilon/4:

2...Kb8 3.Ke3
  +-  (3.27)   Depth: 2   00:00:00
2...Kb8 3.Ke3 Ka7
  +-  (3.01)   Depth: 3   00:00:00
2...Kb8 3.Ke3 Ka7 4.Kd3
  +-  (2.92)   Depth: 4   00:00:00
2...Kb8 3.Kd2 Ka8 4.Ke3 Ka7
  +-  (3.01)   Depth: 5   00:00:00
2...Kb8 3.Kd2 Ka8 4.Ke3 Ka7 5.Kd3
  +-  (2.92)   Depth: 6   00:00:00
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7
  +-  (3.01)   Depth: 7   00:00:00  1kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Ke3
  +-  (2.92)   Depth: 8   00:00:00  1kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka8 6.Kc3 Ka7
  +-  (2.88)   Depth: 9   00:00:00  2kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka8 6.Kc3 Ka7 7.Kd3
  +-  (2.92)   Depth: 10   00:00:00  3kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kc3 Ka7
  +-  (2.88)   Depth: 11   00:00:00  4kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kc3 Ka7 8.Kd3
  +-  (2.92)   Depth: 12   00:00:00  6kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kb3 Kc8 8.Kc3 Kc7
  +-  (2.90)   Depth: 13   00:00:00  8kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kb3 Kc8 8.Kc3 Kb8 9.Kc2
  +-  (2.90)   Depth: 14   00:00:00  10kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kb3 Kc8 8.Kb2 Kb8 9.Kc3 Ka7
  +-  (2.88)   Depth: 15   00:00:00  12kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kb3 Kc8 8.Kb2 Kc7 9.Kc3 Kb8 10.Kc2
  +-  (2.90)   Depth: 16   00:00:00  15kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Kc1 Kb8 9.Kc2 Ka8 10.Kc3 Ka7
  +-  (2.88)   Depth: 17   00:00:00  19kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Kc1 Kb8 9.Kc2 Ka8 10.Kc3 Ka7 11.Kd3
  +-  (2.92)   Depth: 18   00:00:00  28kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kf3 Kb8 10.Kf2 Kc8 11.Ke2 Kc7
  +-  (2.82)   Depth: 19   00:00:00  34kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kd1 Kb8 11.Kc2 Ka8 12.Kc3 Ka7
  +-  (2.88)   Depth: 20   00:00:00  40kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kd1 Kb8 11.Kc2 Ka8 12.Kc3 Ka7
  +-  (2.88)   Depth: 21   00:00:00  47kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kd1 Kb8 11.Kc2 Ka8 12.Kc3 Ka7 13.Kd3
  +-  (2.92)   Depth: 22   00:00:00  53kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kd1 Kb8 11.Kc2 Ka7 12.Kb2 Kb8 13.Kc3 Ka7
  +-  (2.88)   Depth: 23   00:00:00  60kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kf1 Kb8 11.Kf2 Ka8 12.Ke2 Kb8 13.Ke3 Ka7 14.Kd3
  +-  (2.92)   Depth: 25   00:00:00  68kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kf1 Kb8 11.Kf2 Ka8 12.Ke2 Kb8 13.Ke3 Ka8 14.Kd2 Ka7
  +-  (2.80)   Depth: 25   00:00:00  77kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kf1 Kb8 11.Kg2 Ka7 12.Kf3 Kb8 13.Ke3 Ka8 14.Kd2 Ka7 15.Ke3
  +-  (2.92)   Depth: 26   00:00:00  100kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kf1 Kb8 11.Kg2 Ka7 12.Kf3 Kb8 13.Ke3 Ka8 14.Kd2 Kb8 15.Kc3 Ka7
  +-  (2.88)   Depth: 27   00:00:00  109kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kf1 Kb8 11.Kg2 Ka7 12.Kf3 Kb8 13.Ke3 Ka8 14.Kd2 Kb8 15.Kc3 Ka7 16.Kd3
  +-  (2.92)   Depth: 28   00:00:00  118kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kf1 Kb8 11.Kg2 Ka7 12.Kf3 Kb8 13.Ke3 Ka8 14.Kd2 Kb8 15.Kc2 Ka8 16.Kc3 Ka7
  +-  (2.88)   Depth: 29   00:00:00  127kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kf1 Kb8 11.Kg2 Ka7 12.Kf3 Kb8 13.Ke3 Ka7 14.Kf2 Kb8 15.Kf3 Ka8 16.Ke2 Kb8 17.Kf2
  +-  (2.90)   Depth: 30   00:00:00  137kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kd1 Kb8 11.Kc2 Ka7 12.Kb3 Ka8 13.Kc3 Kb8 14.Kb2 Kc8 15.Ka3 Kb8 16.Ka2 Ka8 17.Kb3 Ka7
  +-  (2.72)   Depth: 31   00:00:00  146kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Ke1 Ka7 10.Kd1 Kb8 11.Kc2 Ka7 12.Kb3 Ka8 13.Kc3 Kb8 14.Kb2 Kc8 15.Ka3 Kb8 16.Ka2 Ka8 17.Kb3 Ka7 18.Kc3
  +-  (2.78)   Depth: 32   00:00:00  154kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc2 Kc8 12.Kd2 Kb8 13.Kc3 Kc8 14.Kb3 Kc7 15.Kb2 Kb8 16.Ka2 Ka8 17.Kb3 Kb8 18.Ka4
  +-  (2.64)   Depth: 33   00:00:00  183kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc2 Kc8 12.Kd2 Kb8 13.Kc3 Kc8 14.Kb3 Kc7 15.Kb2 Kb8 16.Ka2 Kc7 17.Kb3 Kc8 18.Ka4
  +-  (2.76)   Depth: 34   00:00:00  211kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc2 Kc8 12.Kd2 Kb8 13.Ke1 Ka8 14.Kf2 Kb8 15.Kg1 Kc8 16.Kh1 Kb8 17.Kg2 Ka7 18.Kf3
  +-  (2.64)   Depth: 35   00:00:00  227kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Ke1 Ka8 12.Kf2 Kb8 13.Ke3 Ka7 14.Kf3 Kb8 15.Kf2 Ka8 16.Ke2 Kb8 17.Kf1 Kc7 18.Kg1
  +-  (2.66)   Depth: 36   00:00:00  253kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Ka7 12.Kb2 Kb8 13.Kc3 Ka7 14.Kb3 Kb8 15.Kb2 Ka8 16.Ka3 Kb8 17.Kb3 Kc8 18.Ka2
  +-  (2.64)   Depth: 37   00:00:00  262kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Ka7 12.Kb2 Kb8 13.Kc3 Ka7 14.Kb3 Kb8 15.Kb2 Ka8 16.Ka3 Kb8 17.Kb3 Kc8 18.Ka2
  +-  (2.33)   Depth: 38   00:00:00  275kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Ke1 Kb8 11.Kf2 Ka8 12.Kg2 Kb8 13.Kh2 Ka7 14.Kg1 Kb8 15.Kg2 Ka8 16.Kf2 Ka7 17.Kf1 Ka8 18.Ke2
  ±  (0.74)   Depth: 39   00:00:00  299kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Ke1 Kb8 11.Kf2 Ka8 12.Kg2 Kb8 13.Kh2 Ka7 14.Kg1 Kb8 15.Kg2 Ka8 16.Kf2 Ka7 17.Kf1 Ka8 18.Ke2
  ²  (0.52)   Depth: 40   00:00:00  318kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Ke1 Kb8 11.Kf2 Ka8 12.Kg2 Kb8 13.Kh2 Ka7 14.Kg1 Kb8 15.Kg2 Ka8 16.Kf2 Ka7 17.Kf1 Ka8 18.Ke2
  ²  (0.43)   Depth: 41   00:00:00  331kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Ke1 Kb8 11.Kf2 Ka8 12.Kg2 Kb8 13.Kh2 Ka7 14.Kg1 Kb8 15.Kg2 Ka8 16.Kf2 Ka7 17.Kf1 Ka8 18.Ke2
  ²  (0.39)   Depth: 42   00:00:00  357kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Ka8 11.Kc1 Ka7 12.Kb1 Ka8 13.Ka1 Kb8 14.Ka2 Kc8 15.Kb1 Kc7 16.Kc1 Kc8 17.Kd2 Kb8 18.Kc3
  ²  (0.39)   Depth: 43   00:00:00  389kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Ka8 11.Kc1 Ka7 12.Kb1 Ka8 13.Ka1 Kb8 14.Ka2 Ka8 15.Kb3 Ka7 16.Kb2 Kb8 17.Kc2 Ka8 18.Kc3
  ²  (0.54)   Depth: 44   00:00:00  440kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 45   00:00:00  469kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 46   00:00:00  475kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 47   00:00:00  483kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 48   00:00:00  494kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 49   00:00:00  502kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 50   00:00:00  512kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 51   00:00:00  523kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 52   00:00:00  535kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 53   00:00:00  548kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 54   00:00:00  561kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 55   00:00:00  578kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 56   00:00:00  597kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 57   00:00:00  620kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 58   00:00:00  646kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 59   00:00:00  707kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 60   00:00:01  791kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 61   00:00:01  874kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 62   00:00:01  1107kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 63   00:00:02  1624kN
2...Kb8 3.Kd2 Ka8 4.Ke3 Kb8 5.Kd3 Ka7 6.Kc2 Kb8 7.Kd1 Ka7 8.Ke2 Ka8 9.Kd2 Ka7 10.Kd1 Kb8 11.Kc1 Kc7 12.Kb1 Kc8 13.Ka2 Kb8 14.Ka1 Kc7 15.Kb1
  =  (0.00)   Depth: 64   00:00:02  2150kN

(Uri, MyTown 01.07.2007)

If you want a case when rybka fails in finding the right move then
Rybka cannot find Nxa7(at least not in a short time) when glaurung has no problem to find it.

2N5/p1k5/Pn1p1p2/1PpPpPp1/2P1P1Pp/7P/8/7K w - - 0 1
Parent - - By Felix Kling (Gold) Date 2007-07-01 12:27 Edited 2007-07-01 12:31
No, Rybka 2.3.2 is stronger in pawn endings, that's what I "feel" but also what a test that I run shows:

6k1/pp3ppp/4p3/8/8/2P5/PP3PPP/6K1 w - -


I started a tournament (1+0, 50 games) with this position and 2.3.2a won against 2.2n2 with 29-21 (I don't thin 2.3 would produce a completly diffrent result).
Of course this position should be better for white, so the colour was changed after every game...
Parent - By Uri Blass (*****) Date 2007-07-01 16:50
It may be stronger in pawn endgame that are not obvious to understand but her evaluation is clearly wrong in some pawn endgames that are easy to understand.

Uri
Parent - By plicocf (***) Date 2007-07-01 16:14
Uri, I learned with you that 1 position don´t proves nothing.
By the way, Rybka it´s now fonding the move of your game,
Kf6.

Paulo Soares
Parent - By Gaмßito (****) Date 2007-07-01 17:17
Deep Fritz 10 understands that to take the knight on b6 is a mistake. It understands relatively quick that it is much better to take the pawn on a7.  
 
Note also that in this position Deep Fritz seems to have the same ''problem'' that Rybka seems to have in some positions too. For a long time it seems completely to be stuck into depth 21, and only after almost 29 minutes, it reach to more than 30 plies easily.  

2N5/p1k5/Pn1p1p2/1PpPpPp1/2P1P1Pp/7P/8/7K w - - 0 1


Analysis by Deep Fritz 10:

1.Nxb6 Kxb6 2.Kg2 Kc7 3.Kf3 Kc8 4.Ke3 Kb8 5.Kd3 Kc7 6.Ke3 Kb6
  +-  (1.88)   Depth: 9/15   00:00:00  19kN
1.Nxb6 Kxb6 2.Kg2 Kc7 3.Kf3 Kc8 4.Ke3 Kb8 5.Kd2 Kc7 6.Ke2
  +-  (2.00)   Depth: 10/15   00:00:00  36kN
1.Nxb6 Kxb6 2.Kg2 Kc7 3.Kf3 Kc8 4.Kf2 Kc7 5.Ke2 Kb6
  +-  (2.00)   Depth: 11/16   00:00:00  42kN
1.Nxb6 Kxb6 2.Kg2 Kc7 3.Kf3 Kc8 4.Kf2 Kc7 5.Ke2 Kb6 6.Kf3 Ka5
  +-  (1.97)   Depth: 12/17   00:00:00  55kN
1.Nxb6 Kxb6 2.Kg2 Kc7 3.Kf3 Kc8 4.Kf2 Kc7 5.Ke2
  +-  (1.95)   Depth: 13/19   00:00:00  118kN
1.Nxb6 Kxb6 2.Kg2 Kc7 3.Kf3 Kc8 4.Kf2 Kc7 5.Ke2 Kb6 6.Kf3
  +-  (1.95)   Depth: 14/21   00:00:00  186kN
1.Nxb6 Kxb6 2.Kg2 Kc7 3.Kf3 Kc8 4.Ke3 Kc7 5.Kd3 Kc8 6.Kc3 Kb8 7.Kc2
  +-  (2.09)   Depth: 15/23   00:00:00  291kN
1.Nxb6 Kxb6 2.Kg2 Kc7 3.Kg1 Kb8 4.Kf1 Kc8 5.Ke1 Kb8 6.Kd1 Kc7 7.Kd2
  +-  (2.23)   Depth: 16/25   00:00:00  370kN
1.Nxb6 axb6 2.Kg2 Kb8 3.Kf1 Ka8 4.Ke2 Kb8 5.Ke1 Ka7 6.Kd1 Kb8
  +-  (2.47)   Depth: 17/28   00:00:01  719kN
1.Nxb6 axb6 2.Kg2 Kb8 3.Kf1 Ka8 4.Ke2 Kb8 5.Ke1 Ka7 6.Kd1 Kb8 7.Ke2
  +-  (2.47)   Depth: 18/30   00:00:07  1016kN
1.Nxb6 axb6 2.Kg2 Kb8 3.Kf1 Ka8 4.Ke2 Kb8 5.Ke1
  +-  (2.47)   Depth: 19/32   00:00:07  1527kN
1.Nxb6 axb6 2.Kg2 Kb8 3.Kf1 Ka8 4.Ke2 Kb8 5.Ke1
  +-  (2.43)   Depth: 20/45   00:00:09  2340kN
1.Nxb6 axb6
  +-  (2.14)   Depth: 21/52   00:00:15  8445kN
1.Nxb6 axb6 2.Kg2 Kb8 3.Kf1 Ka8 4.Kf2 Kb8 5.Kg1 Kc8
  ²  (0.47)   Depth: 21/58   00:00:17  9242kN
1.Nxa7
  ²  (0.47)   Depth: 21/58   00:00:20  12163kN
1.Nxa7
  ²  (0.63)   Depth: 21/58   00:00:21  12659kN
1.Nxa7
  ±  (0.94)   Depth: 21/58   00:00:21  13456kN
1.Nxa7
  +-  (1.57)   Depth: 21/58   00:00:29  20340kN
1.Nxa7 Nxc4 2.Kg2 Nd2 3.Nc6 Kb6 4.Nxe5 c4 5.Nd7+ Ka7 6.Kf2 Nxe4+ 7.Ke3 Nc3 8.Nxf6 Nxb5 9.Ne4 c3 10.f6 Kxa6
  +-  (2.90)   Depth: 21/97   00:30:34  1744mN
1.Nxa7 Nxc4
  +-  (2.62)   Depth: 22/43   00:30:42  1752mN
1.Nxa7 Nxc4 2.Kg2
  +-  (2.62)   Depth: 22/53   00:30:46  1756mN
1.Nxa7
  +-  (2.90)   Depth: 23/50   00:31:04  1773mN
1.Nxa7 Nxc4
  +-  (2.62)   Depth: 24/61   00:32:27  1846mN
1.Nxa7 Nxc4 2.Kg2
  +-  (2.62)   Depth: 24/61   00:32:40  1857mN
1.Nxa7
  +-  (2.90)   Depth: 25/54   00:33:58  1920mN
1.Nxa7
  +-  (3.18)   Depth: 25/58   00:34:45  1959mN

(, AMD 01.07.2007)

Regards,
Gambito.
Parent - By Vasik Rajlich (Silver) Date 2007-07-01 16:27
Rybka 2.3.2 has a new algorithm for king and pawn endings. On average, this algorithm seems to be better, but sometimes (like here) it's worse.

Vas
Up Topic Rybka Support & Discussion / Rybka Discussion / Rybka 2.3.2a Problem

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill